Flood Hazard Mapping of Dharla River Floodplain Using Hec-Ras 1D/2D Coupled Model

FLOOD HAZARD MAPPING OF DHARLA RIVER FLOODPLAIN USING HEC-RAS 1D/2D COUPLED MODEL

TASMIA TAZIN

DEPARTMENT OF WATER RESOURCES ENGINEERING BANGLADESH UNIVERSITY OF ENGINEERING AND TECHNOLOGY DHAKA 1000, BANGLADESH

FEBRUARY, 2018

FLOOD HAZARD MAPPING OF DHARLA RIVER FLOODPLAIN USING HEC-RAS 1D/2D COUPLED MODEL

A THESIS SUBMITTED TO THE DEPARTMENT OF WATER RESOURCES ENGINEERING IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF

MASTER OF SCIENCE IN WATER RESOURCES ENGINEERING

BY TASMIA TAZIN

DEPARTMENT OF WATER RESOURCES ENGINEERING BANGLADESH UNIVERSITY OF ENGINEERING AND TECHNOLOGY DHAKA 1000, BANGLADESH

FEBRUARY, 2018

TO MY PARENTS

ACKNOWLEDGEMENTS

It is indeed a great privilege for the author to express her deepest gratitude to her thesis supervisor, Dr. Md. Sabbir Mostafa Khan, Professor, Department of Water Resources Engineering, BUET for giving the unique opportunity to work on such an important topic. His continuous guidance, invaluable suggestions, affectionate encouragement, generous help and invaluable acumen are greatly acknowledged.

Acknowledgements are very due to Dr. A. F. M. Saiful Amin, Professor, Department of Civil Engineering, BUET for his careful review and suggestions. His precious comments, constructive criticism and valuable suggestions contributed greatly to this dissertation.

Author would like to express her indebtedness to Purnima Das and Abdul Hadi Al Nafi Khan for sharing knowledge and ideas on modelling used in this research.

It is also a great pleasure for the author to express his gratefulness to Sarder Udoy Raihan for supporting author during her entire data collection period and for sharing knowledge.

Author would like to thank to the board of members Dr. Md. Mostafa Ali, Head, Department of Water Resources Engineering, BUET; Dr. Md. Abdul Matin, Professor, Department of Water Resources Engineering, BUET and Dr. Maminul Haque Sarker, Deputy Executive Director, Development Centre for Environmental and Geographic Member (External) Information Services (CEGIS) for their valuable comments and suggestions.

The author would like to thank her parents for their encouragement and inspiration. Without their support she would not have finished her M.Sc.

Author is grateful to her husband, Md. Tahmidul Islam for his contribution to this study at various stages of work. She appreciates and admires his patience and encouragement throughout her study.

She also thanks to her sister, brothers, in-laws and other members of her family for their continuous support. Above all, she is grateful to the Almighty Allah for empowering her to bring this thesis to its satisfactory completion.

ABSTRACT

Development of flood hazard map in Dharla River floodplain, located in the north-west zone of Bangladesh using 1D/2D couple hydrodynamic model simulation has been reported. Maps have been developed with data of administrative upazila and landuse pattern of the study area using flood depth as a hydraulic characteristic factor of flood. The hydrodynamic model for mapping were developed using the Hydrologic Engineering Center River Analysis System (HEC-RAS) in concert with HEC-GeoRAS. HEC- GeoRAS set procedures, tools, and utilities for processing Geographic Information Systems (GIS) data by using a graphical user interface on a GIS platform. Automated GIS processing procedures in HEC-GeoRAS provided a useful and expeditious method for repetitive hydraulic model development during analysis of the Dharla River floodplain. Reach length, stream centerline, main channel bank, flow path lines and cross sections have been determined using HEC-GeoRAS. The geometric data has been imported into HEC-RAS using a data exchange format developed by HEC. The resultant water depth exported from HEC-RAS simulations has been processed by HEC-GeoRAS for flood inundation delineation and hazard map generation.

Calibration and verification of the hydrodynamic model were performed in 2013 and 2014 respectively with observed water level data using Manning’s roughness coefficient (n). Model simulation result has showed that 23.8% and 34 % of total study area were inundated under water in 2017 and 1998 respectively. According to the analysis of flood water depth in year 2017 and 1998, it was found that area of F1 (0 m- 0.9 m) was significant from May to September. From the hazard mapping, out of ten upazilas, Lalmonirhat Sadar, Phulbari and Kurigram Sadar along the Dharla River were found to be the most vulnerable to flood hazard. It was also found that Chilmari, Bhurungamari and Kaliganj upazilas which are the outermost upazilas of Dharla River floodplain were very less susceptible to flooding. Considering the agriculture landuse pattern, Boro - Fallow - T.aman was found to be the most vulnerable crop and Rabi Crop - B.aus - Fallow was the less vulnerable crop to the flood events of 2017 and 1998 in the study area. Generally, the study showed that the methodology for river flood analysis using the 1D–2D coupled hydrodynamic model is generic and can be applied to similar geographical conditions.

vii

CONTENTS

Page DECLARATION v ACKNOWLEDGEMENTS vi ABSTRACT vii CONTENTS viii LIST OF FIGURES xi LIST OF TABLES xiv LIST OF ABBREVIATIONS xv

Chapter 1 : INTRODUCTION 1.1 General 1 1.2 Geophysical Significance of Bangladesh 3 1.3 Major River Systems 3 1.4 Importance and Significance of the Study 4 1.5 Objectives 7 1.6 Organization of this Dissertation 7

Chapter 2 : FLOOD AND FLOOD MANAGEMENT 2.1 General 9 2.2 Natural Hazard 9 2.3 Flood Hazard Map 9 2.4 Definition of Flood and its Types 10 2.4.1 Coastal (Surge) Flood 11 2.4.2 Fluvial (River Flood) 11 2.4.3 Pluvial (Surface) Flood 12 2.5 Floods in Study Area 12 2.6 Causes of Flooding 14 2.7 Statistics of Flooding in Bangladesh 15 2.8 Flood History in Study Area 17 2.9 Flood Mitigation Strategies 21 2.9.1 Structural Measures 21

viii

Page 2.9.2 Non-Structural Measures 22

Chapter 3 : PREVIOUS STUDIES 3.1 General 23 3.2 Study on Hazard Mapping 23 3.3 Uses of HEC-RAS in Floodplain Inundation Modeling 26 3.4 Flood Study using Satellite Images 32

Chapter 4 : SALIENT FEATURES OF THE MODEL 4.1 General 34 4.2 HEC-RAS 34 4.2.1 User Interface 34 4.2.2 Hydraulic Analysis Components 35 4.2.3 Data Storage and Management 37 4.2.4 Graphics and Reporting 37 4.2.5 RAS Mapper 38 4.3 Theoretical Basis for One Dimensional and Two Dimensional 38 Hydrodynamic Calculation 4.3.1 1D Steady Flow Water Surface Elevation 38 4.3.2 1D/2D coupled Hydraulic Modelling 44 4.4 Geographic Information System 44 4.4.1 General 44 4.4.2 Data Models 45 4.5 HEC-GeoRAS 46 4.5.1 General 47 4.5.2 Overview of Requirements 47 4.5.3 Software Requirements 47 4.5.4 Data Requirements 47 4.5.5 Getting Started 47 4.5.6 HEC-GeoRAS Menus 48

Page

Chapter 5: METHODOLOGY AND MODEL SETUP 5.1 General 49 5.2 Study Area 49 5.3 Overall View of the Methodology 51 5.3.1 Preparation Phase 52 5.3.2 Execution Phase 59 5.3.3 Comparison and Hazard Mapping Phase 76

Chapter 6: RESULT AND DISCUSSION 6.1 Calibration of HEC-RAS Model 80 6.2 Validation of HEC-RAS Model 82 6.3 Qualitative Comparison between Model Simulated and 83 Observed Flood Map ( Satellite Image) 6.3.1 Qualitative Comparison between Model and Observed 83 Satellite Image (28 July 2017) 6.4 Analysis of Model Simulated Flood, Year 2017 85 6.4.1 Flood Inundation Map and Depth Analysis 85 6.4.2 Flood Affected frequency 93 6.4.3 Development of Hazard Map 94 6.5 Analysis of Historical Flood Event, 1998 102 6.5.1 Flood Inundation Map and Depth Analysis 102 6.5.2 Development of Hazard Map 111

Chapter 7: CONCLUSIONS AND RECOMMENDATIONS 7.1 Conclusions 118 7.2 Recommendations 120 REFERENCES 121 APPENDICES A Features of Model 128 B Morphological Data 133

LIST OF FIGURES

Page Figure 1-1 Location of Bangladesh 4 Figure 1-2 Basin map of Ganges, Brahmaputra and Meghna River 5 Figure 1-3 Major river floodplains of Bangladesh 6 Figure 2-1 Types of flood 13 Figure 2-2 Discharges in the Ganges, Brahmaputra and Meghna River 15 Figure 2-3 Comparison of hydrograph on Dharla at Kurigram station 19 Figure 2-4 Present flood status 21 Figure 4-1 Representation of terms in the energy equation 39 Figure 4-2 Application of momentum principle 41 Figure 5-1 Identification of study area 50 Figure 5-2 Summary of steps of methodology in flow chart 51 Figure 5-3 Locations of discharge and water level station of Dharla River 53 Figure 5-4 Locations of cross-section of Dharla River 54 Figure 5-5 Digital Elevation (DEM) of Bangladesh 56 Figure 5-6 Digital Elevation (DEM) modification 57 Figure 5-7 (a) Superimposed shape file on modified DEM 58 Figure 5-7 (b) Clipped DEM of shape file 58 Figure 5-7 (c) Dem of study area 58 Figure 5-7 (d) Raster to TIN generation of the study area 58 Figure 5-8 (a) River centerline and bank line of Dharla River 61 Figure 5-8 (b) River flow paths of Dharla River 61 Figure 5-9 1D geometric features of Dharla river 64 Figure 5-10 Locations of boundary condition 65 Figure 5-11 (a) Upstream boundary condition for calibration 2013 66 Figure 5-11 (b) Downstream boundary condition for calibration 2013 66 Figure 5-12 (a) Upstream boundary condition for validation 2014 67 Figure 5-12 (b) Downstream boundary condition for validation 2014 67 Figure 5-13 Location of model calibration and validation 69 Figure 5-14 2D flow area computational mesh 71 Figure 5-15 Introduction of lateral structure 72

Page Figure 5-16 Introduction of boundary condition line 75 Figure 6-1 Observed and simulated stage hydrograph from 1st January, 2013 81 to 1st January, 2014 Figure 6-2 Statistical parameter of unsteady flow calibration, 2013 81 Figure 6-3 Observed and simulated stage hydrograph from 1st January, 2014 82 to 1st January, 2015 Figure 6-4 Statistical parameter of unsteady flow validation, 2014 83 Figure 6-5 Qualitative comparison between model flood map and satellite 84 image Figure 6-6 Flood inundation map developed by model simulation at Dharla 86 River floodplain on 10 May 2017 Figure 6-7 Flood inundation map developed by model simulation at Dharla 87 River floodplain on 10 June 2017 Figure 6-8 Flood inundation map developed by model simulation at Dharla 88 River floodplain on 10 July 2017 Figure 6-9 Flood inundation map developed by model simulation at Dharla 89 River floodplain on 10 August 2017 Figure 6-10 Flood inundation map developed by model simulation at Dharla 90 River floodplain on 20 September 2017 Figure 6-11 Trend of model simulated inundation area at Dharla River 92 floodplain in 2017 Figure 6-12 Inundated area according to inundation depth, 2017 90 Figure 6-13 Flood affected frequency map 93 Figure 6-14 Administrative unit map of study area 96 Figure 6-15 Hazard map on administrative unit on 10 May 2017 96 Figure 6-16 Hazard map on administrative unit on 10 June 2017 97 Figure 6-17 Hazard map on administrative unit on 10 July 2017 97 Figure 6-18 Hazard map on administrative unit on 10 August 2017 98 Figure 6-19 Hazard map on administrative unit on 20 September 2017 98 Figure 6-20 Observation of hazard rank on administrative unit (upazila) of 99 study area Figure 6-21 Agricultural landuse map 100 Figure 6-22 Crops pattern in the study area 101 Figure 6-23 Hazard rank on agricultural landuse in the study area in 2017 101

xii

Page Figure 6-24 Flood inundation map developed by model simulation at Dharla 104 River floodplain on 10 May 1998 Figure 6-25 Flood inundation map developed by model simulation at Dharla 105 River floodplain on 10 June 1998 Figure 6-26 Flood inundation map developed by model simulation at Dharla 106 River floodplain on 10 July 1998 Figure 6-27 Flood inundation map developed by model simulation at Dharla 107 River floodplain on 10 August 1998 Figure 6-28 Flood inundation map developed by model simulation at Dharla 108 River floodplain on 20 September 1998 Figure 6-29 Trend of model simulated inundation area at Dharla River 109 floodplain in 1998 Figure 6-30 Inundated area according to inundation depth, 1998 110 Figure 6-31 Administrative unit map of study area 112 Figure 6-32 Hazard map on administrative unit on 10 May 1998 112 Figure 6-33 Hazard map on administrative unit on 10 June 1998 113 Figure 6-34 Hazard map on administrative unit on 10 July 1998 113 Figure 6-35 Hazard map on administrative unit on 10 August 1998 114 Figure 6-36 Hazard map on administrative unit on 20 September 1998 114 Figure 6-37 Observation of hazard rank on administrative unit (upazila) of 115 study area Figure 6-38 Crops pattern in the study area 116 Figure 6-39 Hazard rank on agricultural landuse in the study area in 1998 117

xiii

LIST OF TABLES

Page Table 2-1 Year-wise flood affected area in Bangladesh 17 Table 2-2 Impact scenario of flood on 28 July, 2016 19 Table 2-3 Summary of flood impact August, 2017 20 Table 2-4 Structural measures for flood 21 Table 5-1 Summary of data type 52 Table 6-1 Model evaluation parameters 82 Table 6-2 Inundation area on different dates of 2017 91 Table 6-3 Calculation of flood area according to inundation depth, 2017 92 Table 6-4 Hazard rank on administrative unit (upazila), 2017 99 Table 6-5 Inundation area on different dates of 1998 109 Table 6-6 Calculation of flood area according to inundation depth, 1998 110 Table 6-7 Hazard rank on administrative unit (upazila), 1998 115

xiv

LIST OF ABBREVIATIONS

ArcGIS Arc Geographic Information System BTM Bangladesh Transverse Mercator BWDB Bangladesh Water Development Board DEM Digital Elevation Model DTM Digital Terrain Model DL Danger Level ESRI Environmental Systems Research Institute FFWC Flood Forecasting and Warning Centre FCDI Flood Control, Drainage and Irrigation GIS Geographic Information System GBM Ganges- Brahmaputra- Meghna GUI Graphical User Interface HEC Hydrologic Engineering Center HEC-RAS Hydrologic Engineering Centre-River Analysis System HEC-GeoRAS Hydrologic Engineering Centre-Geospatial River Analysis System IPCC Intergovernmental Panel on Climate Change LANDSAT Land Remote-Sensing Satellite (System) MSL Mean Sea Level NASA National Aeronautics and Space Administration NSE Nash-Sutcliffe Efficiency PWD Public Works Datum RAS River Analysis System RS Remote Sensing SRTM Shuttle Radar Topographic Mission TIN Triangular Irregular Networks USGS United States Geological Survey USACE United States Army Corps of Engineers WL Water Level RL Reduce Level

xv Chapter One INTRODUCTION

1.1 GENERAL

In many regions and countries, flood is the most devastating natural hazard. Flood affects the social and economic aspects of the population (Smith, 1999) and claims more lives than any other natural phenomena (Dilley et al., 2005). Frequency with which ﬂood occurs is increasing in many regions of the world (Ahmad et al., 2010) and eventually it has become a major concern around the globe.

In the past century, changing climate is quite convincing based on several studies (Carrier et al., 2016; IPCC, 2014) which has led to increasing temperature in some places while increasing precipitation at the other places (Kalra and Ahmad, 2012, 2011). Increased precipitation results elevation of stream flow. In addition to climate change, the changes in land use and urbanization increase the non-pervious area resulting in increasing the runoff from the watershed by reducing the infiltration (Thakur et al., 2017). The flood events are accompanied by the change in land use and intensification of precipitation due to climate change (Thakur et al., 2017). According to IPCC (2013), Bangladeshis are highly vulnerable to climate changes where both monsoon rainfall and sea level will be raised. Because of increased monsoon rainfall and raised sea level, flood inundation will be affected. Bangladesh is under sub-tropical monsoon climate where annual average precipitation is 2,300 mm, varying from 1,200 mm in the north-west to over 5,000 mm in the north-east (FFWC, 2015). The country is mostly flat with few hills in the southeast and the north-east part (Rahman, 2015). It consists of the flood plains of the Ganges, the Brahmaputra and the Meghna rivers and their numerous tributaries and distributaries. The Ganges, Brahmaputra and Meghna river systems together, drain the huge runoff generated from large area with the highest rainfall areas in the world (FFWC, 2015). As a low-lying country, at least, 20 % areas are flooded every year and in case of severe flood 68% areas are inundated in Bangladesh (Disaster Management Bureau, 2010).

The flood hazard problem in recent times is getting more and more frequent and acute due to growing population size and human socioeconomic activities in the floodplain at an ever-increasing scale (Rahman, 2015). Monsoon flood inundation of about 20% to Chapter 1 25% area of the country is assumed beneficial for crops, ecology and environment, inundation of more than that causing direct and indirect damages and considerable inconveniences to the population (FFWC, 2015). Majority of flood disaster’s victims are poor people, who suffer most and are the first casualty of such incidents (WWAP, 2006). Bangladesh has experienced floods of a vast magnitude in 1974, 1984, 1987, 1988, 1998, 2000 and 2004 (FFWC, 2005). Floods of 1988, 1998 and 2004 inundated about 61%, 68 % and 38% of the total area of the country, respectively (Rahman et al., 2014).

Floods are the most significant natural hazard causing suffering to a large number of people and damage to property in Bangladesh (Rouf, 2015). Different reports estimate that the flood damage was US $ 1.4, 2.0, 2.3, and 1.1 billion in the 1988, 1998, 2004 and 2007 severe flood’s year in Bangladesh respectively. Recent catastrophic floods took place in 1988, 1998, 2004, and 2007 causing losses from one to over two million metric tons of rice, or 4–10 % of the annual rice production (Islam et al., 2010).

For the purpose of flood management there are various options that have been long practiced in Bangladesh. Engineered structural measurement options being the principal strategy for the mitigation of flood damage provided some benefits, specially increase in agricultural production at earlier period. The issues of flood management should be considered from different angles of improvement of quality of life, impact on physical environment, socio-economic condition and environmental preservation. In Bangladesh it is being practiced some structural measures such as Flood Embankment, Channel Improvement, River Training, Coastal Embankment to combat the flood sufferings. Among these structural measures, construction of embankment is most popular and very old practice. With the experience over the last few decades, it was observed that the structural measures do not usually bring only blessings. They also have adverse effect such as rise in bed levels and obstruction to drainage (Islam et al., 2010). Flood Forecasting and Warning System as a secondary strategy started from early ‘70s contributed to the improvement of the capacity for flood preparedness and mitigation of flood losses as a non-structural measure (Hossain, 2015). Importance of this strategy has been realized after the floods of 1987, ’88, ’98. This option consists of the Flood Plain Zoning & Management; Policies for Infrastructure Planning and Development in the Floodplains; Flood Proofing; Disaster Preparedness & Response Planning and Flood Forecasting and Warning. In 1972, the Flood Forecasting and Warning Center (FFWC)

Introduction was founded under the Bangladesh Water Development Board (BWDB) to work as the national focal point with respect to flood monitoring, forecasting, warning, and dissemination of information (Rouf, 2015). Currently, FFWC utilizes advanced software such as ‘‘MIKE11’’ and ‘‘Flood Watch’’ to provide real-time forecasts and warning services during the monsoon season (Islam and Tsujimoto, 2012).

As a non-structural measurement to mitigate the effect of flood, numerous mathematical modeling have been introduced all over the world for providing cost effective, reliable and crucial mechanism for flood preparedness, damage control and management of flood disasters by early warning system. In this study Hydrologic Engineering Center River Analysis System (HEC-RAS) has been used to develop flood inundation map by hydraulic modeling analysis in concert with HEC-GeoRAS. HEC-GeoRAS is a set of procedures, tools, and utilities for processing geographic information systems (GIS) data in Arc GIS, using a graphical user interface. Finally, geographical information system (GIS) has been used to develop hazard map.

1.2 GEOPHYSICAL SIGNIFICANCE OF BANGLADESH

Bangladesh lies approximately between 20o30ʹ and 26o40ʹ north latitude and 88o03ʹ and 92o40ʹ east longitude. It is one of the biggest active deltas in the world with an area of about 1,47,570 sq-km. India borders the country in west, north and most part of east. The Bay of Bengal is in the south, Myanmar borders part of the south-eastern area (Figure 1- 1). The country is mostly flat with few hills in the southeast and the northeast part. Generally ground slopes of the country extend from the north to the south and the elevation ranging from 60 meters to one meter above Mean Sea Level (MSL) at the boundary at Tentulia (north) and at the coastal areas in the south. The country consists of the flood plains of the Ganges, the Brahmaputra and the Meghna rivers and their numerous tributaries and distributaries.

1.3 MAJOR RIVER SYSTEMS

It has 405 rivers including 57 transboundary rivers, among them 54 originated from India including three major rivers the Ganges, the Brahmaputra and the Meghna (FFWC, 2015). Ganges, Brahmaputra and Meghna river systems together, drain the large runoff generated from large area with the highest rainfall areas in the world. Their total

Chapter 1

Figure 1-1: Location of Bangladesh [Source: www.escola.britannica.com, accessed on 13th January, 2018] catchment area is approximately 1.72 million sq km of which only about 7.5% lies in Bangladesh and the rest, 92.5% lies outside the territory (Figure 1-2). It is assumed that an average flow of 1,009,000 million cubic meters passes through these river systems during the monsoon season. Most of the rivers are characterized by having sandy bottoms, flat slopes, substantial meandering, banks susceptible to erosion and channel shifting. The river system of Bangladesh is one of the most extensive in the world, and the Ganges and the Brahmaputra are amongst the largest rivers on earth in terms of catchment size, river length and discharge.

1.4 IMPORTANCE AND SIGNIFICANCE OF THE STUDY

In this study, hydraulic model is generated to translate stream flow to water level conditions. Such model is useful in forecasting the water level conditions of large rivers where sufficient lead-time is accorded through translation of upstream flow hydrograph to downstream communities at risk. This model has been interfaced with Geographical Information System (GIS) to provide dynamic water level conditions on maps of communities. The forecast product will be quite useful to communities and emergency

Introduction organizations, as it will provide precise information about areas that will be inundated and time of occurrence.

Figure 1-2: Basin map of Ganges, Brahmaputra and Meghna River [Source: www.lahistoriaconmapas.com, accessed on 13th January, 2018]

Present hydrodynamic models are available in 1D, 2D and 1D/2D coupled hydrodynamic form which allow the simulation of different flood scenarios (Quirogaa et al., 2016). These numerical models are important tools for understanding flood events, flood hazard assessment and flood management planning. In addition, HEC-RAS(1D) was used widely to develop flood inundation map in many studies (Hazarika, 2007; Hicks and Peacock, 2005). Where 1D modeling approaches could be useful in some contexts, mainly for artificial channels, it presents several limitations for overflow analysis (Srinivas et al., 2009). When water begins to overflow, it becomes a 2D phenomenon. So, in this study flood inundation has been done using a combined one dimensional (1D) and two dimensional (2D) hydrodynamic model which includes flood plains as 2D part and river as 1D part. Main advantage of 1D/2D coupled models is the similarity between model behavior and physical behavior (Moore, 2011). For Koiliaris River, China, the combined 1D/2D HEC- RAS model performed better than the 1D HEC- RAS model for a specific study reach by using topographic data at a high spatial resolution (Patel et al., 2017).

Chapter 1 Delineation of flood plain and development of hazard map may help in planning and management of those flood plain areas of the Dharla River to reduce the future probable hazard through early warning system and technical approach. Model simulated floodplain

Figure 1-3: Major river floodplains of Bangladesh (Alam et al., 1990 ) mapping and analysis will provide more effective and standardized results and save time and resources. The outcome of this study will help the planner to prepare river flood warning maps to reduce the sufferings of people, damage of crops and vegetation, and destruction of infrastructures.

Introduction It has been observed that the floodplains of the major rivers (the Ganges, the Brahmaputra and the Meghna) have been established (Figure 1-3) but the floodplains of their numerous tributaries and distributaries have not been yet developed.

1.5 OBJECTIVES

The objective of this study is to show the capacity of HEC-RAS 1D/2D coupled model to reproduce flood stages and flood inundation of the Dharla River floodplain. Specific objectives of research are as follows i. Calibration and validation of HEC-RAS 1D Model ii. To setup HEC-RAS 1D/2D coupled model to generate flood inundation map iii. Qualitative comparison between developed flood inundation map and observed flood map iv. To generate a hazard map incorporating land use pattern

Expected outcome of the researches are as follows: i. A Calibrated hydrodynamic model of the Dharla River will be generated. ii. A model will be developed which will be useful in the planning, designing, operating and maintaining of flood control structures. iii. Monsoon flood inundation map of the Dharla River floodplain will be produced. iv. Hazard map will be generated which will be useful in the context of management purpose in the study area, an agricultural growth center of Bangladesh. v. Applicability of the HEC-RAS 1D/2D coupled model and HEC open-source models in modeling flood inundation to the Dharla River floodplain which can be a basis for application to other floodplains having similar characteristics.

1.6 ORGANIZATION OF THIS DISSERTATION

This dissertation has been organized in seven chapters as follows:

Chapter One provides information about the geophysical physical significance of Bangladesh, it’s major river systems, importance and significance of the study and objectives of this dissertation.

Chapter 1 Chapter Two contains details of flood and flood management in Bangladesh. This chapter introduces the definition of hazard and flood hazard. It also discusses about flood types, types of flood occurs in study area, causes of flooding, flood history in Bangladesh and in the study area and finally the flood mitigation strategies which are adopted.

Chapter Three briefly describes findings from previous studies of other authors about flood hazard mapping, flood inundation modeling and uses of satellite images. Collected and reviewed thesis reports, journals, books, tools, user manuals and a summary of their finding are outlined here.

Chapter Four discusses about the salient features of models used in this study. User interfaces, data storage and management, reporting capabilities of the models have been discussed in this chapter. Theoretical background of simulation has also been presented.

Chapter Five describes the methodology and the model setup of the study. This chapter discussed the methodology from the data collection phase to the flood hazard mapping phase. In middle of the methodology, details of preparation phase and execution phase have also been presented including introduction of the study area.

Chapter Six describes the results and findings of this study. Performance of calibration and validation of model have been presented here. This chapter shows qualitative comparison between model simulated flood maps and observed flood maps. Analysis of model simulated inundation maps and the results of developed hazard map have been also discussed in this chapter.

Finally, Chapter Seven draws conclusions from analysis of results and then made recommendations based on the analysis from this study.

Chapter Two FLOOD AND FLOOD MANAGEMENT

2.1 GENERAL

Flood hazard is the most common natural disasters that affect societies around the world. Dilley et al. (2005) estimated that more than one-third of the world’s land area is flood prone affecting 82 percent of the world’s population. In this chapter discussion has been done about natural hazard, flood hazard map and flood types. This chapter is also directed to an overview of the present situation of flooding as well as flood management measures that is practiced in Bangladesh.

2.2 NATURAL HAZARD

Hazard is something that is dangerous and it causes damage to humans, property, or the environment. Natural hazard is the probability of occurrence of a potentially damaging phenomenon within a specified period of time and where risk is the actual exposure of something of human value to a hazard and is often regarded as the combination of probability and loss within a given area (Bhuiyan, 2014). We may define hazard as a potential threat to humans and their welfare and risk as the probability of a specific hazard consequence. When large numbers of people exposed to hazard are killed, injured or damaged in some way, the event is termed as a disaster. Hazards associated with flooding can be divided into primary hazards that occur due to contact with water, secondary effects that occur because of the flooding, such as disruption of services, health impacts such as famine and disease, and tertiary effects such as changes in the position of river channels.

2.3 FLOOD HAZARD MAP

A hazard map is a map that highlights areas that are affected by or vulnerable to a particular hazard. It is typically created for natural hazards, such as earthquakes, volcanoes, landslides, flooding and tsunamis. Hazard maps help prevent serious damage and deaths (Udono and Sah, 2002). Chapter 2 Flood hazard mapping is a vital component for appropriate land use planning in flood prone areas. It creates easily-read, rapidly-accessible charts and maps which facilitate the administrators and planners to identity areas of risk and prioritize their mitigation/response efforts as a non structural measure.

Geographic Information Systems (GIS) are frequently used to produce flood hazard maps. They provide an effective way of assembling information from different maps and digital elevation (Sanyal and Lu, 2004). Using GIS, the extent of flooding can be calculated by comparing local elevations with extreme water levels. Flood hazard maps can be developed using land cover, elevation, physiographic and geological features and drainage network data. Flood-affected frequency and flood depth can be used as hydraulic components. Hazard index has to be assigned according to inundation depth. But other factors such as frequency of flood, duration of flood, etc. should be considered (Islam and Sado, 2000).

In order to understand the flooding and flood management, it is better having looked into the land types that will be helpful to delineate hazard categories (Hossain, 2013). These are: i. Medium highland, F1: land which is normally flooded up to 90 cm deep during the flood season ii. Medium lowland, F2: land which is normally flooded between 90 cm and 180 cm deep during the flood season iii. Lowland, F3: land which is normally flooded between 180 cm and 300 cm deep during the flood season iv. Very lowland, F4: land which is normally flooded deeper than 300 cm during the flood season

2.4 DEFINITION OF FLOOD AND ITS TYPES

Flooding is the most common environmental hazard worldwide. This is due to the vast geographical distribution of river floodplains and low-lying coastal areas. Flood can be described as an overflow of water onto normally dry land. The inundation of a normally dry area caused by rising water in an existing waterway, such as a river, stream, or drainage ditch. Ponding of water at or near the point where the rain fell. This water comes from the sea, lakes, rivers, canals or sewers. It can also be rainwater.

Flood and Flood Management There are several different kinds of flood, and each one bears a different impact in terms of how it occurs, the damage it causes, and how it is forecasted. Following are the brief description of the types of flood.

2.4.1 Coastal (Surge) Flood

A coastal flood, as the name suggests, occurs in areas that lie on the coast of a sea, ocean, or other large body of open water. It is typically the result of extreme tidal conditions caused by severe weather. Storm surge produced when high winds from hurricanes and other storms push water onshore is the leading cause of coastal flooding and often the greatest threat associated with a tropical storm. In this type of flood, water overwhelms low-lying land and often causes devastating loss of life and property. The severity of a coastal flood is determined by several factors, including the strength, size, speed, and direction of the storm. The onshore and offshore topography also plays an important role. To determine the probability and magnitude of a storm surge, coastal flood models consider this information in addition to data from historical storms that have affected the area, as well as the density of nearby development.

2.4.2 Fluvial (River) Flood

Fluvial, or riverine flooding occurs when excessive rainfall over an extended period of time causes a river to exceed its capacity. It can also be caused by heavy snow melt and ice jams. The damage from a river flood can be widespread as the overflow affects smaller rivers downstream, often causing dams and dikes to break and swamp nearby areas.

There are two main types of riverine flooding: i. Overbank flooding occurs when water rises overflows over the edges of a river or stream. This is the most common and can occur in any size channel — from small streams to huge rivers. ii. Flash flooding is characterized by an intense, high velocity torrent of water that occurs in an existing river channel with little to no notice. Flash floods are very dangerous and destructive not only because of the force of the water, but also the hurtling debris that is often swept up in the flow.

Chapter 2 The severity of a river flood is determined by the amount of precipitation in an area, how long it takes for precipitation to accumulate, previous saturation of local soils, and the terrain surrounding the river system. In flatter areas, floodwater tends to rise more slowly and be more shallow, and it often remains for days. In hilly or mountainous areas, floods can occur within minutes after a heavy rain. To determine the probability of river flooding, models consider past precipitation, forecasted precipitation, current river levels, and temperatures.

2.4.3 Pluvial (Surface) Flood

A pluvial, or surface water flood, is caused when heavy rainfall creates a flood event independent of an overflowing water body. One of the most common misconceptions about flood risk is that one must be located near a body of water to be at risk. Pluvial flooding debunks that myth, as it can happen in any urban area — even higher elevation areas that lie above coastal and river floodplains.

There are two common types of pluvial flooding: i. Intense rain saturates an urban drainage system. The system becomes overwhelmed and water flows out into streets and nearby structures. ii. Run-off or flowing water from rain falling on hillsides that are unable to absorb the water. Hillsides with recent forest fires are notorious sources of pluvial floods, as are suburban communities on hillsides.

Pluvial flooding often occurs in combination with coastal and fluvial flooding, and although typically only a few centimeters deep, a pluvial flood can cause significant property damage.

2.5 FLOODS IN STUDY AREA

In the study area, usually two types of flood occur which cause severe damages every year. These two types are riverine flood and rainfall flood. The area affected by these floods in study area and including other parts of Bangladesh is presented in Figure 2-1. Riverine floods from the spilling of major rivers and their tributaries and distributaries generally rise and fall slowly over 10–20 days or more and can cause extensive damage to property and the loss of life. Depth and extent of floods and associated damage are extensive when the major rivers reach their peaks simultaneously. Rain floods are caused

Flood and Flood Management by high intensity local rainfall of long duration in the monsoon. From year to year, the extent and depth of rainwater flooding varies with the monsoon, depending on the amount and intensity of local precipitation and current water levels in the major rivers that control drainage from the land.

Figure 2-1: Types of flood (Brammer and Khan, 1991)

In Bangladesh, there are some other types of flood that can be encountered such as flash flood, storm surge flood and tidal flood. Area affected by these types of flooding is presented in Figure 2-1. Flash floods occur in the eastern and northern rivers, along the borders of Bangladesh. They are characterized by a sharp rise in

Chapter 2 water level and high water flow velocity, a result from exceptionally heavy precipitation occurring over neighboring hills and mountains in India. Storm surge floods occur in the coastal area of Bangladesh, which consists of large estuaries, extensive tidal flats, and low-lying islands. Storm surges generated by tropical cyclones cause widespread damage to property and the loss of life in coastal area (Rouf, 2015). In case of tidal flood Bangladesh faces semi diurnal tide i.e., two flood tide and two ebb tide in a day in an hour consecutive time interval. Coincidence of heavy rainfall and flood tide occurred water logging in urban area located in coastal part of our country during monsoon. In Chittagong city in Chaktai and Moheshkhali khal catchments such type flood is a common phenomenon in every year monsoon (Rahman, 2015).

2.6 CAUSES OF FLOODING

Floods in Bangladesh occur for number of reasons. The main causes are excessive precipitation, low topography and flat slope of the country; but others include: i. Tectonic uplift of the Himalayas means that erosion rates of sediment increase as the rivers have more potential for erosion. This mass of sediment is dumped in Bangladesh choking the river channels making them more inefficient and reducing hydraulic radius. Sediment is dumped and flooding can occur. ii. Monsoon rainfall – some parts of the Ganges basin receive 500mm of rainfall in a day during the monsoon. iii. Deforestation of the Himalaya – reducing interception rates which means shorter lag time and higher peak discharges. iv. Three massive rivers converge in Bangladesh – the Ganges, Brahmaputra and Meghna – massively swells discharges. Discharges from these major rivers are shown in Figure 2-2. v. Cyclones from the Bay of Bengal cause and contribute to coastal flooding. vi. Snowmelt affects the rivers too, as ice and snow melting from glaciers and mountain peaks in the Himalaya works its way into rivers. vii. The Himalaya also forces relief or geographic rainfall, increasing rainfall totals and then river levels further.

Flood and Flood Management

2.7 STATISTICS OF FLOODING IN BANGLADESH

Flood is a natural phenomenon in Bangladesh and occurs on an annual basis. Rivers are large by global standards, and can inundate over 30% of the land mass at a time.

Brahmaputra-Jamuna

Qmax=100000 m3/s

Qmin=4000 m3/s

Ganges Lower Meghna Qmax=78000 m3/s Qmax=100000 m3/s Qmin=700 m3/s Qmin=4000 m3/s

Figure 2-2: Discharges in the Ganges, Brahmaputra and Meghna River (Source: www.lib.pmo.gov.bd, accessed on 13th January, 2018)

Bangladesh is prone to serious and chronic flooding. Even in an average year 18% of the landmass is inundated and previous floods have affected 75% of the country (as in 1988). 75% of the country is below 10m above sea level and 80% is classified as floodplain as Bangladesh is principally the delta region of South Asia’s great rivers.

Chapter 2 Bangladesh floods on a regular basis, recent notable and catastrophic floods have occurred in 1988 (return period of 1 in every 50 to 100 years), 1998, 2004, 2007 and 2010. Flood statistics have summarized in Table 2-1.

The catastrophic floods of 1987 occurred throughout July and August and affected 57,300 km2 of land, (about 40% of the total area of the country) and was estimated as an once in 30-70 year event. The seriously affected regions were on the western side of the Brahmaputra, the area below the confluence of the Ganges and the Brahmaputra and considerable areas north of Khulna.

The flood of 1988, which was also of catastrophic consequence, occurred throughout August and September. The waters inundated about 82,000 km2 of land, (about 60% of the area) and its return period was estimated at 50–100 years. Rainfall together with synchronization of very high flows of all the three major rivers of the country in only three days aggravated the flood. Dhaka, the capital of Bangladesh, was severely affected. The flood lasted 15 to 20 days.

In 1998, over 75% of the total area of the country was flooded, including half of the capital city Dhaka. It was similar to the catastrophic flood of 1988 in terms of the extent of the flooding. A combination of heavy rainfall within and outside the country and synchronization of peak flows of the major rivers contributed to the river. 30 million people were made homeless and the death toll reached over a thousand. The flooding caused contamination of crops and animals and unclean water resulted in cholera and typhoid outbreaks. Few hospitals were functional because of damage from the flooding and those that were had too many patients, resulting in everyday injuries becoming fatal due to lack of treatment. 700,000 hectares of crops were destroyed, 400 factories were forced to close, and there was a 20% decrease in economic production. Communication within the country also became difficult.

The 2004 floods lasted from July to September and covered 50% of the country at their peak. At the time of the July 2004 floods 40% of the capital, Dhaka was under water. 600 deaths were reported and 30 million people were homeless. 100,000 people alone in Dhaka suffered from diarrhea from the flood waters. Bridges were destroyed, the death toll rose to 750 and the airport and major roads were flooded. This hampered relief efforts. The damage to schools and hospitals was estimated at $7 billion. Rural areas also

Flood and Flood Management suffered, the rice crop was devastated as were important cash crops such as jute and sugar.

In 2007, more than half of Bangladesh was seriously affected by monsoon flooding. Caused by excessive rainfall in catchment areas of Nepal, Bhutan and Northern Indian, floods in July and September affected 13.3 million people – 6 million of them children – in 46 districts.

Table 2-1: Year-wise flood affected area in Bangladesh (FFWC, 2015) Year Flood Affected Area Year Flood Affected Area Year Flood Affected Area Sq km % Sq km % Sq km % 1954 36,800 25 1976 28,300 19 1998 1,00,250 68 1955 50,500 34 1977 12,500 8 1999 32,000 22 1956 35,400 24 1978 10,800 7 2000 35,700 24 1960 28,400 19 1980 33,000 22 2001 4,000 2.8 1961 28,800 20 1982 3,140 2 2002 15,000 10 1962 37,200 25 1983 11,100 7.5 2003 21,500 14 1963 43,100 29 1984 28,200 19 2004 55,000 38 1964 31,000 21 1985 11,400 8 2005 17,850 12 1965 28,400 19 1986 6,600 4 2006 16,175 11 1966 33,400 23 1987 57,300 39 2007 62,300 42 1967 25,700 17 1988 89,970 61 2008 33,655 23 1968 37,200 25 1989 6,100 4 2009 28,593 19 1969 41,400 28 1990 3,500 2.4 2010 26,530 18 1970 42,400 29 1991 28,600 19 2011 29,800 20 1971 36,300 25 1992 2,000 1.4 2012 17,700 12 1972 20,800 14 1993 28,742 20 2013 15,650 10.6 1973 29,800 20 1994 419 0.2 2014 36,895 25 1974 52,600 36 1995 32,000 22 2015 47,200 32 1975 16,600 11 1996 35,800 24

2.8 FLOOD HISTORY IN STUDY AREA

Flood inundation is a phenomenon that results from overtopping or overflowing of floodwater to the river banks. In our country, this situation at a particular place occurs when the river water level exceeds the danger level of that particular place. The danger level of Kurigram station in Dharla river is 26.50 m. Every year, most of flood affected people of Dharla flood plain that is Kurigram, Lalmonirhat districts of Rangpur division are severely facing food and fodder insecurity.

Chapter 2 It has been recorded that during the historical flood event in Bangladesh the water level in Dharla River was above the danger level and caused the floodplain nearby flooded severely. During the devastating floods of 1998 and 1988 in Bangladesh, the water level in Dharla was 27.22 m and 27.25 m respectively. It has been found from FFWC that the maximum recorded water level in Dharla was 27.66 m in the Kurigram station.

The WL of Dharla River at Kurigram registered two distinct peaks during the monsoon 2012, in June and July. It crossed the DL for three times during the monsoon and flowed above DL for 5 days. WL at Kurigram attained peak of 26.74 mPWD on 29th June at 18:00 hours, which was 24 cm above the DL (26.50 m), then fall of WL was recorded and again rise upto 26.68 m (18 cm above the DL) in the 3 rd week of July (FFWC, 2012).

The WL of Dharla River at Kurigram registered its monsoon peak during the monsoon 2014, in last week of August. It crossed the DL once during the monsoon 2014 and flowed above DL for 4 days. WL at Kurigram attained peak of 26.95 m PWD on 28th August at 12:00 hours, which was 45 cm above the DL (26.50 m) (FFWC, 2014). Kurigram is a priority district because approximately 650,000 people are affected and reports indicate that over 120,000 people are presently displaced in 2014. 642, 264 people were affected in Kurigram (38% of the total population of the affected) and 81,091 people were affected in Lalmonirhat (11% of the total population of the district affected) due to flood in 2014 (HCTT, 2014).

The WL of Dharla River at Kurigram registered its monsoon peak during the Monsoon 2015, in 1st week of September. It crossed the DL twice during the monsoon 2015 at the 3rd week of August and then again 1st day of September and flowed above DL for total 13 days. WL at Kurigram attained peak of 26.99 mPWD on 2nd September at 12:00 hours, which was 49 cm above the DL (26.50 m).The significant stations that were above and remained over DLs are Dharla at Kurigram for 13 days (FFWC, 2015).The Figure 2-3 showing water level at Kurigram station in river Dharla for the year 2004, 2007 and 2015.

In the year 2016 the water level was 27.2 m on 30, July in Kurigram station. Below the Table 2-2 shows flood affected information on 28 July, 2016 in Lalmonirhat and Kurigram district (GUK, 2016).

In the year 2017 Kurigram was marooned in floodwater due to incessant rainfall and rise of water levels of Dharla River. According to Nirapad (2017), Dharla River was marked

Flood and Flood Management flowing 22 centimeters above the danger level during the second week of August. People went through undesirable sufferings as many houses have gone under water located in the

27 Danger Level (DL) 26.5

Water Level (m) Level Water 24

2004 23 2007 2015

22 6-Jun 13-Jul 19-Aug 25-Sep Date

Figure 2-3: Comparison of hydrograph on Dharla at Kurigram station

basin of Dharla River. In Lalmonirhat around 25,000 ha of agricultural land damaged. Some 300 fishing ponds were washed away by flood waters across five upazilas. In Kurigram around 42,300 ha of vegetable cultivation is inundated affecting around 300,000 farmers. Table 2-3 shows summary of flood impact of in the year 2017 over Kurigram and Lalmonirhat district.

Table 2-2: Impact scenario of flood on 28 July, 2016 Crop Displaced Inundated Marooned Upazila Union Family People Land family family family (Ha) Lalmonirhat 4 21 34568 172840 354 18203 15700 1890 Kurigram 9 55 146487 1907630 6226 31361 159186 7323

Chapter 2

Table 2-3: Summary of flood impact in August, 2017 No. of Death Death of No. No. of Displaced Displaced of No. Affected Institution Institution Affected Affected Road (km) (km) Affected Road No. of Shelter Center Shelter Center of No. No. of Missing People People Missing of No. No. of Affected Union Union Affected of No. No. of Affected Bridge Bridge Affected of No. No. of Affected People People Affected of No. Affected District Name Name Affected District No. of Damaged House House Damaged of No. No. of Affected Village Village Affected of No. No. of Affected Upazila Upazila Affected of No. No. of Affected Tubewell Tubewell Affected of No. Affected Crops land (Ha) land (Ha) Affected Crops Affected Embankment (km) (km) Affected Embankment

Fully Fully Fully Fully Fully Fully Partially Partially Partially Partially Partially Partially Partially Partially

Kurigram 9 62 724 0 511032 24649 88969 0 50031 22 1 2 684 0 142.5 0 0 23 80 47006 12719

Lalmonirhat 5 35 510 5288 36671 1322 9169 0 31400 6 0 0 366 0 222.7 0 7.5 1 0 0 0

Flood and Flood Management

2.9 FLOOD MITIGATION STRATEGIES

2.9.1 Structural Measures

Considering the issues of securing peoples’ life and property, livelihood, food etc. the Govt. put emphasis on protecting Medium High and Medium Low Lands from floods through construction of embankments. Since 1960s Bangladesh has implemented about 628 nos. of large, medium and small-scale FCDI projects. Total investment was to the tune of US$ 4.0 billion (Hossain, 2003). It provided flood protection to 5.37 million ha of land, which is about 35% of area. A picture flooded, non flooded and flood protected area is shown in Figure 2-4. A picture structural measures works are given in Table 2-4.

Figure 2-4: Present flood status (Hossain, 2003)

Table 2-4: Structural measures for flood Item Quantity Embankment 10,000 km Drainage Channel Improvement 3500 km Drainage Structure 5000 nos. Dam 1 no. Barrage 4 nos Pump House 100 nos. River Closure 1250 nos.

2.9.2 Non-Structural Measures

In spite of all the structural activities, it was found that the people living in the Medium High and Medium Low Lands are not immune to flooding during moderate to extreme flood events. Government considered that the minimizing flood loss through non-

Chapter 2 structural means is also very important. Early warning on flood can save life and property. With this end in view, Flood Forecasting and Warning Centre (FFWC) was established in 1972 with 10 Flood Monitoring Stations on the major river systems. After disastrous floods of 1987 & ‘88 the Government realized the importance of FFWC and took steps to modernize the system. New FFWC model was developed on the basis of Mike-II hydrodynamic model and flood-monitoring stations were increased to 30 in 1996. In 1998 flood FFWC was found to be very useful providing the early warning and information on the flood. With the experience of 1998 flood the Government decided to improve it further to cover all the flood prone areas of the country under real time flood monitoring. A project was under taken from year 2000 to improve the FFWC further. It now covers the entire country with 85 Flood Monitoring Stations and provides real time flood information with early warning for lead-time of 24 and 48 hours. FFWC currently, helping the Government, the disaster mangers and the communities living in the flood prone areas in matters of flood preparedness, preparation of emergency mitigation plan, agricultural planning and rehabilitations.

Chapter Three PREVIOUS STUDIES

3.1 GENERAL

A number of studies have been carried out in the past, particularly concerning the floodplain inundation mapping, flood hazard mapping, flood forecasting and others nonstructural measures. Some of these researches have been summarized in this chapter order to derive proper conception about flooding problem and mitigation managements. A good quality floodplain inundation map has been derived as an outcome presented in the next chapters.

Efforts have been made to collect and review the thesis reports, journals, books, tools, user manuals and a summary of their finding are outlined in the sections below.

3.2 STUDY ON HAZARD MAPPING

Masood (2011) studied flood hazard and risk assessment in mid eastern part old Dhaka of Bangladesh. An inundation map for the mid-eastern Dhaka (37.16 km2), Bangladesh was simulated on the basis of Digital Elevation Model (DEM) data from Shuttle Radar Topography Mission (SRTM) and the observed flood data for 32 years (1972-2004). The topography of the project area has been considerably changed due to rapid land-filling by land developers. The collected DEM data was then modified according to the recent satellite image. The inundation simulation has been conducted using HEC-RAS program for 100 year flood. Both present natural condition and condition after construction of proposed levee (top elevation ranges from 8.60 m to 9.00 m) have been considered for simulation in his study. After simulation, it was revealed that the maximum depth is 7.55 m at the south-eastern part of that area and affected area is more than 50%. Finally, according to the simulation result, a Flood Hazard Map was developed using the software ArcGIS. Moreover, risk map was prepared for this area by conducting the risk assessment.

Bhuiyan (2014) assessed the hazard and vulnerability of riverine flood in Khoksabari union of Sirajganj Town surrounded by a network of rivers namely the Jamuna, the Bangali and the Karatoa which makes the Union vulnerable to flooding using Remote Chapter 3 Sensing (RS) and Geographic Information System (GIS). Flood frequency analysis was carried to assess flooding for different flood magnitudes. Flood inundation maps were prepared based on DEM and satellite image for different risk elements using ILWIS software. LANDSAT satellite images were downloaded and used to develop land use map in the study area. The land use map was used for mapping of settlement and fishery by using ILWIS software. The vulnerability function was developed for preparing vulnerability maps for settlements and fisheries. For the development of vulnerability function, depth-damage relation was developed. Present monetary values of settlement and fishery damage were collected through field survey from actual flood of the study area. Vulnerability functions of settlements and fisheries were used to produce raster- based vulnerability maps. It was found that the Pearson Type III distribution is the best fitted distribution for flood frequency analysis in the study area. In 100-year return period flood, inundation percentage of the total agriculture, settlement, fishery and road areas were 48, 35, 53 and 38, respectively. High land (F0) of the study area was 55%, which was not much inundated in normal monsoon flood. It was found vulnerability scaled as low, very low, moderate, high and very high vulnerable settlement areas to be 17, 12, 3.43, 1.03 and 1.35 percent, respectively for 100-year flood. These correspond to a maximum of 20%, 40%, 60%, 80% and more than 80% damage of the respective settlement areas. In 100-year flood magnitude, low, very low, moderate, high and very high vulnerable fishery areas to be 18, 19, 9, 3 and 4 percent were found. These correspond to a maximum of 20%, 40%, 60%, 80% and more than 80% damage of the respective fishery areas. The results of his study may be useful for future flood damage mitigation plan in the study area.

Pathak et al. (2016) studied on modeling of floodplain inundation for Monument Creek, Colorado. In their study, flood plain map was created for 11.4 km reach of monument creek at Colorado Springs, Colorado. The study presented the approach for terrain modeling and flood plain mapping. It incorporated the hydraulic modeling using HECRAS and terrain modeling using Arc Hydro to create floodplain map. The reach of Colorado Spring was successfully modeled and map was delineated showing the flooded areas along the Monument creek. Flooded areas along the creek were delineated but there were some uncertainties, which could not be overlooked. Change was noticed between the topography of the flood plain and digital elevation model, which could be because of not containing high-resolution terrain data. It was found that high-resolution digital data

Previous Studies of the geometry of river are hence required to create an effective terrain models as the accuracy of hydrologic modeling largely depends upon the accuracy of terrain data used in GIS. Vulnerability assessment of the 1935 flood event as well as 200-year flood magnitude, with the aid of GIS and HECRAS in their study, reveals the risk to which the city of Colorado Springs is exposed. In the city the most drainage structures are designed for 100-year flood level, and as such the structures, if exposed to higher magnitude flood level, will get affected by flood waters. In the future, with the increasing warming of the climate and resulting increase in frequency and intensity of rainfall events, it becomes prudent to model flood plain for future.

Islam and Sado (2000) studied flood hazard assessment using NOAA-AVHRR data with administrative districts, and physiographic, geological, elevation and drainage network data. Flood-affected frequency and flood water depth were essential components for the evaluation of flood hazard in their study. The categories of flood-affected frequency and flood water depth were estimated using NOAA satellite data. Flood hazard rank assessment was undertaken on the basis of land cover classification, physiographic divisions, geological divisions, elevation intervals and administrative districts. All these data and maps were developed in digital form and can be used as a GIS database in other fields. It was showed that 71% of hazard ranks in the area were the same for the best combination of thematic data whether these have been estimated with regard to flood affected frequency or flood water depth, and 75% of the administrative districts fall within the same risk zones when estimated using either flood-affected frequency or flood water depth. Finally, flood risk assessment was generated using both flood hazard maps for the administrative districts of Bangladesh considering the synergistic effect of flood affected frequency and flood water depth. It was also showed that 7.50% of areas were at very high risk and 16.34% were at high risk. The capital city also lied in a high risk area. It was stated in their study that generated flood hazard and flood risk maps might help the responsible authorities to better comprehend the inundation characteristics of the flood plains, the protection of which is their responsibility. Finally, these types of flood hazard and risk map in digital form could be used as a database to be shared among the various government and non-government agencies responsible for the construction and development of flood defense.

Chapter 3 Dewan et al. (2007) illustrated the development of flood hazard and risk maps in Greater Dhaka of Bangladesh using geoinformatics. Multi-temporal RADARSAT SAR and GIS data were employed to delineate flood hazard and risk areas for the 1998 historical flood. Flood affected frequency and flood depth were estimated from multi-date SAR data and considered as hydrologic parameters for the evaluation of flood hazard using land-cover, geomorphic units and elevation data as thematic components. Flood hazard maps were created by considering the interactive effect of flood frequency and flood water depth concurrently. Their analysis revealed that a major portion of Greater Dhaka was exposed to high to very high hazard zones while a smaller portion (2.72%) was free from the potential flood hazard. The results of their generated flood risk map according to administrative division showed that 75.35% of Greater Dhaka was within medium to very high risk areas of which 53.39% of areas are believed to be fully urbanized by the year 2010.

Mani et al. (2014) studied on flood hazard assessment with multiparameter approach derived from coupled 1D and 2D hydrodynamic flow model. Hydrodynamic flow modeling was carried using a coupled 1D and 2D hydrodynamic flow model in northern India where an industrial plant is proposed. The model simulated two flooding scenarios, one considering the flooding source at regional/catchment level and another considering all flooding sources at local level. For simulating flooding scenario due to flooding of the upstream catchment, the probable maximum flood (PMF) was routed in the main river and its flooding impact was studied at the plant site, while at the local level flooding, in addition to PMF in the main river, the probable maximum precipitation was considered at the plant site and breaches in the canals near the plant site. The flood extent, depth, level, duration and maximum flow velocity were computed. Three parameters namely the flood depth, cross product of flood depth and velocity and flood duration were used for assessing the flood hazard, and a flood hazard classification scheme was proposed. Flood hazard assessment for flooding due to upstream catchment and study on local scale facilitates on their study the determination of plinth level for the plant site and helps in identifying the flood protection measures.

3.3 USES OF HEC-RAS IN FLOODPLAIN INUNDATION MODELLING

Betsholtz and Nordlöf (2017) studied that hydraulic models can be useful to predict the consequences of flooding events. Under this project, three hydraulic models were 26

Previous Studies constructed using the software HEC-RAS, and compared through a case study on Höje river catchment. The models include (i) a 1-dimensional (1D) model, where river and floodplain flow was modeled in 1D, (ii) a coupled 1D-2D model, where river flow was modeled in 1D and floodplain flow was modeled in 2D, and (iii) a pure 2D model, where river and floodplain flow was modeled in 2D. Important differences between data requirements, pre-processing, model set-up and results were highlighted and summarized, and a rough guide that may be used when deciding the appropriate type of model for a project, was presented. In addition to that, the sub-grid technique generally used in 2D HEC-RAS modeling was studied by investigating the influence of computational mesh structure and coupling between 1D and 2D areas. The results showed that all three models could successfully reproduce a historic flooding event. The 2D and 1D-2Ds model could also provide more detailed information regarding flood propagation and velocities on the floodplain. The results from the 2D mesh analysis showed that model result was very sensitive to mesh alignment along barriers. In rural floodplains with clear barriers, computational cell alignment was more important than computational cell size. With regards to the 1D-2D model, the results showed that the parameters describing the coupling between the 1D and 2D domain had large impact on model results.

Chow et al. (1988) presented a straightforward approach for processing output of the HEC-RAS hydraulic model, to enable two and three dimensional floodplain mapping and analysis in the ArcView geographic information system. The methodology was applied to a reach of Waller Creek, located in Austin, Texas. A planimetric floodplain view was developed using digital orthophotography as a base map. Moreover, synthesized a digital terrain model from HEC-RAS cross-sectional coordinate data and a digital elevation model of the study area. Finally, the resulting surface model was created, which provided a good representation of the general landscape and contains additional detail within the stream channel. Overall, the results of the research indicated that GIS is an effective environment for floodplain mapping and analysis.

Rouf (2015) worked on flood inundation map in a low-lying riverine flood prone area of Bangladesh at Sirajgonj district. In her study, weather forecast model coupled with a hydrologic model and resulted hydrodynamic model for predicting floods in Jamuna River at Sirajgonj district were used. Weather Research and Forecasting model (WRF 3.0) was used to predict rainfall over the basin with lead-time of 6 days in her study. At

Chapter 3 first hydrological model HECHMS 4.0 (Hydrologic Engineering Center - Hydrologic Modeling System) was calibrated and validated with discharge at Bahdurabad station which was derived from Global Weather rainfall Data with Clark’s Unit Hydrograph transformation method. Output from the WRF model with hydrologic model HEC-HMS was then coupled. Before using the model for prediction, the HEC-HMS model with WRF output was calibrated by observed discharge at Bahdurabad station. WRF predicted rainfall for 1st June 2014 to 9th October 2014 to HEC-HMS were introduced and the generated river discharges of sub basin to the HECRAS 4.1.0 (Hydrologic Engineering Center-River Analysis System) hydrodynamic model were ingested for water profile computations along the Jamuna River. This hydrodynamic model was again calibrated and validated with observed water level at Bahdurabad station. The output of calibrated and validated hydrodynamic HEC-RAS model was exported to ArcMap 10.1 where it was visualized as a flood inundation map with the use of the extension of HEC-GeoRAS. These maps have been developed for each day integrating the Digital Elevation Model (DEM) data of Shuttle Radar Topographic Mission (SRTM) and interpolation of water level height obtained from HEC-RAS output at different cross-sections. Observing these maps, it was found that Sirajgonj district suffered highest inundation in the month of August. On 24 August 2014, Shahjadpur Thana, Ullah Para Thana, Tarash Thana and Chauhali Thana were completely inundated with flood water and other thana named Kamarkhand Thana and Royganj Thana were inundated partially. Inundation map was prepared by HEC-GeoRAS, was mainly done by the water occupied channel area as areal extent due to rainfall. Flood due to overtopping was considered here, flooding due to breaching of embankment or other reasons were not considered in her study.

Rahman (2015) studied to develop floodplain extend maps and inundation maps of the Jamuna River. The Jamuna River is most vulnerable to river flood in Bangladesh. His study also dealt with flood pattern change with time and impact of levee on flood inundation area. One dimensional hydraulic model HEC-RAS with HEC-GeoRAS interface in co-ordination with ArcView were applied for the analysis. Collected bathymetric river grid with the topographic DEM were merged to produce the complete DEM of the river. Using the complete DEM, stream centerline, banks, flow paths and cross sections data prepared in HEC-GeoRAS were imported. After boundary condition setup, the model was calibrated and validated using known hydrological data collected from BWDB. The coefficient of determination (R2) has been found as 0.985, 0.977, 0.821

Previous Studies and 0.811 for steady calibration, steady validation, unsteady calibration and unsteady validation respectively. It was also found the NSE greater than 0.60 for both calibration and validation. After calibration and validation flood inundation and flood hazard map were generated using post-processing of HEC-GeoRAS. It was found in his study that the percentages of area inundated by 2, 5, 10, 25, 50 and 100-year return periods floods were 38.28, 46.10, 51.14, 54.63, 56.89 and 59.19% respectively. The result in his study showed that of the flooding area had water depth between 1.2 m to 3.6m. The assessment of the flood inundated area of his study showed that 41.99% and 30.83% area are of high hazard and very high hazard respectively for the 100-year return period flood. It was also found that, for the 100 year return period, if levee elevation is raised up to 2.13 m from existing levee elevation, then flood inundation land area decreased from 59.19% to 40%, no land will be inundated, if the levee elevation raised up to 2.56 m.

Hossain (2015) developed 5 days forecasted flood inundation map and hydrograph at house level flood information at Rowmari Upazilla of Kurigram district. This study area is surrounded by the mighty Brahmaputra River and flashy Jinjiram River. In his study a weather prediction model (WRF) was coupled with a hydrologic model (HEC HMS) and a hydrodynamic model (HEC-RAS) for predicting floods at Rowmari upazilla of Kurigram district. WRF 3.2 weather model was configured and used to predict rainfall over the basin 120 hours into future. Output of the weather model was incorporated with calibrated and validated hydrologic model HEC-HMS 4.0 and simulated every day during monsoon to forecast discharge at Bahadurabad. Three mathematical relations were developed between Bahadurabad station to other boundary of hydrodynamic model for forecast boundary generation. Then hydrodynamic model was simulated every day using forecast boundary to generate flood inundation map and forecast hydrograph at Rowmari Upazilla of Kurigram. It was found that the estimated NSE value for the calibration and validation period is 0.85 and 0.82. The found hydrodynamic Model (HEC-RAS) performance during calibration and validation period in terms of R2 and NSE against observed water level data to nearly 1. The Manning's roughness coefficient (n) and the coefficient of expansion/contraction (k) were key parameters in his study to calibrate of HEC-RAS model. In his study, analysis of forecast performance indicates that the forecast for the first 3 days are good and next 2 days are average to poor according to BWDB guideline. It was concluded in his study that the developed flood forecasting

Chapter 3 system is capable of predicting the inundated area of Rowmari Upazilla during a monsoon season.

Kalra and Ahmad (2012) studied two dimensional flow routing capabilities of hydrologic engineering center's river analysis system (HEC-RAS) for flood inundation mapping in lower region of Brazo River watershed subjected to frequent flooding. River reach length of 20 km located at Richmond, Texas was considered. Detailed underlying terrain information available from digital elevation model of 1/9-arc second resolution was used to generate the two-dimensional (2D) flow area and flow geometrics. Stream flow data available from gauging station USGS08114000 were used for the full unsteady flow hydraulic modeling along the reach. Developed hydraulic model was calibrated based on the manning's roughness coefficient for the river reach by comparison with the downstream rating curve. Water surface elevation and velocity distribution obtained after 2D hydraulic simulation were used to determine the extent of flooding. For this, RAS mapper's capabilities of inundation mapping in HEC-RAS itself were used. Mapping of the flooded areas based on inflow hydrograph on each time step were done in RAS mapper, which provided the spatial distribution of flow. The results from their study can be used for flood management as well as for making land use and infrastructure development decisions.

Ahmad et al. (2010b) carried out a study by integrating hydrological models with GIS to estimate the ﬂood zone of Nullah Lai in Rawalpindi, Islamabad, Pakistan. HEC-RAS and HEC-GeoRAS hydrological models were used to delineate the areas vulnerable to ﬂood at different discharge values. A topographic survey of fine resolution of the target area (Kattarian to Gawalmandi Bridges) was used to generate the DEM of the area. Krigging was used to interpolate the elevation data. GIS technology was also used to delineate the variation of topography and to find the inundation depths at various locations in the study area. Inundation area estimated at the discharge value of 3000 m3/sec is 3.4 km2, out of which 2.96 km2 is occupied under the inundation depth from 1 to 5 meters. It was found that maximum inundation depth can go up to 20 meters for this discharge value. Output of their study using HEC-RAS showed that inundated areas and inundation depths are in close approximation with survey based inundation results obtained by JICA. So their study showed that the integrated modeling approach used in this study worked well in

Previous Studies order to delineate areas vulnerable to ﬂood with a good estimation of inundation depths at a specific discharge value.

Moore (2011) on his study created a library of steady inundation maps for communities in Iowa which have a high risk of flooding. A high-resolution coupled one-dimensional/two- dimensional hydrodynamic model of Charles City, Iowa was developed in his study. Channel geometry from bathymetric surveys and surface topography from LiDAR were combined to create the digital elevation model (DEM) used in numerical simulations. Coupled one and two dimensional models were used to simulate flood events; the river channel and structures were modeled one-dimensionally, and the floodplain was modeled two-dimensionally. Spatially distributed roughness parameters were estimated using the 2001 National Land Cover Dataset. Simulations were performed at a number of mesh resolutions, and the results were used to investigate the effectiveness of re-sampling simulation results using higher- resolution DEMs. The effect of removing buildings from the computational mesh was also investigated.

Patel et al. (2017) carried a study on assessment of flood inundation mapping of Surat city by coupled 1D/2D hydrodynamic modeling. Surat city of India, situated 100 km downstream of Ukai Dam and 19.4 km upstream from the mouth of River Tapi, has experienced the largest flood in 2006. The peak discharge of about 25,770 m3s-1 released from the Ukai Dam was responsible for a disaster. Two hundred ninety-nine cross sections, two hydraulic structures and five major bridges across the river were considered for 1D modeling, whereas a topographic map at 0.5 m contour interval was used to produce a 5 m grid, and SRTM (30 and 90 m) grid was considered for Surat and the Lower Tapi Basin. Tidal level at the river mouth and the release from the Ukai Dam during 2006 flood were considered as the downstream and upstream boundaries, respectively. The simulation was done under the unsteady flow condition and validated for the year 2006. The simulated result of their showed that 9th August was the worst day in terms of flooding for Surat city and a maximum 75–77% area was under inundation. Out of seven zones, the west zone had the deepest flood and inundated under 4–5 m. Furthermore, inundation was generated under the bank protection work (i.e., levees, retaining wall) constructed after the 2006 flood. The simulated results showed that the major zones were safe against the inundation under 14,430 m3s-1 water releases from Ukai

Chapter 3 Dam except for the west zone. In their study, it was showed the 2D capability of new HEC-RAS 5 for flood inundation mapping and management studies.

3.4 FLOOD STUDY USING SATELLITE IMAGES

Sultana (2015) studied on flash flood forecasting using estimated precipitation by global satellite mapping in the north-east region of Bangladesh. The objectives of her study were to simulate Rainfall Runoff Model of the Northeast Bangladesh, the Meghalaya River catchments, the Barak River catchments and the Tripura River catchment with GSMap precipitation data and generate Flash Flood Forecast using Hydrodynamic Model incorporating WRF predicted precipitation. The result on her study showed underestimation of runoff. As a result bias correction in GSMap rainfall was needed in her study prior to application into operational flood prediction. She derived 7-day moving average bias-adjustment with six years of data from 2009 to 2014 comparing the gauge observed rainfall. The bias-adjustments were applied to every catchment. Then it was found that, these bias-adjusted rainfalls when applied to the NAM model resulted in improvement in runoff for all catchments. The calibrated hydrodynamic model showed good -result in flood forecasting in her study. Overall, findings from her study indicated that the GSMap underestimates rainfall significantly over Barak , trans boundary and north-east catchments. The accuracy of GSMap can be improved by applying a bias- adjustment. Prediction of water level using bias-adjusted rainfall estimates can improve the accuracy of water level prediction with considerable increase in the predictive capability of flood prediction for which the hydrological model needs to be calibrated.

Hossain (2015) studied a comparison that has been conducted between flood inundation maps generated from MODIS images and inundation maps generated from model output by FFWC from the year 2004 to 2014. From this comparison it was found that, among the five BWDB zones, inundation maps generated by FFWC in Northwest zone (NW) (R2=0.915), North central zone (NC) (R2=0.896), Northeast zone (NE) (R2=0.929) and Southeast zone (SE) (R2=0.959) have a strong correlation with the inundation maps generated from MODIS images of that zones. It was also found that MODIS inundation maps have a very poor correlation in Southwest zone (SW) (R2=0.058) which is because of fewer water level measuring stations in that zone and the model which has been used to prepare flood maps, does not consider tidal effect of this zone. It was also found that, correlation between MODIS inundation maps and FFWC inundation maps of a zone 32

Previous Studies depends on the number of water level observing station presents in the that zone. However it was found that the overall correlation between these two types of inundation maps as R2=0.701. Studying flood pattern of Bangladesh was also his objective of this study. It was seen that, flood pattern found from MODIS inundation maps has a good consistency with observed data found in annual flood reports of FFWC. Besides, the method developed in this study is very effective in the area where ground observation data are not available. It was concluded in his study that, this method of flood inundation mapping from MODIS image has a huge potential of observing and monitoring flood pattern and flood extent in Bangladesh.

Chapter Four SALIENT FEATURES OF THE MODELS

4.1 GENERAL

Several number of commercial and non-commercial software tools available for numerical modeling and analysis in GIS. The major tools used in this study are one and two dimensional numerical model HEC-RAS 5.0.3 beta version and Arc GIS for spatial data processing and HEC-GeoRAS for interfacing between HEC-RAS and Arc GIS. HEC-RAS and HEC-GeoRAS, an open source model which have excellent Graphical User Interfaces (GUI), were developed by US Army Corps of Engineers. Arc GIS was developed Environmental Systems Research Institute (ESRI) which enables to view, edit, create, and analyze geospatial data. Descriptions of these software tools are presented below.

4.2 HEC-RAS

HEC-RAS is a computer program that models the hydraulics of water flow through natural rivers and other channels. Prior to the recent update to Version 5.0 the program was one-dimensional, meaning that there is no direct modeling of the hydraulic effect of cross section shape changes, bends, and other two- and three-dimensional aspects of flow. The release of version 5.0 introduced two-dimensional modeling of flow as well as sediment transfer modeling capabilities. The program was developed by the US Department of Defense, Army Corps of Engineers in order to manage the rivers, harbors, and other public works under their jurisdiction; it has found wide acceptance by many others since its public release in 1995.

HEC-RAS is designed to perform one and two-dimensional hydraulic calculations for a full network of natural and constructed channels. The following is a description of the major capabilities of HEC-RAS.

4.2.1 User Interface

The user interacts with HEC-RAS through a graphical user interface (GUI). The main focus in the design of the interface was to make it easy to use the software, while still Salient Features of the Model maintaining a high level of efficiency for the user. The interface provides for the following functions: i. File Management ii. Data Entry and Editing iii. Hydraulic Analyses iv. Tabulation and Graphical Displays of Input and Output Data v. Inundation mapping and animations of water propagation vi. Reporting Facilities vii. Context Sensitive Help

4.2.2 Hydraulic Analysis Components

The HEC-RAS system contains several river analysis components for: (i) steady flow water surface profile computations; (ii) one- and two-dimensional unsteady flow simulation; (iii) movable boundary sediment transport computations; and (iv) water quality analysis. A key element is, that all four components use a common geometric data representation and common geometric and hydraulic computation routines. In addition to these river analysis components, the system contains several hydraulic design features that can be invoked once the basic water surface profiles are computed.

Steady Flow Water Surface Profile

This component of the modeling system is intended for calculating water surface profiles for steady gradually varied flow. The system can handle a full network of channels, a dendritic system, or a single river reach. The steady flow component is capable of modeling subcritical, supercritical, and mixed flow regimes water surface profiles. The basic computational procedure is based on the solution of the one-dimensional energy equation. Energy losses are evaluated by friction (Manning's equation) and contraction/expansion (coefficient multiplied by the change in velocity head). The momentum equation may be used in situations where the water surface profile is rapidly varied. These situations include mixed flow regime calculations (i.e., hydraulic jumps), hydraulics of bridges, and evaluating profiles at river confluences (stream junctions).

Chapter 4 One and Two Dimensional Unsteady Flow Simulation

This component of the HEC-RAS modeling system is capable of simulating one- dimensional; two-dimensional; and combined one/two-dimensional unsteady flow through a full network of open channels, floodplains, and alluvial fans. The unsteady flow component can be used to performed subcritical, supercritical, and mixed flow regime (subcritical, supercritical, hydraulic jumps, and draw downs) calculations in the unsteady flow computations module. An example of unsteady flow simulation has been shown in Appendix A-1.

The hydraulic calculations for cross-sections, bridges, culverts, and other hydraulic structures that were developed for the steady flow component were incorporated into the unsteady flow module.

Special features of the unsteady flow component include: extensive hydraulic structure capabilities Dam break analysis; levee breaching and overtopping; Pumping stations; navigation dam operations; pressurized pipe systems; automated calibration features; User defined rules; and combined one and two-dimensional unsteady flow modeling.

Sediment Transport/ Movable Boundary Computations

This component of the modeling system is intended for the simulation of one-dimensional sediment transport/movable boundary calculations resulting from scour and deposition over moderate time periods (typically years, although applications to single flood events are possible).

The sediment transport potential is computed by grain size fraction, thereby allowing the simulation of hydraulic sorting and armoring. Major features include the ability to model a full network of streams, channel dredging, various levee and encroachment alternatives, and the use of several different equations for the computation of sediment transport.

The model is designed to simulate long-term trends of scour and deposition in a stream channel that might result from modifying the frequency and duration of the water discharge and stage, or modifying the channel geometry. This system can be used to evaluate deposition in reservoirs, design channel contractions required to maintain navigation depths, predict the influence of dredging on the rate of deposition, estimate

Salient Features of the Model maximum possible scour during large flood events, and evaluate sedimentation in fixed channels.

Water Quality Analysis

This component of the modeling system is intended to allow the user to perform riverine water quality analyses. An advection-dispersion module is included with this version of HEC–RAS, adding the capability to model water temperature. This new module uses the QUICKEST-ULTIMATE explicit numerical scheme to solve the one-dimensional advection-dispersion equation using a control volume approach with a fully implemented heat energy budget. Transport and Fate of a limited set of water quality constituents is now also available in HEC-RAS. The currently available water quality constituents are: Dissolved Nitrogen (NO3-N, NO2-N, NH4-N, and Org-N); Dissolved Phosphorus (PO4- P and Org-P); Algae; Dissolved Oxygen (DO); and Carbonaceous Biological Oxygen Demand (CBOD).

4.2.3 Data Storage and Management:

Data storage is accomplished through the use of "flat" files (ASCII and binary), the HEC- DSS (Data Storage System), and HDF5 (Hierarchical Data Format, Version 5). User input data are stored in flat files under separate categories of project, plan, geometry, steady flow, unsteady flow, quasi-steady flow, sediment data, and water quality information. Output data is predominantly stored in separate binary files (HEC and HDF5). Data can be transferred between HEC-RAS and other programs by utilizing the HEC-DSS. A view of data storage capability of HEC-RAS is shown in Appendix A-2.

Data management is accomplished through the user interface. The modeler is requested to enter a single filename for the project being developed. Once the project filename is entered, all other files are automatically created and named by the interface as needed. The interface provides for renaming, moving, and deletion of files on a project-by-project basis.

4.2.4 Graphics and Reporting

Graphics include X-Y plots of the river system schematic, cross-sections, profiles, rating curves, hydrographs, and inundation mapping (Appendix A-3). A three-dimensional plot of multiple cross-sections is also provided. Inundation mapping is accomplished in the

Chapter 4 HEC-RAS Mapper portion of the software. Inundation maps can also be animated, and contain multiple background layers (terrain, aerial photography etc). Tabular output is available. Users can select from pre-defined tables or develop their own customized tables. All graphical and tabular output can be displayed on the screen, sent directly to a printer (or plotter), or passed through the Windows Clipboard to other software, such as a word-processor or spreadsheet. Reporting facilities allow for printed output of input data as well as output data. Reports can be customized as to the amount and type of information desired.

4.2.5 RAS Mapper

HEC-RAS has the capability to perform inundation mapping of water surface profile results directly from HEC-RAS (Appendix A-4). Using the HEC-RAS geometry and computed water surface profiles, inundation depth and floodplain boundary datasets are created through the RAS Mapper. Additional geospatial data can be generated for analysis of velocity, shear stress, stream power, ice thickness, and floodway encroachment data. In order to use the RAS Mapper for analysis, you must have a terrain model in the binary raster floating-point format (.flt). The resultant depth grid is stored in the .flt format while the boundary dataset is store in ESRI's Shape file format for use with geospatial software.

4.3 THEORETICAL BASIS FOR ONE DIMENSIONAL AND TWO DIMENSIONAL HYDRODYNAMIC CALCULATION

4.3.1 1D Steady Flow Water Surface Elevation

HEC -RAS is currently capable of performing 1D water surface profile calculations for steady gradually varied flow in natural or constructed channels. Subcritical, supercritical, and mixed flow regime water surface profiles can be calculated. Topics discussed in this section include: equations for basic profile calculations, applications of the momentum equation.

Salient Features of the Model Equations for Basic Profile Calculations

Water surface profiles are computed from one cross section to the next by solving the Energy equation with an iterative procedure called the standard step method. The Energy equation is written as follows (HEC-RAS 2016):

푍 + 푌 + = 푍 + 푌 + + ℎ (4-1)

Where,

Z1, Z2 = elevation of the main channel inverts

Y1, Y2 = depth of water at cross sections

V1, V2 = average velocities (total discharge/ total flow area)

α1, α2 = velocity weighting coefficients g = gravitational acceleration

he = energy head loss

A diagram showing the terms of the energy equation is shown in Figure 4-1.

Figure 4-1: Representation of terms in the energy equation

The energy head loss (he) is expressed as

푎 푉2 푎 푉2 ℎ = 퐿푆̅ + 퐶 2 2 − 1 1 (4-2) 2푔 2푔

Chapter 4 Where,

L = discharge weighted reach length

푆̅ = representative friction slope between two sections

C = expansion or contraction loss coefficient

The distance weighted reach length, L, is calculated as:

퐿푙표푏푄푙표푏+퐿푐ℎ푄푐ℎ+퐿푟표푏푄푟표푏 퐿 = (4-3) 푄푙표푏+푄푐ℎ+푄푟표푏

Where,

Llob, Lch, Lrob= x-section reach length specified for flow in the left overbank, main channel and right overbank respectively

Qlob + Qch + Qrob = arithmetic average of the flows between sections for the left overbank, main channel and right overbank respectively.

Application of Momentum Equation

Whenever the water surface passes through critical depth, the energy equation is not considered to be applicable. The energy equation is only applicable to gradually varied flow situations, and the transition from subcritical to supercritical or supercritical to subcritical is a rapidly varying flow situation. There are several instances when the transition from subcritical to supercritical and supercritical to subcritical flow can occur. These include significant changes in channel slope, bridge constrictions, drop structures and weirs, and stream junctions. In some of these instances empirical equations can be used (such as at drop structures and weirs), while at others it is necessary to apply the momentum equation in order to obtain an answer. The momentum equation is derived from Newton's second law of motion:

∑ 퐹 = 푚푎 (4-4)

Force = Mass x Acceleration (change in momentum)

Salient Features of the Model Applying Newton's second law of motion to a body of water enclosed by two cross sections at locations 1 and 2 (Figure 4-2), the following expression for the change in momentum over a unit time can be written:

P − P + W − F = Qρ∆ρ∆V (4-5)

Where,

P = Hydrologic pressure force at locations 1 and 2.

Wx= Force due to the weight of water in the X direction.

Fx= Force due to external friction losses from 2 and 1.

Q= Discharge

ρ= Density of water

∆Vx= Change on velocity from 2 to 1, in the X direction.

Figure 4-2: Application of momentum principle

The force in the X direction due to hydrostatic pressure is:

푃 = 훾 퐴푌푐표푠휃 (4-6)

The assumption of a hydrostatic pressure distribution is only valid for slopes less than 1:10. The cosθ for a slope of 1:10 (approximately 6 degrees) is equal to 0.995. Because the slope of ordinary channels is far less than 1:10, the cosθ correction for depth can be

Chapter 4 set equal to 1.0 (Chow, 1959). Therefore, the equations for the hydrostatic pressure force at sections 1 and 2 are as follows:

푃 = 훾퐴푌 (4-7)

푃 = 훾퐴푌 (4-8)

Where,

γ= unit weight of water

Ai= Wetted area of the cross section at locations 1 and 2

Yi= Depth measured from water surface to the centroid of the cross sectional area at locations 1 and 2.

The Weight of Water Force is:

Weight of water = (unit weight of water) x (volume of water)

푊 = 훾 퐿 (4-9)

푊 = 푊 × sin 휃 (4-10)

sin 휃 = = 푆 (4-11)

푊 = 훾 퐿푆 (4-12)

Where,

L= Distance between sections 1 and 2 along the X axis

푆= Slope of the channel, based on mean bed elevations zi= Mean bed elevation at locations 1 and 2

Salient Features of the Model Force to External Friction is:

퐹 = 휏푃퐿 (4-13)

Where,

휏 = Shear Stress

푃= Average wetted perimeter between section 1 and 2

휏 = 훾푅푆̅ (4-14)

Where,

푅 = Average Hydraulic Radius (R=A/P)

푆̅ = Slope of the energy grade line (friction slope)

̅ 퐹 = 훾 푆̅ 퐿 (4-15)

퐴 +퐴 퐹 = 훾 1 2 푆̅ 퐿 (4-16) 2

Mass time acceleration is

푚푎 = 푄휌∆푉 (4-17)

휌 = and ∆푉 = (훽 푉 − 훽 푉 )

푚푎 = (훽 푉 − 훽 푉 ) (4-18)

Where,

β= momentum coefficient that accounts for a varying velocity distribution in irregular channels

Substituting back into equation 4-5 and assuming Q can vary from 2 to 1

훾퐴 푌 − 훾퐴 푌 + 훾 퐿푆 − 훾 퐿푆̅ = 훽 푉 − 훽 푉 (4-19)

Chapter 4

+ 퐴 푌 + 퐿푆 − 퐿푆̅ = + 퐴 푌 (4-20)

+ 퐴푌 + 퐿푆 − 퐿푆̅ = + 퐴푌 (4-21)

This is the functional form of the momentum equation that is used in HEC-RAS. All applications of the momentum equation within HEC-RAS are derived from this equation.

4.3.2 1D/2D Coupled Hydraulic Modeling

This study is focused on the development of 1D/2D coupled hydrodynamic modeling for the Dharla River flooplain through HEC-RAS 5.0.3 published by USACE. The equations for 1D/2D coupled modeling have been stated in (Patel et al., 2017). The HEC-RAS 5.0.3 is fully solved in using the 2D Saint-Venant equation (Brunner 2016b; Manual 2016; Quirogaa et al., 2016):

+ + = 0 (4-22)

+ + = − − 푔ℎ + 푝푓 + (ℎ휏 ) + ℎ휏 (4-23)

+ + = − − 푔ℎ + 푞푓 + ℎ휏 + ℎ휏 (4-24) where h is the water depth (m), p and q are the specific flow in the x and y direction (m2 s-1), 휁 is the surface elevation (m), g is the acceleration due to gravity (m s-2), n is the -3 Manning resistance, 휌 is the water density (kg m ), 휏, 휏 and 휏 are the components of the effective shear stress and 푓 is the Coriolis (s-1) (Quirogaa et al., 2016).

4.4 GEOGRAPHIC INFORMATION SYSTEM

4.4.1 General

GISs are defined as computer systems capable of assembling, storing, manipulating, and displaying geographically referenced information (USGS, 1998). Originally developed as a tool for cartographers, GIS has recently gained widespread use in engineering design and analysis, especially in the fields of water quality, hydrology, and hydraulics. GIS provides a setting in which to overlay data layers and perform spatial queries, and thus create new spatial data.

Salient Features of the Model The results can be digitally mapped and tabulated, facilitating efficient analysis and decision-making. Structurally, GIS consists of a computer environment that joins graphical elements (points, lines, polygons) with associated tabular attribute descriptions. This characteristic sets GIS apart from both computer-aided design software (geographic representation) and databases (tabular descriptive data). For example, in a GIS view of a river network, the graphical elements represent the location and shape of the rivers, whereas the attributes might describe the stream name, length, and flow rate. This one-to- one relationship between each feature and its associated attributes makes the GIS environment unique. In order to provide a conceptual framework, it is necessary to first define some basic GIS constructs.

4.4.2 Data Models

Geographic elements in a GIS are typically described by one of three data models: vector, raster, or triangular irregular network. Each of these is described below.

Vector

Vector objects include three types of elements: points, lines, and polygons (Appendix A- 5). A point is defined by a single set of Cartesian coordinates [easting (x), northing (y)]. A line is defined by a string of points in which the beginning and end points are called nodes, and intermediate points are called vertices (Smith, 1995). A straight line consists of two nodes and no vertices whereas a curved line consists of two nodes and a varying number of vertices. Three or more lines that connect to form an enclosed area define a polygon. Vector feature representation is typically used for linear feature modeling (roads, lakes, etc.), cartographic base maps, and time-varying process modeling.

Raster

The raster data structure consists of a rectangular mesh of points joined with lines, creating a grid of uniformly sized square cells (Appendix A-6). Each cell is assigned a numerical value that defines the condition of any desired spatially varied quantity (Smith, 1995). Grids are the basis of analysis in raster GIS, and are typically used for steady-state spatial modeling and two-dimensional surface representation. A land surface representation in the raster domain is called a digital elevation model (DEM).

Chapter 4 Triangular Irregular Network (TIN)

A TIN is a triangulated mesh constructed on the (x, y) locations of a set of data points. To form the TIN, a perimeter around the data points is first established, called the convex hull. To connect the interior points, triangles are created with all internal angles as nearly equiangular as possible. This procedure is called Delaunay triangulation. By including the dimension of height (z) for each triangle vertex, the triangles can be raised and tilted to form a plane. The collection of all such triangular planes forms a representation of the land surface terrain in a considerable degree of detail (Appendix A-7). The TIN triangles are small where the land surface is complex and detailed, such as river channels, and larger in flat or gently sloping areas. Additional elevation data, such as spot elevations at summits and depressions and break lines, can also be included in the TIN model. Break lines represent significant terrain features like a streams or roads.

In three-dimensional surface representation and modeling, the TIN is generally the preferred GIS data model. Some reasons for the TIN model preference include the following: i. Requires a much smaller number of points than does a grid in order to represent the surface terrain with equal accuracy ii. Can be readily adapted to variable complexity of terrain iii. Supports point, line, and polygon features iv. Original input data is maintained in the model and honored in analysis

4.5 HEC-GEORAS

4.5.1 General

HEC-GeoRAS is an ArcGIS extension specifically designed to process geo-spatial data for use with the Hydrologic Engineering Center River's Analysis System (HEC-RAS). The extension allows users to create an HEC-RAS import sample containing geometric attribute data form an existing digital terrain model (DTM) and complementary data sets. Water surface profile results may also be processed to visualize inundation depths and boundaries. HEC-GeoRAS extension for ArcGIS used an interface method to provide a direct link to transfer information between the ArcGIS and the HEC-RAS.

Salient Features of the Model

4.5.2 Overview of Requirements

HEC-GeoRAS 10.2 is an extension for use with ArcGIS 10.3 that provides the user with a set of procedures, tools, and utilities for the preparation of GIS data for import in to RAS and generation of GIS data from RAS output. While the GeoRAS tools are designed for users with limited geographic information systems (GIS) experience, extensive knowledge of ArcGIS is advantageous. Users, however, must have experience modeling with HEC-RAS and have a thorough understanding of river hydraulics to properly create and interpret GIS data sets.

4.5.3 Software Requirements

HEC-GeoRAS 10.2 is an extension use for ArcGIS 10.3. Both the 3D Analyst extension and the Spatial Analyst extension are required. The full functionality of HEC-GeoRAS 10.3 requires HEC-RAS 5.0 beta, or later, to import and export all of the GIS data options. Older versions of HEC-RAS may be used, however, with limitations on importing roughness coefficients, ineffective flow data, blocked obstruct ions, levee data, hydraulic structures, and storage area data. Further, data exported from older versions of HEC-RAS should be converted to the latest XML file structure using the SDF to XML conversion tools provided.

4.5.4 Data Requirements

HEC-GeoRAS requires a DTM in the form of a TIN or a GRID. The DTM must be a continuous surface that includes the bottom of the river channel and the floodplain to be modeled. Because all cross-sectional data will be extracted from the DTM, only high resolution DTMs that accurately represent the ground surface should be considered for hydraulic modeling.

4.5.5 Getting Started

Start ArcMap. Load the HEC-GeoRAS tools by selecting Tools. Customize from the main ArcMap inter face and placing a check box next to HEC-GeoRAS. The Spatial Analyst and 3D Analyst extension will automatically load whenever required by the tools. When the HEC-GeoRAS extension loads, menus and tools are automatically added to the ArcMap interface. Menus are denoted by text and tools appear as buttons. These menus and tools are intended exported HEC-RAS simulation results. The HEC-GeoRAS tool bar is to aid the user in stepping through the geometric data development process and post-

Chapter 4 processing of exported HEC-RAS simulation results. The HEC-GeoRAS tool bar is shown in Appendix A-8.

4.5.6 HEC-GeoRas Menus

The HEC-GeoRAS menu options are RAS Geometry, RAS Mapping, ApUtilities, and Help. These menus are discussed below.

RAS Geometry

The RAS Geometry menu is for pre-processing geometric data for import into HEC-RAS. Items are listed in the RAS Geometry dropdown menu in the recommended (and sometimes required) order of completion. Items available from the RAS Geometry menu items are shown in Appendix A-9.

RAS Mapping

The RAS Mapping menu is for post-processing exported HEC-RAS results. Items available from the RAS Mapping dropdown menu are listed in the required order of complete RAS Mapping menu are shown in Appendix A-10.

ApUtilites

Features available from the ApUtilities menu are used behind the scenes to manage the data layers created through GeoRAS. Also available from the ApUtilities menu is functionality to assign a unique HydroID to features. Only experienced users should use the items on the ApUtilities menu.

Help The Help menu will provide general online help information and version is consistent with the ArcGIS product it is being used with provide the version number.

HEC-GeoRAS Tools

There are several tools are provided in the toolbar. A tool waits for user action after being activated and will either invoke a dialog or change the mouse pointer, indicating the need for further action. Appendix-11 shows the Summary of HEC-GeoRAS tools.

Chapter Five METHODOLOGY AND MODEL SETUP

5.1 GENERAL

Flood hazard mapping forms the foundation of risk management decision making process by providing information essential to understanding the nature and characteristics of the community’s vulnerability to flooding. The estimation of the flood depth, the extents of flooded area are essential in a flood-prone area for flood risk management. This chapter describes a brief discussion about the methodology and model setup to achieve the study objectives.

5.2 STUDY AREA

The Dharla River is one of Bangladesh's trans-boundary rivers. It originates in the Himalayas where it is known as the Jaldhaka River, and then it flows through the Jalpaiguri and Cooch Behar districts of West Bengal, India, one of the seven main rivers to do so. Here the river enters Bangladesh through the Lalmonirhat district and joins with the Jaldhaka River and flows as the Dharla River until it empties into the Brahmaputra River near the Kurigram district. Near Patgram upazila, it again flows eastwards back to India. It then moves south and enters Bangladesh again through Phulbari upazila of Kurigram district and continues a slow meandering course. In this study the river reach considered from when it re-enters in Bangladesh through Phulbari upazilla of Kurigram district and until it reaches to Brahmaputra River near Kurigram.

The Dharla watershed lies in North-West Zone of Bangladesh. The length of the river within Bangladesh is about 56 km (Patel et al., 2017). The flood plain about in average of 18 km from the left bank and 20 km from the right bank of the river is the study area (Figure 5-1). The river reach length considered in this study is about 49 km. The average bed slope of Dharla River is 0.00018 (Rahman et al., 2011). The bed slope of the river may vary from 0.0001 to 0.0005 (Patel et al., 2017). Only Dharla River’s floodplains have been considered for this study which includes Lalmonirhat district and Kurigram district. These two districts are in Rangpur division. Lalmonirhat district lies between 25º46´ and 26º33´ north latitudes and between 89º01´ and 89º36´ east longitudes. The Chapter 5 total area of the district is 1247.37sq km. Kurigram district lies between 25023' and 26014' north latitudes and between 89027' and 89054' east longitudes. The total area of the district is 2245.04 sq km.

Courtesy: Google Earth Pro

Figure 5-1: Identification of study area

Methodology and Model Setup

5.3 OVERALL VIEW OF THE METHODOLOGY

Modeling of any physical phenomenon is successful when practicing a process of an iterative development. Model refinements are based on the availability and quality of data, hydrological understanding and scopes of the study. The general approach that has been followed in the current study can be summarized in the flow chart given in Figure 5- 2. To achieve the study objectives, a brief description of the methodology and approaches that are followed has been provided in this section.

Figure 5-2: Summary of steps of methodology in flow chart

Chapter 5

5.3.1 Preparation Phase

I. Data Collection

In order to develop the mathematical flood model, various kinds of data, recent and previous years have been collected and compiled. These data also form the basis for further analysis and interpretation of the model results leading to accurate assessment of hydrological condition of Dharla River floodplain. According to the modeling requirements, a significant amount of data includes water level, discharge, cross-section have been collected and setup flood model using this data. All collected data are summarized in the Table 5-1. A brief description of all the data used in the study is presented below.

Table 5-1: Summary of data type Data Type Data Source Data Location Periods(year) Discharge data BWDB Taluk-Simulbari (SW 76) 1996-2017 Discharge data BWDB Kurigram (SW 77) 1996-2017 Water level BWDB Taluk-Simulbari (SW 76) 1996-2017 Water level BWDB Kurigram (SW 77) 1996-2017 Cross section BWDB Dharla River (RMDLA 1-RMDLA 10) 2014 DEM USGS Bangladesh 2014 Satellite image USGS/Earth North-West (NW) of Bangladesh 2017 Explorer

a. Discharge Data

The flow hydrographs of the Dharla River at Taluk-Simulbari and Kurigram gauge station were collected from the Bangladesh Water Development Board (BWDB) for the year 1996-2017 .The hydrograph comprises of data at an interval of one day. The discharge data of Taluk-Simulbari has been used as the upstream boundary condition in the model.

b. Water Level Data

The stage hydrographs of the Dharla River at Taluk-Simulbari and Kurigram gauge station were collected from the Bangladesh Water Development Board (BWDB) for the year 1996-2017. The hydrograph comprises of data at an interval of one day. The water level data of Kurigram has been used as the downstream boundary condition in the

Methodology and Model Setup model. The data of year 2013 has been used to calibrate and the year of 2014 data has been used to validate the hydro-dynamic model.

Locations of two stations where the historical water discharge and water level data of the Dharla River have been collected from BWDB are shown in Figure 5-3.

Figure 5-3: Locations of discharge and water level station of Dharla River

c. Cross-section

River cross-sections of Dharla River are collected for the years of 2014 from Morphology Department of Bangladesh Water Development Board (BWDB). BWDB collects cross- section data at 10 different stations in Dharla River designated by RMDLA-1 to

Chapter 5 RMDLA-10. RMDLA means River Morphology Dharla. These cross-sections depict the shape and morphology of a river. Locations of the stations are shown in Figure 5-4.

Figure 5-4: Locations of cross-section of Dharla River

d. Digital Elevation Model (DEM)

A digital elevation model (DEM) is a digital model or 3D representation of a terrain's surface, created from terrain elevation data. DEM data identifies the elevations of the earth surface and to locate natural and relevant features on it (Rouf, 2015). A DEM can be represented as a raster (a grid of squares, also known as a height map when representing elevation) or as a vector-based triangular irregular network (TIN). The DEM could be acquired through techniques such as photogrammetry, lidar, land surveying, etc. DEMs

Methodology and Model Setup are commonly built using data collected using remote sensing techniques, but they may also be built from land surveying. This data is required to formulate mathematical models for the study area. Each cell, and cell face, of the computational mesh of 2D flow area is pre-processed in order to develop detailed hydraulic property tables based on the underlying terrain used in the modeling process (Brunner et al., 2015).

The Shuttle Radar Topography Mission (SRTM) data has emerged as a global elevation data in the past one decade because of its free availability, homogeneity and consistent accuracy compared to other global elevation dataset. The present study explores the hydrological modeling with the help of the SRTM digital elevation model (DEM). In this study DEM image The Digital Elevation Model (DEM) of Bangladesh in raster format has been collected from the FTP server of the Shuttle Radar Topographic Mission (SRTM) of National Aeronautics and Space Administration (NASA). Figure 5-5 shows the DEM of Bangladesh.

e. Satellite Image

For the purpose of comparison between model simulated flood map and observed flood map, Landsat 7 ETM Plus and Sentinel- 2 satellite images have been collected from United States Geological Survey (USGS)/Earth Explorer. The collected satellite images are of North-West (NW) region of Bangladesh in the year of 2017 for the visualization of inundated area in these years. These Images have been used for the comparison with the model simulated flood map.

II. Modification of Digital Elevation Model (DEM)

DEMs are increasingly used for visual and mathematical analysis of topography, landscapes and landforms, as well as modeling of surface processes. The accuracy of a DEM is determined by data type and actual sampling technique of the surface during DEM creation. A DEM offers the most common way of showing topographic information and even enables the modeling of flow across topography, a controlling factor in distributed models of landform processes (Khan et al., 2017).

The Digital Elevation Model (DEM) of Bangladesh was collected from the FTP server of the Shuttle Radar Topographic Mission (SRTM) of National Aeronautics and Space Administration (NASA). The DEM was in geographical coordinate system

Chapter 5 (GCS_WGS_1984). Geographic coordinate systems indicate location using longitude and latitude based on a sphere (or spheroid) while projected coordinate systems use X and Y based on a plane. Projections manage the distortion that is in evitable when a spherical earth is viewed as a flat map (Masood, 2011).

Figure 5-5: Digital Elevation (DEM) of Bangladesh

All the data in the DEM have been projected on to the Bangladesh Transverse Mercator (BTM). The data comprises of a resolution of 30m x 30m. The elevation of the DEM has been measured with respect to the mean sea level. In this study all the elevations including topography of river cross sections, water surface elevation have been considered are measured from Public Work Datum (PWD). PWD is a horizontal datum believed originally to have zero at a determined Mean Sea Level (MSL) at Calcutta. PWD is located approximately 1.5 ft below the MSL established in India under the British Rule and brought to Bangladesh during the Great Trigonometric Survey. To adjust this

Methodology and Model Setup difference in elevation, a slight modification of the collected DEM has been done. The DEM both before modification and after modification are shown in Figure 5-6.

(a) DEM before modification (b) DEM after modification

Figure 5-6: Digital Elevation (DEM) modification

III. Clipping of Study Area and Triangulated Irregular Network (TIN) Generation

After taking the modified DEM of Bangladesh, the Shape file of Lalmonirhat and Kurigam that are adjacent to Dharla River was superimposed which is shown in Figure 5- 7(a). The DEM of two districts has been clipped from the modified digital elevation model using the Clipping Tool in ArcToolbox. The clipped DEM is shown in Figure 5-7 (b). And after that, study area has been clipped from the DEM of Lalmonirhat district and Kurigram district using the shape file of study area. This is shown in Figure 5-7(c).

The purpose of the Raster to TIN tool is to create a Triangulated Irregular Network (TIN) whose surface does not deviate from the input raster by more than a specified Z tolerance. It is used to convert raster from a DEM to a TIN surface model. It is done by using the Raster to TIN tool in the ArcToolbox. TIN generation of the study area has been shown in Figure 5-7(d).

Chapter 5

(a) (b)

Figure 5-7: (a) Superimposed shape file on modified DEM, (b) Clipped DEM of shape file, (c) Dem of study area and (d) Raster to TIN generation of the study area

Methodology and Model Setup

5.3.2 Execution Phase

I. Preprocessing in HEC-GeoRAS

HEC-GeoRAS is an Arc GIS extension specifically designed to process geo spatial data to incorporate with the Hydrologic Engineering Center River’s Analysis System (HEC- RAS). The extension allows users to create an HEC-RAS import file containing geometric attribute data form an existing digital terrain model (DTM) and complementary data sets. Water surface profile results may also be processed to visualize inundation depths and boundaries. HEC-GeoRAS is organized as preprocessing (preRAS) and post- processing (postRAS) facilitated by menu and buttons. In this study only preprocessing is done in HEC-GeoRAS.

a. Pre-processing to Develop the RAS GIS Import File

The RAS GIS Import File consists of geometric data necessary to perform hydraulic computations in HEC-RAS. Cross-sectional elevation data are derived from an existing Digital Terrain Model (DTM) of the channel and surrounding land surface, while cross- sectional properties are defined from points of intersection between RAS layers. The DTM may be in the form of a TIN or GRID. In this study DTM has been used as a TIN format.

The goal of this section is to develop the spatial data required to generate a HEC-RAS import file with a 3-D river network and defined 3-D cross sections. This extraction comprises several steps. These are development of a stream centerline, cross-sections cut lines, main channel banks, and flow path lines as shape files.

River Centerline Creation:

The Dharla River has been represented by the stream centerline layer. The river has been created by starting from the upstream end and working downstream following the deepest part of the channel. The river name has been assigned as Dharla and reach name has been assigned as upstream.

Generally while creating a stream centerline layer some rules have been maintained. Following are the rules that have been maintained while creating stream centerline layer for Dharla River. Figure 5-8 (a) shows river centerline created for this study.

Chapter 5 i. The stream centerline must have been in the downstream direction i.e. reach has been started at the upstream end and finished at the downstream end. ii. River reach must have a unique combination of its river name and reach name which has assigned in this study.

Riverbanks Creation:

The riverbanks separate the main channel from the overbank areas when flooding occurs. It differentiates the resistance of the main channel and the overbanks. This is important for a steady-state simulation. Before beginning to create the banks theme, the following rules were important to remember to digitize:

i. There were exactly two bank lines per cross section. It was important to make sure to have a left and right bank defined, but no more than that. ii. Bank lines may be broken. The theme does not have to be a continuous poly line along a side of the channel. iii. Orientation of the banks lines was unimportant. Starting on the left or right of the stream centerline, as well as upstream or downstream was not matter. Creating this theme is optional when using GeoRAS. For this exercise, the banks theme was created so can be seen how accurate or inaccurate digitization ended up being. Figure 5-8 (a) shows river banks created for this study.

Flow Path Creation:

The Flow Path Centerlines theme was used to identify the hydraulic flow path in the left overbank, main channel, and right overbank. Creating the flow path centerline layer has assisted in properly laying out the cross-sectional cut lines. As the stream centerline already exists in this study, river centerline has been copied for the flow path in the main channel. Flow paths are created in the direction of flow (upstream to downstream). To complete assigning flow paths, the flow paths were digitized for the left and right overbanks. The following rules regarding flow path development were adhered:

i. All flow paths (left overbank, main channel, and right overbank) are drawn from upstream to downstream. A visual representation of the direction of the flow paths has been used to maintain this rule using a line with arrow. The arrows should point towards downstream.

Methodology and Model Setup ii. All three flow paths should be ensured at each cross section. The flow paths are used to derive downstream reach lengths in HEC-RAS. It is needed to be concerned with the buildings in the GRID. It should be noticed that the three flow paths for each reach, flowing from upstream to downstream. Once the digitization of the flow paths were completed, each flow path must now be identified as a left, right, or channel flow path. Obviously the channel is the flow path along the center of the river channel. Determining the left and right flow path is accomplished by an upstream to downstream perspective. Flow paths created for this study are shown in Figure 5-8 (b).

(a) (b)

Figure 5-8: (a) River centerline and bank line of Dharla River and (b) River flow paths of Dharla River

Cross-Sectional Cut Lines Creation:

The location, position and expanse of cross sections are represented by the Cross Section Cut Line theme. This theme will identify the planar location of the cross sections and the station elevation data being extracted from the DTM along each cut line for use in HEC-

Chapter 5 RAS. The rules which were followed for developing this theme in this study are as follows:

i. Cross sectional cut lines must be pointed from the left overbank to the right overbank. Thus, each cut line from left to right was drawn, as it could look downstream. ii. As it could look at the Preprocess View, that would be from right to left. Cross sectional cut lines must cross each of the three flow paths and the two banks exactly once. iii. Cross sectional cut lines should be perpendicular to the direction of flow. (In some cases, this might be difficult to accomplish and still follow the other rules). iv. Cross sectional cut lines should not intersect.

To begin digitizing the Cross Section Cut Line theme, it was specified as xscutlines.shp and began digitizing. Ten cross-section locations were chosen for the total study reach and drawn according to the morphological station position (Figure 5-9). There are four key points to remember:

i. For buildings, it should be acted as if they do not exist in the TIN. ii. Locations should be eluded where possible bridges and/or overpasses exist in the TIN. iii. It should be ensured to place cross sections at upstream and downstream boundaries and iv. Cut lines are to be started and ended well beyond the extent of the flow paths, since this theme will subsequently become the extent of the floodplains bounding polygon.

Generation of Additional Attributes and 3D Spatial Data:

The polyline themes that have been created was used to extract the 3-D attributes of the GRID through the theme’s spatial relationship with the terrain. Extract spatial data were used to determine Manning’s roughness value, ‘n’, for the HEC-RAS model. HEC- GeoRAS has been used to process the intermediate data, the stream centerline (3D) and XS surface line (3D) in the subsequent steps of this study.

Methodology and Model Setup Centerline Completion:

It was important to check all the data were extracted or not for stream centerline. For this reason, completion of the centerline theme was done ensuring all features from the RAS Geometry menu under “Stream Centerline Attributes”. The attribute table was then checked to ensure the data were properly extracted.

Cross-Section Attributing:

It was also important to ensure the all cross-section data were properly extracted. To complete the cross section layer, all features were ensured from the RAS Geometry toolbar under the “XS Cut Line Attributes’’. The 2D feature class of XS Cut Lines was intersected with the GRID to create a feature class with 3D cross section. After that, the attribute table and cross sections are examined in order to check their correctness.

b. Exporting GIS Data to HEC-RAS:

The generation of the HEC-RAS import file was the last step of the HEC-GeoRAS preprocessing. The idea was to create a HEC-RAS input file in RAS Import format which includes the terrain elevation extracted from the TIN, the 3-D stream centerline and the 3- D cross sections themes as z values (z value is the elevation above public work datum and, for our case, is in units of meter). The “Extract GIS DATA” was clicked under the menu “RAS Geometry” from the HEC-GeoRAS toolbar. The default name GIS2RAS was accepted and saved in them selected folder.

II. Processing on HEC-RAS (Model Development)

Mathematical modeling is an advance technology in engineering practice for predicting flood water level. Hydrological and one-dimensional hydrodynamic models have setup using HEC-package to forecast flood inundation map and hydrograph. Using this modeling software, a mathematical model for flood forecasting system has been developed. The various key steps during processing the model are described below.

a. 1D Hydrodynamic Model:

HEC-RAS 5.0 is an integrated software system, designed for interactive use in a multi- tasking environment and used to perform one-dimensional water surface calculations.

Chapter 5 Four files are required to run a HEC-RAS 5.0 project. First, the `Project File' that acts as a file management tool and identifies which files are used in the model. The `Plan File' sets the model conditions as subcritical, supercritical, or mixed flow and runs the simulation. The `Geometry File' contains all the geometric attributes for the model. The `Steady Flow File' establishes the steady-state flow and boundary conditions at numerous points in time for the model. The `Unsteady Flow' File establishes the unsteady-state flow and boundary conditions at numerous points in time for the model. The 1D hydrodynamic model of the study areas has been developed along with the mighty Dharla River. The total length of river in the model is around 56 km.

Importing 1D Geometry Data:

The geometric data editor was opened from the project window. The `import GIS data' from the `File' menu was opened. This operation needs to specify the unit system. The unit system was selected as SI units. The model comprises around 10 number of cross- sections of the river which were imported for the development of 1D hydrodynamic model (Figure 5-9). Initially for 1D calibration and validation, two boundary condition have been considered along the river upstream and downstream where boundary input data are collected from BWDB. The HEC-RAS 1D Model has been calibrated for hydrological event of 2013 and validated for hydrological event of 2014.

Figure 5-9: 1D geometric feature of Dharla River 64

Methodology and Model Setup Boundary Condition:

Boundary condition is the conditions or phenomenon occurring at the boundaries of the model. The data of 2013 and 2014 have been used as a boundary condition for calibration and validation for unsteady flow simulations.

Boundary Condition for Unsteady Analysis (1D):

The one-dimensional hydrodynamic model has one upstream boundary and one downstream boundary. Figure 5-9 shows the locations of the two boundary condition. To calibrate the model, the boundary conditions have been used from the observed data for the year of 2013. To validate the model, the boundary conditions have been used from the observed data for the year of 2014. The discharge hydrograph has been used as a upstream boundary condition and stage hydrograph has been used as a downstream boundary condition. The upstream discharge boundary and downstream water level boundary condition for calibration and validation are shown in Figure 5-10 and 5-11 respectively.

Figure 5-10: Locations of boundary condition

Chapter 5

1500

1000 /s) 3 Flow (m Flow 500

0 Jan/1 Mar/1 May/1 Jul/1 Sep/1 Nov/1 Jan/1 Date

(a) 27

Stage (m) Stage 24

22 Jan/1 Mar/1 May/1 Jul/1 Sep/1 Nov/1 Jan/1 Date (b)

Figure 5-11: (a) Upstream boundary condition for calibration 2013 and (b) Downstream boundary condition for calibration 2013

Methodology and Model Setup

1400

1200

1000 /s)

3 800

600 Flow (m Flow

400

200

0 Jan/1 Mar/1 May/1 Jul/1 Sep/1 Nov/1 Jan/1 Date

(a) 27

Stage (m) Stage 24

22 Jan/1 Mar/1 May/1 Jul/1 Sep/1 Nov/1 Jan/1

Date (b)

Figure 5-12: (a) Upstream boundary condition for validation 2014 and (b) Downstream boundary condition for validation 2014

Chapter 5 Model Calibration:

To simulate the model with base and different flow conditions, it is necessary to test the model’s performance. Sets of field data are prerequisite for the testing. This testing provides an impression about the degree of the accuracy of the model in reproducing river processes. This process is known as calibration. Including consistent and rational set of theoretically defensible parameters and inputs of the model provide the basis for finalizing these inputs and parameter with good comparison of the model-generated outputs with the observed data. For this study, one-dimensional HEC-RAS model has been calibrated hydro-dynamically for the year 2013. Since there is no intermediate hydrologic station in between Taluk-Simulbari station and Kurigram station, calibration has been done in this study using Taluk-Simulbari station (Figure 5-12).

Model Validation:

Model validation involves testing of a model with a data set representing ‘observed’ field data. This data set represents an independent source different from the data used to calibrate the model. Due to the uncertainty of prediction, this step is very important prior to widespread application of model output. The calibrated HEC-RAS based model has been used to validate the flow for the year 2014.Since there was no intermediate hydrologic station in between Taluk-Simulbari station and Kurigram station, the station which was considered in this study for validation is Taluk-Simulbari station (Figure 5- 12).

Model Performance Evaluation:

The quantitative statistics were divided into two major categories: standard regression and error index. Standard regression statistics determine the strength of the linear relationship between simulated and measured data. Error indices quantify the deviation in the units of the data of interest (Legates and McCabe, 1999). For this study, two methods have been used: coefficient of determination R2 as Standard regression statistics and Nash and Sutcliffe simulation efficiency (NSE) as dimensionless.

Methodology and Model Setup i. Coefficient of determination (R2)

Coefficient of determination (R2) describes the degree of collinearity between simulated and measured data (Rahman, 2015). R2 ranges from 0 to 1, with higher values indicating less error variance and typically values greater than 0.5 are considered acceptable (Legates and McCabe, 1999). R2 has been widely used for model evaluation, these statistics are oversensitive to high extreme values (outliers) and insensitive to additive and proportional differences between model predictions and measured data (Legates and McCabe, 1999). Correlation Coefficient,

∑((∑ )(∑ )) 푟 = (5-1) (∑ )(∑ )(∑ )(∑ )

Where, x is the observation and y is the simulated value for the constituent being evaluated. Coefficient of determination , R2 equals to r2.

Figure 5-13: Location of model calibration and validation

Chapter 5 ii. Nash-Sutcliffe Efficiency (NSE)

The Nash-Sutcliffe efficiency (NSE) is a normalized statistic that determines the relative magnitude of the residual variance (noise) compared to the measured data variance (information) (Nash and Sutcliffe, 1970). NSE indicates how well the plot of observed versus simulated data fits the 1:1 line. NSE is computed as shown in equation 5-2

2 푛 표푏푠 푠푖푚 ∑푖=1(푌푖 −푌푖 ) 푁푆퐸 = 1 − 2 (5-2) 푛 표푏푠 푚푒푎푛 ∑푖=1(푌푖 −푌푖 )

obs sim Where, Yi is the i-th observation for the constituent being evaluated, Yi is the i-th mean simulated value for the constituent being evaluated, Yi is the mean of observed data for the constituent being evaluated, and n is the total number of observations. NSE ranges between -∞ and 1.0 (1 inclusive), with NSE =1 being the optimal value (Rahman, 2015). Values between 0.0 and 1.0 are generally viewed as acceptable levels of performance, whereas values <0.0 indicates that the mean observed value is a better predictor than the simulated value, which indicates unacceptable performance (Gupta et al., 2009; Servat and Dezetter, 1991)also found NSE to be the best objective function for reflecting the overall fit of a hydrograph.

b. One Dimensional and Two Dimensional Coupled Hydrodynamic Model:

Layering Terrain Data:

HEC-RAS supports the use of digital terrain model for representing the bare earth ground surface. RAS DTM support is for raster data of many formats, but once processed, the data will be stored in the Geo Tiff format. A terrain layer is of primary importance for computing hydraulic properties (elevation-volume, elevation-wetted perimeter, elevation profiles, etc), inundation depths and floodplain boundaries.

This can be done in RAS Mapper once the terrain model has been associated with a geometry and plan. This terrain can be used to visualize the floodplain geometry. So the grid-cell size must be small enough that it captures the features of the terrain. Prior to this, digital terrain model of study area has been created using new layer tool for preprocessing geometric data for 2D flow areas, computing flood depths and inundation boundaries from simulation results. While creating a terrain layer projection has been specified by

Methodology and Model Setup selecting Bangladesh Transverse Mercator as an ESRI projection file (*.prj).Once a projection has been included, all the data will be projected into the selected coordinate system.

2D Flow Area Computational Mesh:

Two dimensional flow areas are regions of a model in which the flow through that region will be located at the beginning of a reach (as an upstream boundary to a reach), at the end of a reach (as a downstream boundary to a reach), or they can be located laterally to a reach. A polygon boundary for the 2D flow area has been drawn in the right side of Dharla River and left side of the Dharla River.The HEC-RAS 2D modeling capability uses a Finite-Volume solution scheme. This algorithm was developed to allow for the use of a structured or unstructured computational mesh. A 200m*200m grid resolution has been defined for the computational mesh (Figure 5-14).

Figure 5-14: 2D flow area computational mesh

Lateral Structures:

At any lateral structures HEC-RAS has the ability to model lateral weir, gated spillways, culverts, diversion rating curves and an outlet time series. A single lateral weir, a weir and

Chapter 5 separate set of gates, a weir and group of culverts, or any combination of weir, gates, gates, culverts, rating curves and a time series outlet can be set up. In general a cross sections should be end at the inside top of the levee, and then the lateral structure option can be used to represent the top of the levee along the stream. On the upstream of Dharla River two lateral structures both side of the river have been entered using lateral structure button from geometry data window. Figure 5-15 shows the introduction of lateral structure in model development.

Figure 5-15: Introduction of lateral structure

Defining Boundary Condition Location Line:

If a model needs a boundary condition then in HEC-RAS a boundary condition line can be drawn along the outer boundary of the area where you want the boundary condition to be located. Numbers of boundary conditions location has been defined and individual unique name has been given to each of the line. After all the boundary condition lines have been drawn along the flow areas, the geometry data has been saved. In unsteady flow data editor, boundary condition types data for each boundary condition lines have

Methodology and Model Setup been defined. In this study location of boundary condition line has been replaced until the model gives the correct inundation area.

Boundary Condition:

Boundary condition is the conditions or phenomenon occurring at the boundaries of the model. There are several different types of boundary conditions available for using. They are as follows i. Flow hydrograph ii. Stage hydrograph iii. Stage and flow hydrograph iv. Rating curve v. Normal depth vi. Lateral inflow hydrograph vii. Uniform lateral inflow hydrograph viii. Groundwater inflow ix. Time series of gate openings x. Elevation controlled gate xi. Navigation dam xii. Internal boundary stage and/or flow

The following is a short discussion of some of the boundary condition

Flow hydrograph:

A flow hydrograph can be used as either an upstream boundary or downstream boundary condition, but it is most commonly used as an upstream boundary condition.

Stage hydrograph:

A stage hydrograph can be used as either an upstream or downstream boundary condition. The user has the choice of either attaching a HEC-DSS file and pathname or ntering the data directly into a table

Stage and flow hydrograph:

Chapter 5 The stage and flow hydrograph option can be used together as either an upstream or downstream boundary condition. The upstream stage and flow hydrograph is a mixed boundary condition where the stage hydrograph is inserted as the upstream boundary until the stage hydrograph runs out of data; at this point the program automatically switches to using the flow hydrograph as the boundary condition. This type of boundary condition is primarily used for forecast models where the stage is observed data up to the time of forecast and the flow data s a forecasted hydrograph

Rating curve:

The rating curve option can be used as a downstream boundary condition. The user can either read the rating curve from HEC-DSS or enter it by hand into the editor. When using a rating curve, make sure that the rating curve is a sufficient distance downstream of the study area, such that any errors introduced by the rating curve do not affect the study reach.

Normal depth:

The Normal depth option can only be used as a downstream boundary condition for an open-ended reach. This option uses Manning’s equation to estimate a stage for each computed flow. To use this method the user is required to enter a friction slope (slope of the energy line) for the reach in the vicinity of the boundary condition. The slope of the water surface is often a good estimate of the friction slope, however this is hard to obtain ahead of time. The average bed slope in the vicinity of the boundary condition location is often used as an estimate for the friction slope.

As recommended with rating curve option, when applying this type of boundary condition it should placed far enough downstream, such that any errors it produces shall not affect the results at the study reach.

Boundary Condition Line for Unsteady Flow Analysis (1D and 2D Coupled Model):

Here in this study, different boundary conditions have been applied for different time series data for the calibration of the 2D flow area for unsteady flow simulations. There are six peripheral boundaries in the hydrodynamic model (HEC-RAS) of the study area out of which four are inflow boundaries, and the two are outflow boundaries. Flow hydrograph

Methodology and Model Setup was used to serve the purpose of inflow into the flood plain and to consider the effect of Teesta River and Dudhkumar River and Jamuna River. For the purpose of passing out the water from floodplain, normal depth condition has been used. The channel and the boundary condition lines incorporated in the hydrodynamic model have been shown in Figure 5-16.

Figure 5-16: Introduction of boundary condition line

Assuming Friction Slope as a Normal Depth Boundary Condition:

In case of using energy slope as a boundary condition, discrete energy slope at the downstream cross section is the correct assumption. But without computing, this is impossible to come by, and in order to compute, an energy slope has to be assumed. There are a few ways an energy slope for downstream boundary can be calculated and has been used in this study. i. Measurement of the average bed slope of your stream in the profile plot.

Chapter 5 ii. Measurement of the bed slope of the last two cross sections at the downstream boundary.

Water Surface Profile Generation:

The 1D and 2D coupled hydrodynamic model has been used to generate water surface profiles for unsteady flow conditions for the year 2017 and historical flood event 198. Water surface profiles of five individual day (10 May 2017, 10 June 2017, 10 July 2017, 10 August 2017 and 20 September 2017) have been generated for 2017. The generated water surface profile data have been exported in GIS format data to develop flood inundation map and flood depth to produce flood hazard map. Same procedure has been performed for the historical flood event of 1998 considering five consecutive days of 10 May, 10 June, 10 July, 10 August and 20 September.

5.3.3 Comparison and Hazard Mapping Phase I. Mapping of Result and Visualization

Geospatial HEC-RAS results are managed in two distinct methods. (a) Dynamic map (b) Stored map a. Dynamic map has been generated on-the-fly at the current view. This map is recomputed in RAS each time the map extents change based on the zoom level. It can also be animated within RAS Mapper. For each HEC-RAS Plan, specific output dynamics layers (water depth, velocity, and water surface elevation) have been automatically created for immediate visualization and analysis of the RAS simulation results. b. Stored map data has been created and written to disk for permanent storage, exporting, sharing and analysis. Stored maps, however, are computed using the base resolution of the underlying Terrain Layer and stored to disk for permanent data storage. The datasets have been created and edited using the Manage Results Maps. Different types of map can be created and they are available based on the type of run performed. In this study, depth map type and inundation boundary map type have been created for exporting and analysis purpose. For serving the purpose of comparison and hazard mapping these maps have been used. Other than these, various maps can be produced in HEC-RAS. They are as follows i. Water Surface Elevation

Methodology and Model Setup ii. Velocity iii. Flow (1D) iv. Shear Stress v. Stream Power vi. Depth*Velocity vii. Depth*Velocity^2 viii. Arrival Time ix. Duration x. Recession xi. Percent time Inundation

II. Satellite imagery Acquisition and Preparation

The main data sources for comparing the model simulated results with for this study were Landsat-7 Enhanced Thematic Mapper Plus (ETM+) and Sentinel-2 images taken during the flood of 2017.The images have been taken considering the availability and the most possible cloud free. The full scenes of Landsat-7 ETM+ and Sentinen-1 covering the North West part of Bangladesh and parts of neighboring countries were considered for this study. Vector data for the boundary of study area was prepared using an existing shape file. After geometric correction, sub-scenes corresponding to the area of study were extracted from the full scenes using a vector layer for the study area.

Cloudy skies are expected during big floods, and Landsat-7 Enhanced Thematic Mapper Plus (ETM+) and Sentinen-1 images can’t observe a cloud covered ground surface, so the presence of clouds over the damaged areas after an event limits the usefulness of this data and difficulty arises with the interpretation of whether a given area beneath the clouds is dry or covered by water. I employed low cloud covered and an algorithm was developed to interpret these cloud covered pixels. After differentiating all of the pixels into three categories (land, water and cloud), this algorithm was used to estimate what lay beneath the cloud covered pixels. The shadow of a cloud was included with the cloud covered area. Cloud covered pixels for the flood images which had represented water in the dry season image were interpreted as water, but the remaining cloud covered pixels were divided into two types through the use of digital elevation data. A flocculus of cloud covered pixels, which was surrounded by non-water pixels was considered to be non- water, while cloud covered flocculus surrounded by water pixels was compared with the

Chapter 5 neighboring water pixels. If the elevation of a cloud covered pixel of interest (candidate pixel) was less than the land surface elevation of a neighboring water pixel plus water depth, then the candidate pixel was considered to be a water pixel (it was assumed that the flow of water from one pixel to another pixel occurs across the common perimeter only), otherwise it was considered to be a non-water pixel. For cloud covered pixels initially determined to be non-water in the same flocculus, however, these procedures were repeated to compare them with neighboring water pixels in all directions until the decision was finalized for the flocculus. On the other hand, pixels that were initially determined to be water were not allowed to be changed to non-water pixel status. This procedure was used for all the flocculi in the image. Digital elevation and water depth data are necessary to apply this procedure.

Image for the flood of 28 July 2017 was analyzed. Image was taken under water content conditions. To differentiate between water and non-water, unsupervised land cover classification and supervised land cover classification were performed for the satellite image.

III. Development of flood hazard maps

Generally, flood hazard assessment is the calculation of adverse effects of flooding for a particular area. One or more parameters, such as flood duration, flood depth, flood wave velocity and rate of rise of water level can be used to estimate flood hazard, which mainly depends on the area investigated and the characteristics of the flood (UN 1991). In this study, flood depth was considered in estimating flood hazard. A simple and modified procedure, similar to the technique used by (Dewan et al., 2007, Islam and Sado, 2000) was adopted in this study for flood hazard assessment.

In order to assess flood hazard for each category of administrative unit, elevation data, a weighted score was estimated and hazard ranks were decided. This was accomplished by overlaying the GIS database onto the derived flood depth maps. The following steps were involved in assessing the flood as hazard. a. Floodwater depth maps were overlaid with administrative unit, landuse pattern to estimate the percent of area occupied by each category of floodwater depth. b. A weighted score for the acquired area percentage of each category of administrative unit, landuse pattern was estimated using equation 5-3.

Methodology and Model Setup Weighted score= Class 1*1.0+Class 2*3.0+Class 3*5.0+Class 4*7.0 (5-3) After calculating the weighted score, points for each category of administrative unit, landuse pattern was estimated on the basis of a linear interpolation between 0 and 10. c. On the basis of estimated points for each category of administrative unit, landuse pattern, hazard ranks were determined. This was carried out by deciding four hazard ranks. The higher the ranking value, the more susceptible that particular class was considered to be vulnerable of flooding.

Chapter Six RESULT AND DISCUSSION

6.1 CALIBRATION OF HEC-RAS MODEL

The 2013 dataset is used for calibration of the model and it is validated with 2014 dataset. Flow hydrographs from rating curve and stage hydrograph have been used as a boundary parameter. The mean daily water level data at observed station is compared with model- simulated output for calibration. In this study, the one dimensional long channel Dharla River model has been simulated using the daily hydrograph for twelve months from January to December. The channel is 56 km long. Various Manning’s roughness coefficients, n= 0.010, 0.015, 0.025, 0.045, are adopted for respective simulation. All simulations are run to the unsteady state. Results show that the upstream water surface elevation is increased by increasing the Manning’s roughness coefficients value. Those are consistent with the law of water flowing. Finally, ‘n’ value as 0.025 have been fixed as Manning’s ‘n’ for all the cross sections in Dharla River which gives acceptable result. The comparison of observed and simulated stage hydrograph at Taluk-Simulbari gauging station for Manning’s ‘n’ value of 0.025 for main channel is shown in Figure 6-1.

Typical value of Manning’s roughness coefficient ‘n’ is 0.025 for river (Chow et al., 1988) which indicates our Manning’s value is acceptable. It can be shown from Figure 6- 1 that the trend and shape of the simulated and observed hydrograph are almost similar. The simulated water level and observed water level almost matches from June to October. There are little differences from January to May and November to December. This may be occurred because of low flow profile in the beginning of the year and end of the year. From the simulation it has been found that highest and lowest simulated water level is 32.26 m and 26.73 m in 12 July 2013 and 8 April 2013.

In unsteady calibration, the coefficient of determination R2 and Nash Sutcliffe Efficiency (NSE) have been found 0.951 and 0.47 respectively which indicate that the simulated value is closer to the observed value. Statistical parameters for the calibration are shown in Figure 6-2.

Result and Discussion

34 Observed WL Simulated WL

30 Stage(m)

Jan/1 Mar/1 May/1 Jul/1 Sep/1 Nov/1 Jan/1 Date

Figure 6-1: Observed and simulated stage hydrograph from 1st January, 2013 to 1st January, 2014

32 NSE=0.47

2 31 R =0.951

29 Simulated WL (m) 28

26 27 28 29 30 31 32 Observed WL (m)

Figure 6-2: Statistical parameter of unsteady flow calibration, 2013

Chapter 6

6.2 VALIDATION OF HEC-RAS MODEL

Using the calibrated Manning’s roughness coefficient (n) value, validation for the model has been performed for the year of 2014. The result of validation has also showed satisfactory. The comparison of observed and simulated stage hydrograph at Taluk- Simulbari gauging stations are shown in Figure 6-3. The Figure 6-3 shows the simulated stage hydrograph is in close agreement with observed hydrograph where there were slight differences in simulated water level and observed water level as like as calibration. In unsteady validation, the coefficient of determination R2 and Nash and Sutcliffe Efficiency (NSE) have been found 0.983 and 0.79 respectively which indicate that the validated value is closer to the observed value (F igure 6-4).

34 Observed WL Simulated WL

30 Stage(m)

Jan/1 Mar/1 May/1 Jul/1 Sep/1 Nov/1 Jan/1 Date

Figure 6-3: Observed and simulated stage hydrograph from 1st January, 2014 to 1st January, 2015

Unsteady flow calibration and validation statistical parameter are summarized in Table 6- 1

Table 6-1: Model evaluation parameters

Flow Condition R2 NSE Calibration (2013) Unsteady Flow 0.951 0.47 Validation (2014) Unsteady Flow 0.983 0.79

Result and Discussion

NSE=0.79 31 R2=0.983

29 Simulated WL (m)

27 27.5 28 28.5 29 29.5 30 30.5 31 31.5 Observed WL (m)

Figure 6-4: Statistical parameter of unsteady flow validation, 2014

6.3 QUALITATIVE COMPARISON BETWEEN MODEL SIMULATED AND OBSERVED FLOOD MAP (SATELLITE IMAGE)

After the calibration and validation of the one dimensional HEC-RAS model, an attempt has been taken to develop an one dimensional and two dimensional coupling model. It has done on purpose of 1D flow of water in the main channel and 2d flow of water in flood plain of Dharla River while water overflows it’s main channel banks. 1D and 2D run shows inundation scenarios of flood that is important to check the flood delineation area and depth. In this section qualitative comparison between model and observed flood maps is shown.

6.3.1 Qualitative Comparison between Model and Observed Satellite Image (28 July 2017)

Qualitative comparison on 28 July 2017 between model simulated flood map and observed available satellite image of Sentinel-2 has been observed. For the better comparison purpose image has been categorized as water and non-water zone. Figure 6-5 shows the comparison between observed flood map and model simulated flood map.

Chapter 6

Common places of inundated area between model simulated flood map and observed flood map have been marked by circle to visualize in Figure 6-6. From figure it can be seen that the inundation areas between simulated and observed are adequately alike.

Figure 6-5: Qualitative comparison between model flood map and satellite image

Result and Discussion

6.4 ANALYSIS OF MODEL SIMULATED FLOOD, YEAR 2017

Qualitative comparison between observed and simulated inundation map gives satisfactory result, further analysis based on flow of 2017 is performed. Flow hydrograph of 2017 from rating curve has been used at Taluk-Simulbari station and stage hydrograph of 2017 has been used at Kurigram station as boundary condition to perform unsteady simulation. The simulation has been performed from May to September. It has been performed for analysis the changing of flood pattern with time. Simulated water profile is used to generate flood inundation maps and flood depth maps of 2017 on different date which are presented and discussed in the following sub-sections.

6.4.1 Flood Inundation Map and Depth Analysis

Floodplain delineation is done by HEC RAS 5.0 software itself and shown in RAS Mapper. No GIS help is needed here. Shape file of different flood extend can be produced for different time. Then the shape file has been exported into GIS and calculated the area of inundation including the main channel.

I. Flood Inundation Map Analysis

Maps of inundation area of particular dates of 10 May 2017, 10 June 2017, 10 July 2017, 10 August 2017 and 20 September 2017 are shown in Figure 6-7 to 6-10. Inundation area and the percentage of inundation area with those dates are shown in Table 6-2. Maximum and minimum inundation areas are 37.6% on 20 September 2017 6.28% on 10 May 2017 respectively. For other dates of 10 June 2017, 10 July 2017 and 10 August 2017, percentages of inundation area are 14.04, 26.96 and 34.6 respectively. The total floodplain area is about 1553.4 square kilometers (sq km). It is observed that the flood inundation areas are 97.55, 218.14, 418.83, 533.78 and 584.1 sq km for 10 May 2017, 10 June 2017, 10 July 2017, 10 August 2017 and 20 September 2017 respectively. Figure 6- 11 shows the relationship between inundation area and the particular dates that are considered. From the figure it can be seen that inundation area increases as time passes from May to September, 2017. From the figure it can be seen that maximum inundation occurs in September.

Chapter 6

Figure 6-6: Flood inundation map developed by model simulation at Dharla River floodplain on 10 May 2017

Result and Discussion

Figure 6-7: Flood inundation map developed by model simulation at Dharla River floodplain on 10 June 2017

Chapter 6

Figure 6-8: Flood inundation map developed by model simulation at Dharla River floodplain on 10 July 2017

Result and Discussion

Figure 6-9: Flood inundation map developed by model simulation at Dharla River floodplain on 10 August 2017

Chapter 6

Figure 6-10: Flood inundation map developed by model simulation at Dharla River floodplain on 20 September 2017

Result and Discussion

Table 6-2: Inundation area on different dates of 2017 Date Inundated Area Inundated Area (%) (km2) 10/5/2017 97.55 6.28 10/6/2017 218.13 14.04 10/7/2017 418.82 26.96 10/8/2017 533.78 34.36 20/09/2017 584.10 37.60

600

500

400

300

200 Inundated Area Inundated (sq.km)

100

0 May/7 Jun/11 Jul/16 Aug/20 Sep/24 Date

Figure 6-11: Trend of model simulated inundation area at Dharla River floodplain in 2017

II. Flood Inundation Depth Analysis

Flood inundation depth of the study area for 10 May 2017, 10 June 2017, 10 July 2017, 10 August 2017 and 20 September 2017 were observed in Figure 6-6 to Figure 6-10. Depth map have been created in RAS Mapper for the particular dates and then those raster file have been exported into GIS. In the study, inundated areas are defined into four qualitative inundation depth classes viz. High Land (0.3m - 0.9m), Medium Land (0.9m - 1.2 m), Low Land (1.2m - 3.6m) and Very Low Land ( >3.0m) based on the inundation

Chapter 6 depth. The results of this assessment are summarized in Table 6-3 and Figure 6-12. The classification of flood depth areas indicates that through the monsoon period of 2017, water depth of 1.7 m to 3.6 m increases from 15 sq km. to 173 sq km of the total flooded area. On an average of 98.7 sq km and 84.5 sq km areas have been inundated by water depth up to 0.9m and 1 to 1.8m respectively. Water depth greater than 3.6 m prevails through the monsoon period from 3.6 sq km to 156.8 sq km.

Table 6-3: Calculation of flood area according to inundation depth, 2017 Date in 2017 Flood Types F1 F2 F3 F4 Total Water 0.0 - 0.9 0.9 - 1.8 1.8 - 3.6 > 3.6 Depth(m)

10-May Area(sq km) 45.61 33.58 14.72 3.62 97.55 10-Jun Area(sq km) 103.36 76.79 30.50 7.46 218.13 10-Jul Area(sq km) 90.98 82.66 122.57 122.59 418.82 10-Aug Area(sq km) 120.53 109.09 157.85 146.29 533.78 20-Sep Area(sq km) 132.86 120.42 174.05 156.76 584.10

The Figure 6-13 below shows that the total area inundated by water depth among 0.9m to 1.8m,1.8 m to 3.6 m and more than 3.6 m increased considerably with the increase in the intensity of flooding. Total area under water depth of less than 0.9 m and between 0.9m to 1.8m were increased but then showed a decreased value with time and then again it’s area started increasing with the time.

200

F1 F2 F3 150 F4

100

50 Innundated Area (sq.km)

0 10-May 10-Jun 10-Jul 10-Aug 20-Sep Date

Figure 6-12: Inundated area according to inundation depth, 2017

Result and Discussion

6.4.2 Flood Affected Frequency

The concept of flood-affected frequency was used to assess flood hazard by (Islam and Sado, 2000). Flood affected frequency used in this study is for general purpose to observe the most common damages areas for a single flood event. A flood affected frequency map based on flood duration was developed using model simulation of 2017. A total of four inundation map on 10 May, 10 June, 10 July and 10 August from model simulation of 2017 were used to develop flood frequency map. The shape file of 10 May, 10 June, 10 July and 10 August were combined to construct a single image for 2017. This combination provided an opportunity to obtain a common boundary needed to develop a flood-affected frequency map. Figure 6-13 shows a flood affected frequency map where it the common damages areas for consecutive days can be seen.

Figure 6-13: Flood affected frequency map

The pixels that appeared in inundated area in all images were considered the very highly damaged areas. The common inundated areas that appeared in any three of the images and two of the images were deemed as high- and medium-damaged areas, respectively.

Chapter 6

The other areas that appeared in four images but not in common in any combination are deemed as low-damage area. An inundated area that did not appear in any of the images was considered as non-damaged area. Thus, five flood-affected frequency categories were obtained, corresponding to damage rankings of class 1, class 2, class 3, class 4 and class 5 as non-, low-, medium-, high-, and very high-damaged areas, respectively.

6.4.3 Development of Hazard Map

One or more parameters, such as flood affected frequency, flood depth, flood wave velocity and rate of rise of water level can be used to estimate flood hazard, which mainly depends on the area investigated and the characteristics of the flood (UN 1991). In this study, flood depth has been considered in estimating flood hazard administrative unit map of upazila and agricultural landuse map. To use the flood depth as a hydraulic parameter to develop hazard its need to be classified. On this purpose inundation depth from 0 to 0.9 m has been classified as class 1 (F1), 0.9 m to 1.8 m inundation depth was classified as class 2 (F2), inundation depth from 1.8 m to 3.6 m was classified as class 3 (F3) and inundation depth more than 3.6 m is classified as class 4 (F4). Class 1, class 2, class 3 and class 4 were considered as low-, medium-, high-, very high hazardous. Using this depth classification hazard maps were prepared on administrative unit map of upazila and agricultural landuse map. Ascending order of hazard rank indicates the increment of impact of hazard. Five model simulated depth map have been used to develop hazard map on administrative unit map of upazila and agricultural landuse map. The depth maps used to developed hazard map were 10 May 2017, 10 June 2017, 10 July 2017, 10 August 2017 and 20 September 2017.

I. Hazard Map on Administrative Unit

Bangladesh is divided into 8 Divisions and 64 Districts, although these have only a limited role in public policy. For the purposes of local government, the country is divided into Upazila, Municipalities, City Corporations and Union Council. In this study, upazila administrative units have been considered. There were ten upazilas in the study area in administrative unit map of upazila that are shown in Figure 6-14. Based on this map, five hazard maps on 10 May 2017, 10 June 2017, 10 July 2017, 10 August 2017 and 20 September 2017 has been developed (Figure 6-15 to 6-19).

Result and Discussion

It has been found that all upazilas were more or less vulnerable to flooding in 2017. Among them, the upazilas along with the Dharla River were found more hazardous zone of flood in 2017. It is observed from Table 6-4 and Figure 6-20 that the most hazardous upazila in study area is Lalmonirhat Sadar and the second most hazardous is Kurigram Sadar. These two upazila were inundated on five individual dates of 2017 from May to September. Lalmonirhat Sadar was hazard rank four in four days from June to Septmber and Kurigram Sadar was hazard rank four for two days. Phulbari is highly hazardous since it was inundated on five considered date and hazard was ranked four on 10 June 2017. Nageshwari and Ulipur upazilas are same hazardous upazila and ranked after Phulbari. These two upazilas were inundated for five individual dates but Nageshwari ranked four for four days, three for one day whereas Ulipur ranked four for one day, two for one day, three for one day and one for two days. So comparing the trend of ranking of these two upazilas, it can be said that Nageshwari and Ulipur upazilas are high medium hazardous zone. Aditmari and Rajarhat are on an average showed hazard rank two. These two upazilas are less hazard than Nageshwari and Ulipur upazila. After that Kaliganj stands low hazardous zone which was not hazardous zone on 10 May, but ranked two for two days and one for two days. Lastly very low hazardous zones are Bhurungamari and Chilmari which are located in most outer sides of the study area.

Chapter 6

Figure 6-14: Administrative unit map of study area

Figure 6-15: Hazard map on administrative unit on 10 May 2017

Result and Discussion

Figure 6-16: Hazard map on administrative unit on 10 June 2017

Figure 6-17: Hazard map on administrative unit on 10 July 2017

Chapter 6

Figure 6-18: Hazard map on administrative unit on 10 August 2017

Figure 6-19: Hazard map on administrative unit on 20 September 2017

Result and Discussion

Table 6-4: Hazard rank on administrative unit (upazila), 2017 Upazila/ate 10-May-17 10-Jun-17 10-Jul-17 10-Aug-17 20-Sep-17 Phulbari 3 4 3 3 3 Kurigram Sadar 4 4 3 3 3 Rajarhat 1 1 2 2 3 Ulipur 1 1 2 3 4 Lalmonirhat Sadar 3 4 4 4 4 Kaliganj 0 1 1 2 2 Aditmari 0 2 2 3 3 chilmari 0 0 0 1 1 Nageshwari 0 2 4 4 4 Bhurungamari 0 0 1 1 1

Date of 3 Observation

10-May-17 10-Jun-17 2 Hazard rank Hazard 10-Jul-17 10-Aug-17 1 20-Sep-17

Adminitrative Unit (Upazilla)

Figure 6-20: Observation of hazard rank on administrative unit (upazila) of study area

II. Hazard Map on Agricultural Landuse Types

Agriculture is the largest employment sector in Bangladesh. A plurality of Bangladeshis earn their living from agriculture. The floods caused severe damage to the agriculture sector, including crop losses of the main food staple rice, with most of the damage

Chapter 6 concentrated in the northern districts. In case of a high devastating situation, however, loss to agriculture is enormous as improved technical information is not yet available. Crop hazard management as a non-structural measure can be an application of measures to avoid or to minimize the impact of flood on various agricultural development activities. Essentially it means regulating cropping and other land use in order to reduce the vulnerability of crops, livestock, fisheries, forestry, people and properties. To serve the purpose a land use map (Figure 6-21) has been used to find the most hazardous rank of crops that exist in the study area using the model simulation flood on 20 September, 2017. A shape file (Figure 6-22) of existing crops pattern has been extracted from the agricultural landuse map and hazard rank has been made on agricultural landuse in the study area which is shown in Figure 6-23.

Figure 6-21: Agricultural landuse map [Source: www.banglapedia.org. accessed on 13th January, 2018] 100

Result and Discussion

Figure 6-22: Crops pattern in the study area

Figure 6-23: Hazard rank on agricultural landuse in the study area in 2017

101

Chapter 6

From Figure 6-22 it can be seen that there are three types of three combinations crops in the study area. They area Boro - Fallow - T.aman, Rabi Crop - Aus - T.aman, Rabi Crop - B.aus - Fallow. Figure 6-23 shows three hazard ranks where hazard rank three is the most hazardous zone, hazard rank two is medium hazard and hazard rank one is low hazard rank. From the model simulated flood map, it was found that Boro - Fallow - T.aman was the most vulnerable to flood in 2017, Rabi Crop - Aus - T.aman was found less vulnerable to flooding than Boro - Fallow - T.aman. And the lowest flood hazardous crop to flooding in 2017 was Rabi Crop - B.aus - Fallow.

6.5 ANALYSIS OF HISTORICAL FLOOD EVENT, 1998

History of floods in this country is perhaps inseparable from the history of this land. In every century, bengal delta witnessed visit of nearly half a dozen floods. The floods of 1987, 1988 and 1998 were catastrophic, leading to widespread destruction, misery and loss of life. An investigation has been done in this study concerning flood scenarios of historical flood event 1998 in the study area. In 1998, over two-thirds of the total area of the country was flooded. It compares with the catastrophic flood of 1988 so far as the extent of flooding is concerned. A combination of heavy rainfall within and outside the country, synchronisation of peak flows of the major rivers and a very strong backwater effect coalesced into a mix that resulted in the worst flood in recorded history. The flood of 1998 lasted for more than two months. Flow hydrograph of 1998 from rating curve has been used at Taluk-Simulbari station and stage hydrograph of 1998 has been used at Kurigram station as boundary condition to perform unsteady flow simulation. The simulation has been performed from May to September. It has been performed for analysis the changing of flood pattern with time. Simulated water profile is used to generate flood inundation maps and flood depth maps of 1998 on different date which are presented and discussed in the following sub-sections.

6.5.1 Flood Inundation Map and Depth Analysis

Flood inundation is done by HEC RAS and shown in RAS Mapper. GIS help was not needed here. Shape files of different flood extend have been produced for different time. Later the shape file has been exported into GIS and the area of inundation including the main channel were calculated for flood inundation analysis. Classification of flood depth has been defined and the area of individual classes of flood depth has been determined.

102

Result and Discussion

I. Flood Inundation Map Analysis

Five consecutive days from May to September have been observed for the analysis of inundation scenarios in the study area in 1998. The observation dates are 10 May, 10 June, 10 July, 10 August and 20 September. Model simulated inundation flood map of those particular dates are shown in Figure 6-24 to 6-28. Inundation area and the percentage of inundation area are shown in Table 6-5. Maximum and minimum inundation areas are 45.2 % on 10 August 1998 and 12.45 % on 10 May 2017 respectively. Percentages of inundation area on other dates are 30.22, 44.15, 40.63 on 10 June 1998, 10 July 1998 and 20 September 1998 respectively. The total floodplain area is about 1553.4 square kilometers (sq km). It is observed that the flood inundation areas are 193.48, 469.39, 685.57, 701.91and 631.044 sq km for 10 May 1998, 10 June 1998, 10 July 1998, 10 August 1998 and 20 September 1998 respectively. Figure 6-29 shows the relationship between inundation area and the particular dates that are considered. From the figure it can be seen that inundation area increases from May to August and it shows a decrease trend from August to September. The maximum inundated area was found in August, 1998. On an average, it has been found that about 35 percent of the total study area were inundated.

103

Chapter 6

Figure 6-24: Flood inundation map developed by model simulation at Dharla River floodplain on 10 May 1998

104

Result and Discussion

Figure 6-25: Flood inundation map developed by model simulation at Dharla River floodplain on 10 June 1998

105

Chapter 6

Figure 6-26: Flood inundation map developed by model simulation at Dharla River floodplain on 10 July 1998

106

Result and Discussion

Figure 6-27: Flood inundation map developed by model simulation at Dharla River floodplain on 10 August 1998

107

Chapter 6

Figure 6-28: Flood inundation map developed by model simulation at Dharla River floodplain on 20 September 1998

108

Result and Discussion

Table 6-5: Inundation area on different dates of 1998 Date Inundated Area Inundated Area (km2) (%) 5/10/1998 193.48 12.45 6/10/1998 469.39 30.22 7/10/1998 685.57 44.15 8/10/1998 701.911 45.2 9/20/1998 631.044 40.63

800

700

600

500

400

300 Inundated Area (sqkm) InundatedArea (sqkm)

200

100 May/10 Jun/14 Jul/19 Aug/23 Sep/27 Date

Figure 6-29: Trend of model simulated inundation area at Dharla River floodplain in 1998

II. Flood Inundation Depth Analysis

Flood inundation depth of the study area on 10 May 1998, 10 June 1998, 10 July 1998, 10 August 1998 and 20 September 1998 were observed in Figure 6-24 to Figure 6-28. Depth map have been created in RAS Mapper for the particular dates and then those raster file have been exported into GIS. The same qualitative inundation depth classes based on the inundation depth were used for the analysis of flood depth in the study area. The results of this assessment are summarized in Table 6-6 and Figure 6-30. From Table 6-6 it has been found that area of F1 increases from 88.78 sq km to 308.73 sq km and decreases to

109

Chapter 6

290.28 sq km. Area of F2 increases from 68.92 sq km to 120.02sq km and decreases to 92.79 sq km. Area of F3 increases from 29.28sq km to 249.37 sq km and decreases to 229.15 sq km. And the area of F4 increases from 6.5 sq km to 23.79 and decreases to 18.85 sq km.

Table 6-6: Calculation of flood area according to inundation depth, 1998 Date in 1998 Flood Types F1 F2 F3 F4 Total Water 0.0 - 0.9 0.9 - 1.8 1.8 - 3.6 > 3.6 Depth(m) 10-May Area(sq km) 88.78 68.92 29.28 6.5 193.48 10-Jun Area(sq km) 233.15 157.41 64.23 14.58 469.37 10-Jul Area(sq km) 308.85 244.94 109.92 21.85 685.56 10-Aug Area(sq km) 308.73 249.37 120.02 23.79 701.91 20-Sep Area(sq km) 290.28 229.15 92.79 18.85 631.07

It is observed in Figure 6-30 that during the monsoon periods, area of F1 flood depth was significant than all other flood depth classes. And area of F4 flood depth class was smaller. It has also been observed that all four flood depth class F1, F2, F3 and F4 increases along with the time till August and later the areas decreases.

350

F1 300 F2 F3 F4 250

200

150

100 InundatedArea (sq.km)

0 10-May 10-Jun 10-Jul 10-Aug 20-Sep Date

Figure 6-30: Inundated area according to inundation depth, 1998

110

Result and Discussion

6.5.2 Development of Hazard Map

The study has been conducted to produce the flood hazard map in the study area for the historical flood event of 1998 to delineate the the most flood hazardous zone in that devastating flood event. The hazard map has also been developed on administrative unit map of upazila and agricultural landuse map of study area using water depth as hydraulic parameter. The water depth has been classified as before and used to identify the hazardous rank for each of the classification on administrative unit map and agricultural landuse map. Hazard map on administrative unit map of upazila has been generated on 10 May 1998, 10 June 1998, 10 July 1998, 10 August 1998 and 20 September 1998. Hazard map on agricultural landuse map has been developed using the model simulated flood map of 20 September 1998.

I. Hazard Map on Administrative Unit

There were ten upazilas in the study area on which hazard rank has been determined using the flood depth classification. The administrative unit map of upazila has been shown in Figure 6-31. Five developed hazard maps on 10 May 1998, 10 June 1998, 10 July 1998, 10 August 1998 and 20 September 1998 has been shown in Figure 6-32 to 6- 36. Observations of hazard rank on administrative unit maps in 1998 has been summarized in Table 6-7 and Figure 6-37.

It was found that there were no upazila left that was not flooded due to flood inundation of Dharla River floodplain in 1998. Among the ten upazilas, Lalmonirhat Sadar was the most high hazardous upazila which is inundated in five considered days and the upazila ranked four in all five days. Chilmari, Bhurungamari and Kaliganj upazilas were found to be the most less hazardous zones due to flood inundation of Dharla River floodplain. The second most high hazardous zone is Phulbari and Kurigram Sadar is found to be third high hazardous zone after Phulbari upazila. Phulbari was identified as hazard rank four in two days and three in three days and Kurigram Sadar was identified as hazard rank three in all five observed days. Aditmari and Nageshwari upazilas were found fourth hazardous zone. Lastly, Ulipur and Rajarhat were found to be the sixth hazardous upazilas due to occurrence of flooding in Dharla River in 1998.

111

Chapter 6

Figure 6-31: Administrative unit map of study area

Figure 6-32: Hazard map on administrative unit on 10 May 1998

112

Result and Discussion

Figure 6-33: Hazard map on administrative unit on 10 June 1998

Figure 6-34: Hazard map on administrative unit on 10 July 1998

113

Chapter 6

Figure 6-35: Hazard map on administrative unit on 10 August 1998

Figure 6-36: Hazard map on administrative unit on 20 September 1998

114

Result and Discussion

Table 6-7: Hazard rank on administrative unit (upazila), 1998 Upazilla/Date 10-May-98 10-Jun-98 10-Jul-98 10-Aug-98 20-Sep-98 Phulbari 4 4 3 3 3 Kurigram Sadar 3 3 3 3 3 Rajarhat 1 2 2 2 2 Ulipur 1 1 3 3 4 Lalmonirhat 4 4 4 4 4 Sadar Kaliganj 1 1 2 1 1 Aditmari 2 2 3 3 3 chilmari 0 0 1 1 1 Nageshwari 1 3 3 3 3 Bhurungamari 0 1 1 1 1

3 Date of Observation 10-May-98 2

Hazard Rank Hazard 10-Jun-98 10-Jul-98 1 10-Aug-98 20-Sep-98

Adminitrative Unit (Upazilla)

Figure 6-37: Observation of hazard rank on administrative unit (upazila) of study area

115

Chapter 6

II. Hazard Map on Agricultural Landuse Types

Flood hazard map has been developed on agricultural landuse pattern in the Dharla River floodplain using the model simulated results in 1998. An agricultural landuse map in the study area has been made using a collected agricultural landuse map of Bangladesh. The developed agricultural landuse map of the study area has been shown in Figure 6-38. From this Figure 6-38, it can be seen that there are three types of three combinations crops in the study area. They area Boro - Fallow - T.aman, Rabi Crop - Aus - T.aman, Rabi Crop - B.aus - Fallow. There are three hazard ranks were assigned where hazard rank three is the most hazardous zone, hazard rank two is medium hazard and hazard rank one is low hazard rank. A flood hazard map has been developed using the model simulation flood inundation depth map on 20 September 1998.

The developed flood hazard map on agricultural landuse types has been shown in Figure 6-39. It has been found that Boro - Fallow - T.aman was the most vulnerable to flood in 1998 in the Dharla River floodplain, Rabi Crop - Aus - T.aman was found less vulnerable to flooding than Boro - Fallow - T.aman. And the lowest flood hazardous crop to flooding in 1998 was Rabi Crop - B.aus - Fallow.

Figure 6-38: Crops pattern in the study area

116

Result and Discussion

Figure 6-39: Hazard rank on agricultural landuse in the study area in 1998

117

Chapter Seven CONCLUSIONS AND RECOMMENDATIONS

7.1 CONCLUSIONS

For development of inundation model in Dharla River floodplain, 1D/2D couple model has been used in this study. Geographical Information System has been used for further development of flood hazard map from the inundation results of the model. Calibration and validation of 1D hydrodynamic model have been performed in 2013 and 2014 respectively where Manning’s roughness co-efficient ‘n’ was the key parameter for calibration and validation. For calibration and validation, the water level data of the upstream station (Taluk-Simulbari) has been used. The result showed a good correlation between the observed and simulated water level data for Manning’s roughness coefficient ‘n’ of 0.025. The value of correlation coefficient, R2 is 0.951 and 0.983 for calibration and validation respectively and the value of NSE is 0.47 and 0.79 for calibration and validation respectively which were found in acceptable range.

In this study, 1D/2D coupled hydrodynamic model has been used to develop flood inundation model of Dharla River floodplain. Floodplains have been considered as 2D flow area and Dharla River has been considered as 1D flow. Later, the model has been simulated for 2017 and 1998. Comparison has been made between model simulated flood map and observed flood map. Remote sensing data were integrated with GIS tool in this study for comparing with the model simulated flood map. Available satellite image on 28 July 2017 has been used for the comparison purpose. Comparison between observed flood from satellite imagery and model result was satisfactory which leads this study for further analysis of inundation scenarios and finally hazard mapping. From the analysis of flood inundation areas considering five consecutive days from May to September, it has been found that flood inundated area of Dharla River floodplain increases along passage of time. It has been found that 23.8% and 34% of the total study area were inundated under the flood water in 2017 and 1998 respectively.

Development of flood hazard for Dharla River floodplain by using an integrated model flood map and GIS approach has been discussed. The results described in this study provide a means of assessing the potential flood hazard in the study area, which can be Conclusion and Recommendation used to mitigate the negative impacts of future floods. Flood hazard assessments were undertaken and a flood hazard maps for the study area were developed using the model simulated flood depth map of 2017 and 1998.

The flood hazard maps based on each administrative upazila and landuse pattern represent the magnitude of flood damage with respect to flood depth classification. This type of map helps the responsible authorities to better comprehend the inundation characteristics of the floodplains. The results described in this study provide helpful information for flood control planning, and the construction and development of flood countermeasures in the most flood hazardous zone in Dharla River floodplain.

It has been observed in hazard map on administrative unit map that the most flood hazardous upazila in study area is Lalmonirhat Sadar and second most flood hazardous areas are Kurigram Sadar and Phulbari. Chilmari, Bhurungamari and Kaliganj upazilas were found to be the most less hazardous zones due to flood inundation of Dharla River floodplain. Aditmari and Nageshwari upazilas were found third flood hazardous zone. Lastly, Ulipur and Rajarhat were found to be the fourth flood hazardous upazilas due to occurrence of flooding in Dharla River floodplain.

From hazard map on landuse pattern it was found that boro - fallow - t.aman was the most vulnerable to flood in Dharla River floodplain, rabi crop - aus - t.aman was found less vulnerable to flooding than boro - fallow - t.aman. And the lowest flood hazardous crop to flooding in Dharla River floodplain was rabi crop - b.aus – fallow.

It was observed that top most upside of the study area zone is very low flood hazardous because of their high elevation data in those zone. There is less inundation in this area and it means hazard rank is one or zero. In the middle of the study area falls in medium flood hazardous category where the elevation varies 25 m to 34 m. The downside of the study area is high hazardous because of the low elevation which is from 16 m to 25 m. There is more inundation in this elevation zone which refers it hazard rank four.

Compared to the wide range of research conducted in other flood prone countries, research work carried out in Bangladesh on determination of current status of flood map is very limited. As per current state of knowledge any study regarding Dharla River floodplain inundation mapping using 1D/2D coupled model could not found yet. This study has been conducted using GIS tool in a very simple and straight forward method.

119

Chapter 7 For preliminary flood protection measures this method could be very effective. The use of GIS tools is realized to be very effective for flood mapping.

7.2 RECOMMENDATIONS

Bangladesh is a flood-prone country as being located at the confluence of three large rivers the Ganges, Brahmaputra and Meghna. The objective of the study is to develop calibrated hydrodynamic model and hazard map in Dharla River floodplain. In this study flood inundation maps is prepared with SRTM data at 30 m resolutions. Hydrodynamic model is used to develop flood inundation map and hazard map with the flooding of Dharla River. Some actions can be recommended for the improvement of this study for future prediction: i. The study has been done using 1D and 2 D coupling. For the comparison purpose or for better understanding pure 2D analysis can be performed in future. ii. Digital elevation model plays vital role to enhance the capability of model. It is recommended to use high-resolution digital spatial database for real replication of topography for the better performance of the model. iii. For comparison purpose between model simulation and real time data, Radarsat images should be used if available which can capture ground condition neglecting the cloudy sky. iv. For hazard mapping, inundation depth has been assigned in this study. But other factors such as frequency of flood, duration of flood, velocity can be considered. v. Generation of flood risk map can be performed for future study.

120

REFERENCES

Ahmad, B., Kaleem, M.S., Butt, M.J., and Dahri, Z.H., 2010. Hydrological modelling and flood hazard mapping of Nullah Lai. Proc. Pakistan Acad. Sci. 47, 215–226.

Alam, K. and Abdullah, S.K.M., 1990. Geological map of Bangladesh. Geological Survey of Bangladesh.

Betsholtz, A., and Nordlöf, B., 2017. Potentials and limitations of 1D, 2D and coupled 1D-2D flood modelling in HEC-RAS. Lund University.

Bhuiyan, M.S.R., 2014. Flood Hazard and Vulnerability Assessment in a Riverine Flood Prone Area: A Case Study. Bangladesh University of Engineering and Technology (BUET).

Brammer, H., and Khan, H., 1991. Bangladesh case study. In: Disaster mitigation in Asia and the Pacific, Asian Development Bank, Manila.

Brunner G.W., 2016 Combined 1D and 2D Modeling with HEC-RAS. US Army Corps of Engineers, Institute for Water Resources, Hydrologic Engineering Center.

Brunner, G.W., Piper, S.S., Jensen, M.R., and Chacon, B., 2015. Combined 1D and 2D Hydraulic Modeling within HEC-RAS. World Environ. Water Resource Congress. 1432– 1443. https://doi.org/10.1061/9780784479162.141

Carrier, C.A., Kalra, A., and Ahmad, S., 2016. Long-range precipitation forecasts using paleoclimate reconstructions in the western United States. J. Mt. Sci. 13, 614–632. https://doi.org/10.1007/s11629-014-3360-2

Chow, V. Te, Maidment, D.R., and Mays, L.W., 1988. Applied Hydrology. McGraw- Hill, New York.

Chow, V.T., 1959, Open Channel Hydraulics, McGraw-Hill, New York. References Dewan, A.M., Kabir, M.H., Islam, M.M., Kumamoto, T., and Nishigaki, M., 2007. Delineating flood risk areas in greater Dhaka of Bangladesh using geoinformatics. Georisk 1, 190–201. https://doi.org/10.1080/17499510701772097

Dilley, M., Chen, R.S., Deichmann, U., Lerner-Lam, A.L., Arnold, M., Agwe, J., Buys, P., Kjevstad, O., Lyon, B., and Yetman, G., 2005. Natural disaster hotspots : a global risk analysis. World Bank.

Disaster Management Bureau, 2010. National plan for disaster management, Disaster Management and Relief Division, Government of the People’s Republic of Bangladesh.

FFWC, 2005. Flood Annual Report 2005. Flood Forecasting and Warning Centre. Bangladesh Water Development Board, Government of the People’s Republic of Bangladesh Ministry of Water Resources, Dhaka.

FFWC, 2012. Annual Flood Report 2012, Flood Forecasting and Warning Centre. Bangladesh Water Development Board, Government of the People’s Republic of Bangladesh Ministry of Water Resources, Dhaka.

FFWC, 2014. Annual Flood Report 2014, Flood Forecasting and Warning Centre. Bangladesh Water Development Board, Government of the People’s Republic of Bangladesh Ministry of Water Resources, Dhaka.

FFWC, 2015. Annual Flood Report 2015, Flood Forecasting and Warning Centre. Bangladesh Water Development Board, Government of the People’s Republic of Bangladesh Ministry of Water Resources, Dhaka.

GUK, 2016. Flood Situation Report: 30 July 2016. Gana Unnayan Kendra (GUK), Gaibandha, Bangladesh.

Gupta, H. V., Kling, H., Yilmaz, K.K., and Martinez, G.F., 2009. Decomposition of the mean squared error and NSE performance criteria: Implications for improving hydrological modelling. Journal of Hydrology 377, 80–91. https://doi.org/10.1016/J.JHYDROL.2009.08.003

Hazarika, M., 2007. Capacity Building in Applications of Geoinformatics in Flood Hazard Mapping.

122

References HCTT, 2014. Joint Needs Assessment: Flooding in north-western Bangladesh - 08 September 2014. Report prepared by Multi Stakeholders Team for Humanitarian Coordination Task Team (HCTT).

HEC-RAS, 2016. River Analysis System: Hydraulic Reference Manual. Hydrologic Engineering Center, California.

Hicks, F.E., and Peacock, T., 2005. Suitability of HEC-RAS for flood forecasting. Canadian Water Resources Journal. 30, 159–174. https://doi.org/10.4296/cwrj3002159

Hossain, A.N.H.A., 2003. Bangladesh: Flood Management, Water Resources. WMO/GWP Associated Programme on Flood Management.

Hossain, M.S., 2013. Flood Damage and Risk Assessment Model in the Haor Basin of Bangladesh. MSc Thesis. Bangladesh University of Engineering and Technology (BUET). Dhaka

Hossain, M.S., 2015. Local level flood forecasting system using mathematical model incorporating WRF model predicted rainfall. MSc Thesis. Bangladesh University of Engineering and Technology (BUET). Dhaka.

IPCC, 2013. Climate Change 2013 - The Physical Science Basis, Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press, Cambridge and New York. https://doi.org/10.1017/CBO9781107415324

IPCC, 2014. Climate Change 2014: Synthesis Report, Contribution of Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Core Writing Team, R.K. Pachauri and L.A. Meyer (eds.)]. Geneva.

Islam, A.S., Bala, S.K., and Haque, M.A., 2010. Flood inundation map of Bangladesh using MODIS time-series images. Journal of Flood Risk Management. 3, 210–222. https://doi.org/10.1111/j.1753-318X.2010.01074.x

123

References Islam, M.M., and Sado, K., 2000. Development of flood hazard maps of Bangladesh using NOAA-AVHRR images with GIS. Hydrology Sciences Journal 45(3), 337–355. https://doi.org/10.1080/02626660009492334

Islam, M.S., and Tsujimoto, T., 2012. Comparisons of Levee Breach and Successive Disasters in Floodplain between Bangladesh and Japan. Procedia Engineering, Elsevier publication, 28, 860-865. https://doi.org/10.1016/j.proeng.2012.01.824

Kalra, A., and Ahmad, S., 2011. Evaluating changes and estimating seasonal precipitation for the Colorado River Basin using a stochastic nonparametric disaggregation technique. Water Resources Research. 47, 1-26. https://doi.org/10.1029/2010WR009118

Kalra, A., and Ahmad, S., 2012. Estimating annual precipitation for the Colorado River Basin using oceanic-atmospheric oscillations. Water Resources Research. 48(6). https://doi.org/10.1029/2011WR010667

Kalra, A., Li, L., Li, X., and Ahmad, S., 2013a. Improving streamflow forecast lead time using oceanic-atmospheric oscillations for Kaidu River Basin, Xinjiang, China. Journal of Hydrologic Engineering 18, 1031–1040. https://doi.org/10.1061/(ASCE)HE.1943- 5584.0000707

Kalra, A., Miller, W.P., Lamb, K.W., Ahmad, S., and Piechota, T., 2013b. Using large- scale climatic patterns for improving long lead time streamflow forecasts for Gunnison and San Juan River Basins. Hydrologic Processes. 27, 1543–1559. https://doi.org/10.1002/hyp.9236

Khan, S.M., Sarker, M.H., Navera, U.K., Rahman, M.A., Alam, S., Das, P., Noor, L., and Islam, F., 2017. Comparison of various modeling software pertaining to hydrodynamics and morphology: Their scopes and applicability.

Legates, D.R., and McCabe, G.J., 1999. Evaluating the use of “goodness-of-fit” Measures in hydrologic and hydroclimatic model validation. Water Resources Research. 35, 233– 241. https://doi.org/10.1029/1998WR900018

Mani, P., Chatterjee, C., and Kumar, R., 2014. Flood hazard assessment with multiparameter approach derived from coupled 1D and 2D hydrodynamic flow model. Natural Hazards. 70(2), 1553–1574. https://doi.org/10.1007/s11069-013-0891-8

124

References Masood, M., 2011. Flood Hazard and Risk Assessment in Mid- Eastern Part of Dhaka, Bangladesh. Bangladesh Water Development Board (BWDB). Dhaka.

Moore, M.R., 2011. Development of a high-resolution 1D/2D coupled flood simulation of Charles City, Iowa. MSc Theses. University of Iowa. Iowa.

Moriasi, D. N., Arnold, J. G., Van Liew, M. W., and Bingner, R. D., 2002. Model evaluation guidelines for systematic quantification of accuracy in watershed simulations. Transactions of the ASABE. 50(3), 885-900.

Nash, J.E., and Sutcliffe, J.V., 1970. River flow forecasting through conceptual models part I — A discussion of principles. Journal of Hydrology 10(3), 282–290. https://doi.org/10.1016/0022-1694(70)90255-6

Nasreen, M., 2004. Disaster Research: Exploring Sociological Approach to Disaster in Bangladesh. Bangladesh e-Journal of Sociology. 1(2), 1-8.

Nirapad, 2017. Flood Situation Report, updated on 22 August 2017. Nirapad. Dhaka.

Patel, D.P., Ramirez, J.A., Srivastava, P.K., Bray, M., and Han, D., 2017. Assessment of flood inundation mapping of Surat city by coupled 1D/2D hydrodynamic modeling: a case application of the new HEC-RAS 5. Natural Hazards 89(1), 93–130. https://doi.org/10.1007/s11069-017-2956-6

Pathak, P., Kalra, A., and Ahmad, S., 2016. Temperature and precipitation changes in the Midwestern United States: implications for water management. International Journal of Water Resources Development. 33, 1003–1019. https://doi.org/10.1080/07900627.2016.1238343

Quiroga, V.M., Kure, S., Udo, K., and Mano, A., 2016. Application of 2D numerical simulation for the analysis of the February 2014 Bolivian Amazonia flood: Application of the new HEC-RAS version 5. RIBAGUA – Revista Iberoamericana del Agua. 3, 25–33. https://doi.org/10.1016/j.riba.2015.12.001

Rahman, M.M., 2015. Modeling flood inundation of the Jamuna river. MSc Thesis. Bangladesh University of Engineering and Technology (BUET). Dhaka.

125

References Rahman, M.M., Arya, D.S., Goel, N.K., and Dhamy, A.P., 2011. Design Flow and Stage Computations in the Teesta River, Bangladesh, Using Frequency Analysis and MIKE 11 Modeling. Journal of Hydrologic Engineering. 16, 176–186. https://doi.org/10.1061/(ASCE)HE.1943-5584.0000299

Rahman, M.M., Hossain, M.A., and Bhattacharya, A.K., 2014. An analytical study of flood management in Bangladesh. IOSR Journal of Engineering. 4, 2250–3021.

Rogala, J. T., and Soballe, D. M., 1998. A GIS model for predicting winter habitat sustainability in backwaters of the upper Mississippi River. Technical Report No. PSR 98-01. United States Geological Survey.

Rouf, T., 2015. Flood inundation map of Sirajgonj district using mathematical model. MSc Thesis. Bangladesh University of Engineering and Technology (BUET). Dhaka.

Sanyal, J., and Lu, X.X., 2004. Application of remote sensing in flood management with special reference to monsoon Asia: A review. Natural Hazards 33, 283–301. https://doi.org/10.1023/B:NHAZ.0000037035.65105.95

Servat, E., and Dezetter, A., 1991. Selection of calibration objective functions in the context of rainfall-runoff modelling in a Sudanese savannah area. Hydrological Sciences Journal. 36, 307–330. https://doi.org/10.1080/02626669109492517

Smith, D.I., 1999. Floods: physical processes and human impacts by K. Smith and R. Ward, 1998. Earth Surface Processes and Landforms. 24, 569–569. https://doi.org/10.1002/(SICI)1096-9837(199906)24:6<569::AID-ESP985>3.0.CO;2-W

Smith, P., 1995. Hydrologic data development system. CRWR Online Report 95-1. Department of Civil Engineering, University of Texas at Austin.

Srinivas, K., Werner, M., and Wright, N., 2009. Comparing forecast skill of inundation models of differing complexity: The case of Upton upon Severn. Flood Risk Management: Research and Practice – Samuels et al. (eds). Taylor & Francis Group. London. ISBN 978-0-415-48507-4.

126

References Sultana, T., 2015. Flash flood forecasting using estimated precipitation by global satellite mapping in the north-east region of Bangladesh. MSc Thesis. Bangladesh University of Engineering and Technology. Dhaka.

Thakur, B., Parajuli, R., Kalra, A., Ahmad, S., and Gupta, R., 2017. Coupling HEC-EAS and HEC-HMS in precipitation runoff modelling and evaluating flood plain inundation map. World Environmental and Water Resources Congress 2017. 240–251.

Udono, T., and Sah, A.K., 2002. Hazard mapping and vulnerability assessment. Regional Workshop on Total Disaster Risk Management. 7-9 August 2002.

USGS, 1998. Geographic Information Systems. United States Geological Survey.

USACE, 2016. HEC-RAS river analysis system. Hydraulic reference manual, Version 5.0. Institute for Water Resources, Hydrologic Engineering Center. US Army Corps of Engineers. California.

WWAP, 2006. Water, a shared responsibility. UN world water assessment report 2. World Water Assessment Programme (WWAP). Paris.

127

Appendix A FEATURES OF MODEL

A-1 Unsteady flow simulation in HEC-RAS

A-2 Data storage capability of HEC-RAS

Features of model A-3 Graphical and tabular output in HEC-RAS

129

Appendix A A-4 Inundation mapping of water surface profile in RAS Mapper

A-5 Representation of vector features

130

Features of model A-6 Raster DEM

A-7 Representation of TIN land surface

A-8 HEC-GeoRAS toolbar

131

Appendix A A-9 HEC-GeoRAS geometry processing menu items

A-10 HEC-GeoRAS mapping menu items

A-11 Summary of HEC-GeoRAS tools

132

Appendix B MORPHOLOGICAL DATA

A-1 Cross-Section data of station RMDL1 in Dharla River on 2nd November 2014 Sl. No. Distance RL Sl. No. Distance RL Sl. No. Distance RL

1 0 26.79 37 835 25.15 73 1860 24.51 2 0 26.59 38 860 25.1 74 1900 24.65 3 10 26.64 39 880 25.07 75 1935 24.6 4 25 26.69 40 910 25.15 76 1970 24.5 5 40 26.75 41 950 25.16 77 2000 24.48 6 55 26.71 42 975 24.7 78 2030 24.67 7 70 26.51 43 1000 24.73 79 2060 24.63 8 85 26.37 44 1030 24.68 80 2085 24.61 9 100 26.34 45 1060 24.54 81 2115 24.66 10 120 25.93 46 1100 24.51 82 2135 24.89 11 140 25.76 47 1130 24.46 83 2150 24.98 12 160 25.6 48 1155 24.37 84 2180 25.61 13 180 25.58 49 1180 24.3 85 2210 25.63 14 200 24.86 50 1200 24.25 86 2250 25.4 15 230 24.89 51 1205 22.82 87 2275 25.39 16 260 24.8 52 1240 22.75 88 2300 25.37 17 290 24.73 53 1270 22.39 89 2330 24.78 18 310 24.61 54 1300 22.16 90 2360 24.44 19 340 24.31 55 1320 22.72 91 2390 24.36 20 370 24.65 56 1350 22.35 92 2410 24.35 21 400 24.63 57 1375 22.31 93 2440 24.24 22 425 24.73 58 1400 22.29 94 2470 24.32 23 450 24.7 59 1430 23.4 95 2490 24.18 24 475 25.69 60 1460 23.36 96 2510 23.81 25 500 25.82 61 1490 23.48 97 2540 23.7 26 530 25.81 62 1520 23.5 98 2570 23.66 27 560 25.79 63 1550 23.58 99 2600 23.56 28 590 25.69 64 1575 24.2 100 2680 23.31 29 630 25.35 65 1600 24.65 101 2720 23.2 30 650 24.79 66 1640 24.78 102 2750 23.26 31 680 24.84 67 1680 24.83 103 2780 23.34 32 705 24.79 68 1710 24.69 104 2810 23.3 33 730 24.9 69 1740 24.54 105 2825 22.25 34 760 24.82 70 1770 24.46 106 2830 21.81 35 790 24.88 71 1800 24.51 107 2836 20.95 36 810 25.12 72 1830 24.56 108 2840 20.75

Appendix B Sl. No. Distance RL Sl. No. Distance RL Sl. No. Distance RL

109 2845 20.45 125 3070 20.35 141 3310 24.38 110 2850 20.15 126 3080 20.45 142 3340 24.48 111 2860 19.85 127 3090 20.65 143 3370 24.25 112 2880 19.75 128 3095 20.75 144 3400 24.46 113 2890 19.85 129 3100 20.75 145 3430 24.48 114 2900 19.95 130 3104 20.95 146 3460 24.53 115 2920 20.05 131 3110 21.75 147 3490 24.54 116 2940 20.15 132 3120 21.92 148 3520 24.49 117 2960 20.15 133 3130 22.59 149 3550 24.38 118 2980 20.25 134 3140 23.12 150 3580 24.35 119 2990 20.35 135 3160 23.33 151 3610 24.58 120 3000 20.35 136 3180 23.47 152 3625 24.65 121 3015 20.35 137 3200 23.69 153 3650 25.1 122 3030 19.95 138 3230 23.81 154 3650 25.2 123 3045 19.95 139 3250 24.15 124 3060 20.15 140 3280 24.09

A-2 Cross-section data of station RMDL2 in Dharla River on 4th November 2014