100228 DIREC TIONS IN DE VELOPMENT Environment and Sustainable Development Urban Flooding of Greater Dhaka in a Changing Climate Building Local Resilience to Disaster Risk Susmita Dasgupta, Asif Zaman, Subhendu Roy, Mainul Huq, Sarwar Jahan, and Ainun Nishat Urban Flooding of Greater Dhaka in a Changing Climate Direc tions in De velopment Environment and Sustainable Development Urban Flooding of Greater Dhaka in a Changing Climate Building Local Resilience to Disaster Risk Susmita Dasgupta, Asif Zaman, Subhendu Roy, Mainul Huq, Sarwar Jahan, and Ainun Nishat © 2015 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW, Washington, DC 20433 Telephone: 202-473-1000; Internet: www.worldbank.org Some rights reserved 1 2 3 4 18 17 16 15 This work is a product of the staff of The World Bank with external contributions. The findings, interpreta- tions, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) http:// creativecommons.org/licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following conditions: Attribution—Please cite the work as follows: Dasgupta, Susmita, Asif Zaman, Subhendu Roy, Mainul Huq, Sarwar Jahan, and Ainun Nishat. 2015. Urban Flooding of Greater Dhaka in a Changing Climate: Building Local Resilience to Disaster Risk. Directions in Development. Washington, DC: World Bank. doi:10.1596/978-1-4648-0710-7. License: Creative Commons Attribution CC BY 3.0 IGO Translations—If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation. Adaptations—If you create an adaptation of this work, please add the following disclaimer along with the attribution: This is an adaptation of an original work by The World Bank. Views and opinions expressed in the adaptation are the sole responsibility of the author or authors of the adaptation and are not endorsed by The World Bank. Third-party content—The World Bank does not necessarily own each component of the content ­ contained within the work. The World Bank therefore does not warrant that the use of any third-party–owned individual component or part contained in the work will not infringe on the rights ­ of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to reuse a component of the work, it is your responsibility to determine whether permission is needed for that reuse and to obtain permission from the copyright owner. Examples of components can include, but are not limited to, tables, figures, or images. All queries on rights and licenses should be addressed to the Publishing and Knowledge Division, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@ worldbank.org. ISBN (paper): 978-1-4648-0710-7 ISBN (electronic): 978-1-4648-0712-1 DOI: 10.1596/978-1-4648-0710-7 Cover photo: © Rasedul Islam / Drik. Used with the permission of Rasedul Islam / Drik. Further permission required for reuse. Cover design: Debra Naylor, Naylor Design, Inc. Library of Congress Cataloging-in-Publication data has been requested. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Contents Foreword xv Acknowledgments xvii About the Authors xix Executive Summary xxi Abbreviations xxvii Chapter 1 Introduction 1 Overview 1 Impact of Recent Major Floods 3 Waterlogging and Drainage System Capacity 5 Megacity in a Changing Climate 5 Study Purpose and Approach 7 Defining the Detailed Study Area 10 Structure of the Book 14 Notes 14 References 15 PART 1 Methods 19 Chapter 2 Hydrological Modeling 21 Introduction 21 Modeling Framework 21 Rainfall Analyses 29 Process for Modeling Future Scenarios 29 Notes 32 References 32 Chapter 3 Detailed Area Modeling 35 Introduction 35 Modeling Setup 35 Old Dhaka Model Setup 38 Central Dhaka Model Setup 44 Kallyanpur System Model Setup 50 Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7   v   vi Contents Goranchatbari System Model Setup 56 Eastern Dhaka Model Setup 62 DND Model Setup 68 Narayanganj Model Setup 75 Notes 80 References 81 Chapter 4 Vulnerability Assessment 83 Introduction 83 Key Concepts and Approach 83 Development and Application of the FVI 84 Data Collection and FVI Calculation 87 Notes 89 References 89 Chapter 5 Estimating Flood Damage 91 Introduction 91 Alternative Cases and Assumptions for Future Scenarios 91 Data to Project Future Outcomes 92 Direct and Indirect Damage 92 Key Sectors 93 Notes 99 References 100 PART 2 Results 101 Chapter 6 Assessment of Dhaka’s Modeled Zones 103 Introduction 103 Old Dhaka Results 103 Central Dhaka Results 114 Kallyanpur System Results 123 Goranchatbari System Results 133 Eastern Dhaka Results 139 DND Results 144 Narayanganj Results 153 References 157 Chapter 7 Assessment of Dhaka City’s Flood Resilience 159 Introduction 159 Summary of Modeling Findings 159 What Is Dhaka’s Flood Vulnerability? 160 Damage Estimates 164 Cost and Benefits of Adaptation 171 Scope for Further Improvement 172 Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Contents vii Notes 173 References 173 Chapter 8 Complementary Flood-Mitigation Measures 175 Introduction 175 Institutional Challenges to Stormwater Management 175 Improved Approach to Solid Waste Management 177 Conveyance System Maintenance 181 Shaving Peak Flow 184 Last Resort: Truck-Mounted Water Pumps 189 Stormwater Quality Control 189 Khals of Dhaka 193 Closing Remarks 196 Notes 197 References 197 Chapter 9 Toward an Integrated Adaptation Strategy 199 Introduction 199 The Challenge of Managing Urban Flooding 199 Climate Change Impact 200 Recommendations 200 Study Assumptions 203 Complementary Adaptation Measures 203 Toward an Integrated Strategy 204 Appendixes 205 Appendix A Individuals Consulted in Dhaka to Estimate Adaptation Costs 207 Dhaka Water Supply and Sewerage Authority 207 Institute of Water Modelling 207 Bangladesh Water Development Board 207 Other Practitioners 208 Appendix B Basin and Region Modeling and Rainfall Analyses 209 Change in Monsoon River Flows: Basin Model Output 209 Change in Sea Levels 209 Change in Land Use and Land Levels 211 Change in River Water Level: Regional Model Output 212 Change in Rainfall 212 Summary Remarks 217 Note 218 References 218 Appendix C Dhaka City Rainfall Analyses: Previous Study Findings 219 References 220 Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 viii Contents Appendix D DAP-Proposed Water Retention Areas and Water Bodies 221 Appendix E Status of 26 Khals under DWASA Jurisdiction 223 Boxes 1.1 Clarification of Key Concepts 2 1.2 Expected Increase Factor in Extreme 24-hour Precipitation for Dhaka in 2050 8 1.3 Topography of Dhaka at a Glance 13 4.1 CDRI: Assessing Adaptive Capacity at the Local Level 85 5.1 Calculating Indirect Damage from Income Loss 98 Figures 1.1 Overview of Study Approach 7 2.1 Conceptual Framework for the Modeling Exercise 22 2.2 Basin Model Calibration Plot 23 2.3 Comparison of Simulated and Measured Water Levels (m PWD) at Selected BWDB Stations, 2004 28 2.4 Process for Modeling the Detailed Study Area 30 3.1 Comparison of Simulated and Observed Water Levels at Shimrail Pump Station 74 6.1 Total Estimated Damage for Old Dhaka in 2050 108 6.2 Total Cumulative Damage for Old Dhaka, 2014–50 109 6.3 Total Estimated Damage for Central Dhaka in 2050 119 6.4 Total Cumulative Damage for Central Dhaka, 2014–50 120 6.5 Depth and Duration of Urban Flood: Comparisons for Rainfall Extremes (341 mm/24 hours) over the Study Period, Kallyanpur System Model 124 6.6 Total Estimated Damage for Kallyanpur in 2050 128 6.7 Total Cumulative Damage for Kallyanpur, 2014–50 129 6.8 Total Estimated Damage for Goranchatbari in 2050 136 6.9 Total Cumulative Damage for Goranchatbari, 2014–50 137 6.10 Flood-Depth Simulations for 30-Year Rainfall Event in Eastern Dhaka, Showing Effects of Planned Improvements 141 6.11 Total Estimated Damage for Eastern Dhaka in 2050 143 6.12 Total Cumulative Damage for Eastern Dhaka, 2014–50 143 6.13 Total Estimated Damage for DND in 2050 149 6.14 Total Cumulative Damage for DND, 2014–50 150 7.1 Impact of Climate Change on Flooded Area in 2050 160 7.2 Total Estimated Damage for Dhaka in 2050 168 7.3 Total Cumulative Damage for Dhaka, 2014–50 170 7.4 Cumulative Climate Change Damage for Total Dhaka, Using PPP and Man-Days 171 Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Contents ix B.1 Average Monthly Flow of Brahmaputra (Jamuna) River at Chilmari 210 B.2 Change in Monsoon Discharge of Brahmaputra (Jamuna) River at Chilmari 210 B.3 Frequency Distribution of Land Levels for Greater Dhaka 211 B.4 Changes in Peak River Water Levels around Dhaka City Due to Climate Change 213 B.5 Daily Rainfall Time Series at BWDB Dhaka Station 214 B.6 Change in Return Periods of Daily Rainfall Events at BWDB Dhaka Station 214 B.7 Rainfall Data at BMD Dhaka Station, 2003–09 215 B.8 Three-Hourly Rainfall for September 2004 Event at BMD Dhaka Station 216 B.9 Hourly Fractions of Six-Hour Design Storm Event 217 Maps 1.1 Detailed Study Area, Showing Seven Drainage Models 11 2.1 Regional Rivers Surrounding Dhaka City 25 2.2 River System and Hydrological Stations of the North Central Region Hydrodynamic Model 26 3.1 Model Boundaries of Detailed Study Area 36 3.2 Physical Setting of Old Dhaka Model with Adjoining Boundaries 39 3.3 Change in Land Elevation over the Study Period, Old Dhaka Model 40 3.4 Changes in Proposed Land Use over the Study Period, Old Dhaka Model 40 3.5 Planned Drainage System Improvements for Old Dhaka 41 3.6 Physical Setting of Central Dhaka Model with Adjoining Boundaries 45 3.7 Change in Land Elevation over the Study Period, Central Dhaka Model 46 3.8 Changes in Proposed Land Use over the Study Period, Central Dhaka Model 47 3.9 Planned Drainage System Improvements, Central Dhaka 48 3.10 Physical Setting of Kallyanpur System Model with Adjoining Boundaries 51 3.11 Change in Land Elevation over the Study Period, Kallyanpur System Model 52 3.12 Changes in Proposed Land Use over the Study Period, Kallyanpur System Model 53 3.13 Kallyanpur System Model Drainage Network and Subcatchments, 2004 54 3.14 Physical Setting of Goranchatbari System Model with Adjoining Boundaries 57 Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 x Contents 3.15 Change in Land Elevation over the Study Period, Goranchatbari System Model 58 3.16 Changes in Proposed Land Use over the Study Period, Goranchatbari System Model 59 3.17 Goranchatbari System Model Drainage Network and Subcatchments, 2004 60 3.18 Physical Setting of Eastern Dhaka Model with Adjoining Boundaries 63 3.19 Change in Land Elevation over the Study Period, Eastern Dhaka Model 64 3.20 Changes in Proposed Land Use over the Study Period, Eastern Dhaka Model 65 3.21 Eastern Dhaka Drainage Network and Subcatchments, 2004 66 3.22 Physical Setting of DND Model with Adjoining Boundaries 69 3.23 Change in Land Elevation over the Study Period, DND Model 70 3.24 Changes in Proposed Land Use over the Study Period, DND Model 71 3.25 DND Drainage Network and Subcatchments, 2004 72 3.26 Physical Setting of Narayanganj Model with Adjoining Boundaries 75 3.27 Change in Land Elevation over the Study Period, Narayanganj Model 76 3.28 Changes in Proposed Land Use over the Study Period, Narayanganj Model 77 3.29 Drainage System of Narayanganj, 2004 79 6.1 Extreme Rainfall Scenarios in 2050 with 0.25 m Flood Criteria, Old Dhaka 104 6.2 Extreme Rainfall Scenarios in 2050 with 0.1 m Flood Criteria, Central Dhaka 115 6.3 Extreme Rainfall Scenarios in 2050 with 0.25 m Flood Criteria, Kallyanpur System Model 125 6.4 Changes in Extent of Flooding for Extreme Rainfall Event over Study Period, Goranchatbari System Model 133 6.5 Flood-Depth Comparisons of Extreme Rainfall Scenarios in 2050, Eastern Dhaka Model 140 6.6 Flood-Depth Comparisons of Extreme Rainfall Events in 2050, DND Model 145 6.7 Extreme Rainfall Event (341 mm/24 hours) in 2050 with 0.25 m Flood Criteria, Narayanganj Model 155 6.8 Flood Depth of A1FI Rainfall Event in 2050, with Additional Drainage Improvements, Narayanganj Model 156 8.1 Water Bodies of Dhaka 188 8.2 Flood Maps for Extreme Event in 2050 with Planned Improvements and Khal Bank Encroachments 195 Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Contents xi Photos 3.1 Site investigation to measure sediment depth in box-culvert 42 3.2 Drainage pumps at Dholaikhal Pump Station, Old Dhaka 43 3.3 Pumping floodwater out of Hathir Jheel using temporary pumps, Central Dhaka 49 3.4 Kallyanpur Retention Pond in the dry season 53 3.5 Outlet of temporary drainage pumps on the Western Embankment near Rayerbazar, Kallyanpur system 55 3.6 Water inlet after garbage filtering at Goranchatbari Pump Station 61 3.7 Tin-shed house built on stilts in low-lying area of Khilgaon, Eastern Dhaka 62 3.8 Intake pipes of drainage pumps at Shimrail Pump Station, DND system 74 3.9 Drainage outfalls to Lakhya River in Narayanganj township at end of RK Mitra Road, eastern area and end of BK Road, southeastern area 78 6.1 Original Rampura sluice gates and new bypass channel used when gates are closed (Central Dhaka system) 114 6.2 Construction of new pump station at Hatir Jheel, Rampura (Central Dhaka system) 116 6.3 Laying new stormwater pipe in Gulshan area of Central Dhaka system 120 6.4 Resectioning and lining work at Katashur Khal near Mohammadpur (Kallyanpur system) 124 6.5 Temporary pumps located along the Western Embankment, Kallyanpur 125 8.1 DWASA sign next to Segunbagicha Khal warning against encroachment, connecting wastewater line to the khal, and garbage dumping 178 8.2 Motorized solid-waste collector belt at Hatir Jheel outlet of Panthapath Box-Culvert near Sonargaon Hotel (Central Dhaka system) 182 8.3 Hazaribagh Khal outlet (looking toward river) on the Western Flood Embankment near Rayerbazar 190 8.4 Low-lying area of Kallyanpur, showing floodwater mixed with wastewater 191 8.5 Discharge from Kallyanpur Retention Pond into Turag River 193 8.6 Segunbagicha Khal, Central Dhaka system 194 Tables 2.1 Relationship between Population Density and Land Levels 27 2.2 Model Simulations for the Detailed Study Area 31 Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 xii Contents 3.1 Summary of Data Requirements and Availability 36 3.2 Typical Parameters and Values for MIKE Urban Model A 37 3.3 Impervious Percentage for Various Land Uses 38 3.4 Development Baseline: Key Assumptions Comparison between 2004 and 2050, Old Dhaka Model 43 3.5 Development Baseline: Key Assumptions Comparison between 2004 and 2050, Central Dhaka Model 49 3.6 Development Baseline: Key Assumptions Comparison between 2004 and 2050, Kallyanpur System Model 55 3.7 Development Baseline: Key Assumptions Comparison between 2004 and 2050, Goranchatbari System Model 61 3.8 Development Baseline: Key Assumptions Comparison between 2004 and 2050, Eastern Dhaka Model 67 3.9 Runoff Contribution into Khals, Eastern Dhaka Model 68 3.10 Development Baseline: Key Assumptions Comparison between 2004 and 2050, DND Model 73 3.11 Development Baseline: Key Assumptions Comparison between 2004 and 2050, Narayanganj Model 80 4.1 Indicators Used to Develop the Flood Vulnerability Index 86 5.1 Residential Housing Data to Estimate Direct Damage 94 5.2 Household Income Distribution and Affected Property 96 6.1 Comparison of Flooding Results with 0.1 m Flood Criteria, Old Dhaka Model 104 6.2 Comparison of 2050 Flooding Results with 0.25 m Flood Criteria, Old Dhaka Model 105 6.3 2050 Rainfall Event in a Changing Climate, Old Dhaka 106 6.4 FVI Rankings of Old Dhaka Wards under Alternative Flooding Scenarios 107 6.5 Increased Damage in Old Dhaka from Climate Change for Future Scenarios under Alternative Assumptions 110 6.6 Unit Costs for Temporary Pumps, Old Dhaka 110 6.7 Estimated Cost of Meeting the Current Adaptation Deficit, Old Dhaka 111 6.8 Estimated Cost of Meeting the Climate Change Adaptation Deficit, Old Dhaka 112 6.9 Summary of Adaptation Costs, Old Dhaka 113 6.10 2050 Rainfall Event in a Changing Climate, Central Dhaka 115 6.11 FVI Rankings of Central Dhaka Wards under Alternative Flooding Scenarios 117 6.12 Increased Damage in Central Dhaka from Climate Change for Future Scenarios under Alternative Assumptions 121 6.13 Estimated Cost of Meeting the Current Adaptation Deficit, Central Dhaka 122 6.14 Estimated Cost of Meeting the Climate Change Adaptation Deficit, Central Dhaka 122 Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Contents xiii 6.15 Summary of Adaptation Costs, Central Dhaka 123 6.16 2050 Rainfall Event in a Changing Climate, Kallyanpur System Model 126 6.17 FVI Rankings of Kallyanpur Wards under Alternative Flooding Scenarios 127 6.18 Increased Damage in Kallyanpur from Climate Change for Future Scenarios under Alternative Assumptions 130 6.19 Estimated Cost of Meeting the Current Adaptation Deficit, Kallyanpur 131 6.20 Estimated Cost of Meeting the Climate Change Adaptation Deficit, Kallyanpur 132 6.21 Summary of Adaptation Costs, Kallyanpur 132 6.22 Comparison of Flooding Results, Goranchatbari System Model 134 6.23 2050 Rainfall Event in a Changing Climate, Goranchatbari System Model 135 6.24 FVI Rankings of Goranchatbari Wards/Region under Alternative Flooding Scenarios 135 6.25 Increased Damage in Goranchatbari from Climate Change for Future Scenarios under Alternative Assumptions 138 6.26 Estimated Cost of Meeting the Adaptation Deficit, Goranchatbari 138 6.27 Summary of Adaptation Costs, Goranchatbari 139 6.28 Comparison of Flooding Results, Eastern Dhaka Model 139 6.29 2050 Rainfall Event in a Changing Climate, Eastern Dhaka Model 140 6.30 FVI Rankings of Eastern Dhaka Regions under Alternative Flooding Scenarios 142 6.31 Increased Damage in Eastern Dhaka from Climate Change for Future Scenarios under Alternative Assumptions 144 6.32 Flood-Depth Comparisons, DND Model 145 6.33 2050 Rainfall Event in a Changing Climate, DND Model Area 146 6.34 2050 Flood-Depth Comparisons in a Changing Climate, Highlighting the Impact of Khal Encroachment 147 6.35 FVI Rankings of DND Wards/Region under Alternative Flooding Scenarios 148 6.36 Increased Damage in DND from Climate Change for Future Scenarios under Alternative Assumptions 150 6.37 Estimated Cost of Meeting the Current Adaptation Deficit, DND 151 6.38 Estimated Cost of Meeting the Climate Change Adaptation Deficit, DND 152 6.39 Summary of Adaptation Costs, DND 152 6.40 Comparison of 2004 Flooding Results, Narayanganj Model 153 6.41 Comparison of 2050 Flooding Results, Narayanganj Model 154 Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 xiv Contents 6.42 2050 Rainfall Events in a Changing Climate, Narayanganj Model Area 154 7.1 FVI Rankings of Dhaka Study Areas 161 7.2 FVI Rankings of Dhaka Wards and Regions 162 7.3 Main Drivers of Greater Vulnerability for Dhaka’s Top 20 Vulnerable Wards/Regions 163 7.4 Summary of Sector-Wise Damage in Dhaka under Alternative Scenarios 168 7.5 Damage Differences from Climate Change for Future Scenarios Using Alternative Probabilities 169 7.6 Differences in Cumulative Damage (2014–50) from Climate Change for Future Scenarios Using Alternative Probabilities 170 7.7 Total Estimated Cost of Adaptation for Detailed Study Area 172 8.1 Length of Drainage System Infrastructure and Estimated Cleaning Cost by System 182 8.2 Values of Runoff Coefficient for Various Catchments 186 D.1 DAP-Proposed Water Retention Areas and Water Bodies 221 E.1 Status of 26 Khals under DWASA Jurisdiction 223 Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Foreword Urban flooding from intense rainfall is a recurring phenomenon that adversely affects life and livelihoods in Dhaka, the capital and largest city of Bangladesh. Rapid and indiscriminate urbanization has aggravated the problem, contributing to the gradual filling up of low-lying flood plains, rivers, canals, and other water bodies that used to drain or retain water during heavy rainfall events. The floods of 1998, 2004, and 2007 battered the city, affecting communications, livelihoods, and service facilities for many days. The fringe areas, slums, and shanties—where many of the city’s poor reside—suffered more. In the coming decades, the effects of climate change may further aggravate Dhaka’s flood vulnerability. Projections by the Intergovernmental Panel on Climate Change (IPCC) and the World Meteorological Organization suggest more erratic monsoon rainfall in the Ganges, Brahmaputra, and Meghna Basin and an increase in the frequency and intensity of climate extremes in the 21st century. Adapting to the increased risks of severe rainfall events, along with the other climate change impacts, is essential for the development of Greater Dhaka. Adaptation will require climate-smart policies and investments to make Greater Dhaka more resilient to the effects of climate change, including loss of property, habitat, and infrastructure. To formulate effective adaptation, the resi- dents of Greater Dhaka and the Government of Bangladesh, as well as develop- ment partners, need a better understanding of the potential damage related to climate change and the cost of adapting to extreme rainfall events. The World Bank, the Institute of Water Modelling, Development Policy Group, Bangladesh University of Engineering and Technology (BUET), and BRAC University, in consultation with the Dhaka Water Supply and Sewerage Authority, have conducted a study on the potential intensification of urban flooding in the Greater Dhaka area in a changing climate. The study sheds light on the potential damage from extreme weather events and adaptation costs. Earlier analytical work on climate-proofing Dhaka’s infrastructure has been confined mainly to case studies, with relatively limited sets of locations, impacts, and adaptation measures. This study fills that critical knowledge gap by providing itemized estimates of the incremental costs of infrastructure adaptation out to the year 2050. The study is of prime importance, as it identifies vulnerable popu- lations and infrastructure, quantifies outstanding deficits in addressing current Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7   xv   xvi Foreword climate-related risks, and estimates the adaptation cost of avoiding further damage due to climate change. It is my hope that these estimates will serve as ­ a pragmatic tool for Bangladesh’s decision-makers to develop location-specific adaptation plans for the Greater Dhaka area. Johannes C. M. Zutt Bangladesh Country Director The World Bank Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Acknowledgments We would like to express our sincere appreciation to the many people whose dedicated work contributed to the creation of this book. We are particularly grateful to Masahiro Sugiyama, assistant professor at the University of Tokyo, Policy Alternatives Research Institute, who provided us critical data on the expected increase in the extreme 24-hour precipitation increase factor for Dhaka in 2050. We appreciate the support provided by experts at the Institute of Water Modelling, Water Resources Planning division, throughout various stages of the hydrological modeling study. We would like to mention S. R. Prodhan, water sector specialist; David M. Khan, senior specialist; Imtiaz A. Taher, drainage ­ modeling expert; Ismat A. Parvin, associate specialist; Md. Atikul Islam, junior ­ specialist; and junior engineers Md. Feroz Islam, Nabila F. Tasneem, Sadbir Rahman, Johor K. Munni, and Shadman S. Protik. The book benefited from the valuable suggestions and insights provided by numerous experts in Bangladesh. At the Bangladesh University of Engineering and Technology (BUET), Department of Urban and Regional Planning, we would like to note the contributions of Shakil Akther, professor and head; Ishrat Islam, professor (and director, Center for Regional Development Studies); and Gulshan Ara Parvin, former associate professor (and former post-doctoral fellow, Graduate School of Global Environmental Studies, Kyoto University). At the Bangladesh Water Development Board (BWDB), we would like to thank Choudhury Nazmul Alam, superintending engineer; Amirul Hossain, executive engineer; and K. M. Nurul Islam, superintending engineer. At the Dhaka Water Supply and Sewerage Authority (DWASA), we would like to mention Kyaw Sha Ching, superintending engineer and project director and A. K. M. Shahid Uddin, superintending engineer. At Dhaka University, Department of Geography and Environment, we would like to thank professors A. Q. M. Mahbub and Nurul Islam Nazem. We also extend our thanks to Akhter H. Choudhury, professor, Khulna University, Urban and Rural Planning Discipline division; Md. Moinul Islam, town planner, Narayanganj City Corporation; Md. Taibur Rahman, senior assistant chief, Bangladesh Planning Commission; and Khurshid Zabin Hossain Taufique, director, town planning, RAJUK. In addition to those mentioned above, we extend our thanks to the experts who participated in the study’s three stakeholder conferences convened at the Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7   xvii   xviii Acknowledgments World Bank Dhaka Office. We would also like to thank the participants of the study’s two Focus Group Discussions, held at the World Bank Dhaka Office and BRAC University. At the World Bank, we are grateful for the contributions of Johannes C. M. Zutt, Bangladesh country director; Ellen A. Goldstein, former Bangladesh coun- try director; Michael Toman, research manager, DECEE; Kseniya Lvovsky, prac- tice manager, GENDR; Herbert Acquay, former practice manager, GENDR; Bernice K. Van Bronkhorst, practice manager, GSURR; Maria Sarraf, lead envi- ronmental specialist, GENDR; Lia C. Sieghart, program leader, SACBN; Mehrin A. Mahbub, communications officer, SAREC; Cecilia Belita, senior program assistant, GWADR; Sayeeda S. Tauhid, former senior monitoring and evaluation specialist; and Omar Farook, former operations analyst. Finally, we gratefully acknowledge the financial support provided by the Bangladesh Climate Change Resilience Fund (BCCRF). In April 2012, the BCCRF Management Committee endorsed this study, requested by the Government of Bangladesh, as a priority research project. The BCCRF is a multi- donor trust fund supported by the governments of the United Kingdom, Denmark, Sweden, Switzerland, Australia, the United States, and the European Union. Note of Appreciation As it often happens, there are people without whose dedicated work the cre- ation of this book would not have been possible. We would like foremost to ­mention S. M. Mahabubur Rahman, director, Water Resources Planning Division, Institute of Water Modelling, Bangladesh, who played a vital role in designing the study and ensuring the quality of the hydrological modeling, thus providing a solid foundation on which the other components of the analytic work were developed. We also express our sincere appreciation to Kazi Habib Ullah, ­executive engineer, Dhaka Water Supply and Sewerage Authority, and Md. Liakath Ali, climate change and environmental adviser, UK Department for International Development, Bangladesh, for providing technical guidance throughout the study. At the World Bank, our sincere gratitude goes to Norma Adams, consultant editor, whose linguistic skills and creativity enabled us to produce this book from its underlying multidisciplinary technical reports. Our special thanks go to Polly Means, who worked tirelessly to prepare the graphics presented in this book. We also extend special thanks to Shahpar Selim, environmental specialist, for her support at various stages of the study. As a native Bangladeshi, she had knowledge of Dhaka that was essential to building multisectoral cooperation and participatory process for implementing the study. We are especially grateful the ­ to Angie Harney, program assistant, SACBD, for meticulously planning the study workshops and Focus Group Discussions. Finally, special thanks go to B. Elaine Wylie, program assistant, DECEE, and Angie Harney for their support with contract-related logistics. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 About the Authors Susmita Dasgupta (PhD, State University of New York, USA, 1992) is a lead environmental economist in the Development Research Group of the World Bank, specializing in environmental management in developing countries. She has conducted extensive analysis on the health hazards of pollution, poverty/ environment nexus, priority-setting in pollution control, deforestation, biodiver- sity loss, impacts of climate change on coastal zones and climate extremes, climate change adaptation, cost-effective regulations, and monitoring and ­ enforcement of regulations. Her research activities have covered Bangladesh, Brazil, Cambodia, China, Colombia, Cuba, India, the Islamic Republic of Iran, the Lao People’s Democratic Republic, Madagascar, Mexico, Tunisia, Vietnam, and the Republic of Yemen. She has published numerous articles on issues related to development and the environment. Asif Zaman (PhD, University of Melbourne, Australia, 2006) is a water resources consultant with more than 15 years of international work experience in water- related projects, including water-resources planning and management, irrigation and drainage modeling, river-basin modeling, and general hydrologic modeling. He has worked on climate change impact and adaptation projects and environ- mental risk assessments and has prepared training guidelines and manuals. He has written more than 20 technical and scientific publications. His research interests include integrated water resources management, climate change adaptation, and improving communication between technical and non-technical disciplines. Over the past six years, his work has focused mainly on various projects at the Water Resources Planning Division, Institute of Water Modelling. Subhendu Roy (PhD, Boston University, USA, 1997) is currently a professor of economics at Massasoit College in Massachusetts and consultant in envi­ ronmental and health economics. For 15 years, he worked for the Indian Administrative Service, where he focused on developmental issues. He spent the subsequent 14 years in the telecommunications industry, specializing in economic regulations. He has written numerous publications on issues related ­ to environment, health, and economic regulations, with a particular focus on policy aspects. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7   xix   xx About the Authors Mainul Huq (MA, Vanderbilt University, Nashville, Tennessee, USA, 1990) is an economist and consultant on numerous World Bank–financed research projects, most of which focus on environmental issues, particularly climate change and cost-effective adaptation policies, health impact of pollution problems, and mon- etization of damage caused by environmental degradation. He has also worked on projects financed by the Asian Development Bank and World Health Organization. His additional research interests include the economic impact of large-scale infrastructure projects and urban and municipal issues. He has pub- lished numerous articles in international journals on issues related to environ- ment, health, environmental damage, and pollution-control policy. Sarwar Jahan (PhD, University of Illinois at Urbana–Champaign, USA, 1987) is a professor in the Department of Urban and Regional Planning at the Bangladesh University of Engineering and Technology (BUET), Dhaka. His research inter- ests include spatial modeling, urban and regional analysis and development, adaptation to climate change, and disaster management. As a consultant, he has worked on a variety of projects funded by the World Bank, Asian Development Bank, and United Nations Development Program, among other international agencies. He has also provided consultancy and advisory services to various gov- ernment departments. He has been published widely in national and interna- tional research journals, as well as in conference proceedings focused on urban and regional issues. Ainun Nishat (PhD, University of Strathclyde, Glasgow, UK, 1981) is professor emeritus at the Centre for Climate Change and Environmental Research, BRAC University, Dhaka. For decades, he has worked as a catalyst, advocate, educator, and facilitator, championing the wise use of natural resources and sustainable development in Bangladesh. His current research focuses on adaptation to cli- mate change and related policy advocacy. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Executive Summary Dhaka, the capital of Bangladesh and one of the world’s rapidly growing megacities, is an urban hotspot for climate risks. Nearly every year, the city experiences urban flooding and waterlogging following intense rainfall during the rainy season. The city’s poorer inhabitants are usually among the most vulnerable since large, densely populated conglomerations of slums and shanties are located in areas of unplanned and unregulated development. A major concern is that, in a changing climate, Dhaka’s flooding situation may worsen dramatically into disasters. Projections by the Intergovernmental Panel on Climate Change (IPCC) suggest heavier, more erratic rainfall in the Ganges, Brahmaputra, and Meghna (GBM) Basin during the monsoon season. Also, warmer ocean-surface temperatures and rising sea levels will likely mean more intense cyclones in the future. Thus, adaptation to the challenges posed by climate change in this rapidly growing megacity is critical. Against this backdrop, the World Bank, in collaboration with experts from Bangladesh, has undertaken analytical work to provide Dhaka’s local planners an effective approach to mitigate the damage risk of rainfall-induced urban flooding in a changing climate. Funded by the Bangladesh Climate Change Resilience Fund (BCCRF), this study was launched in 2012. Its specific objectives were to evalu- ate the Greater Dhaka area’s local capacity to address urban flooding in the cur- rent climate, assess the impacts of climate change on location-specific inundation for various scenarios, and identify adaptation measures for each of the city’s areas to address the threat of climate change–related impacts. The general physical setting for the study is a 1,528 km2 area in north-central Bangladesh that corresponds to RAJUK’s (Rajdhani Unnayan Kartripakkha—the Capital Development Authority of Bangladesh) Dhaka Metropolitan Development Plan (DMDP). It consists of the Dhaka metropolitan area, the Dhaka- Narayanganj-Demra (DND) area, and the townships of Narayanganj, Savar, Gazipur, and Tongi. For the latter three townships, detailed data on drainage infrastructure, which are required for location-specific mapping of inundation depth and duration, were lacking. Therefore, the detailed study area of 345 km2 was limited to the Dhaka metropolitan area, the DND area, and Narayanganj township. The Dhaka metropolitan area consists of two parts: Eastern Dhaka (121 km2) and Western Dhaka (143 km2). Western Dhaka, in turn, is divided into four areas: Old Dhaka, Central Dhaka, Kallyanpur, and Goranchatbari. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7   xxi   xxii Executive Summary The study comprises four main activities: (a) hydrological modeling and develop- ment of adaptation measures, (b) spatial ranking of flood vulnerability, (c) evaluation of expected damage from flooding, and (d) estimation of adaptation costs. The overall objectives and scope of the hydrological modeling exercise were to assess climate change impacts on location-specific waterlogging for various scenarios, focusing on the inundation depth and duration caused by extreme precipitation events for the standard IPCC AR4 high-emissions, fossil-fuel intensive (A1FI) sce- nario in 2050. The baseline (without climate change scenario) selected for the modeling exercise was the September 2004 extreme rainfall event (341 mm in 24 hours), which caused one of the heaviest urban flooding damages in recent history. The study’s hydrological modeling component adopted a three-step approach to account for the effects of river flows on flooding in the Greater Dhaka area. Because Dhaka is located on the lower reaches of the Ganges-Brahmaputra Delta and surrounded by six rivers and canals, flooding in the study area is also affected by flows in the Brahmaputra (Jamuna) River and its distributaries in north-central Bangladesh. Thus, the first modeling step simulated basin-level flows from the Brahmaputra River in the GBM system since they strongly influence monsoon- season river flows and water levels in the Greater Dhaka area. In the second step, the effects of climate change on the regional rivers were simulated. Finally, in the third step, detailed modeling of the drainage system in and around Dhaka city was simulated. Although river levels affect the drainage ­ system within the city area, this study centered on the worst-case scenario in which river levels are high and all sluice gates are closed; that is, there is no grav- ity drainage out of the city, and the drainage system depends primarily on the performance efficiency of drainage pumps, which is often the case during intense rainfall. For Western Dhaka—Old Dhaka, Central Dhaka, Kallyanpur, and Goranchatbari—the impacts of climate change on river flooding were not con- sidered in the analyses as this main part of the city is protected by flood embank- ments on all sides. In order to conduct the hydrological modeling, local engineers were consulted to understand the functioning of each existing drainage system in the detailed study area. In accordance with recommendations received from the experts, all future scenarios assumed that all improvements to Dhaka’s drainage infrastructure— both planned and proposed by the relevant drainage-system authorities—would be implemented. These included improvements covered in RAJUK’s Detailed Area Plan (DAP), the Sewerage Master Plan of the Dhaka Water Supply and Sewerage Authority (DWASA), and Narayanganj City Corporation’s Concept Vision Plan. Previous recommendations on improving drainage-system perfor- mance were included in the modeling of future scenarios for Eastern Dhaka, Central Dhaka, the DND area, Kallyanpur, and Narayanganj. Modeling assump- tions also included recommendations of Jurutera Perunding Zaaba Sdn. Bhd. (Malaysia), Farhat Consulting Engineers and Architects Limited, and SARM Associates Limited; the Institute of Water Modelling (IWM); and the Institute of Water and Flood Management (IWFM). Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Executive Summary xxiii Based on the hydrological modeling simulations, structural measures beyond cur- rent planned improvements were developed to address the current and climate change adaptation deficits. In order to identify interventions to close the existing gap in drainage capacity, flood modeling simulations were repeated, using alternative pump capacities and diameters of drainage pipes, until the resulting flood maps met pre-set acceptable flooding criteria for inundation depth and duration. These criteria, agreed on by the local experts and other stakeholders, were based on the level of standing water that the local people in each of the modeled zones have learned to cope with over the years as a usual occurrence during the flood season (between 4 and 8 inches). To provide a more complete picture of how the various areas of Dhaka may be affected, the study conducted a detailed vulnerability impact assessment, taking into account all non-flood factors that, when combined with flooding, may exacerbate the impact. Inundation of even the same depth and duration can have significantly different impacts on the population and infrastructure, depending on an affected area’s physical, social, economic, and institutional characteristics. Drawing on the hydrological modeling data, along with information generated from a climate disaster resilience index (CDRI) analysis for Dhaka conducted for this assess- ment, the study developed a flood vulnerability index (FVI) that ranked the population and infrastructure of the city’s major areas and wards, according to their relative exposure, susceptibility, and resilience. It is expected that the result- ing ward-level analysis, which identified Dhaka’s 20 most vulnerable wards/ regions, will provide local policymakers a useful roadmap for future planning to minimize the effects of climate change. Since flooding in an urban area like Dhaka can lead to widespread damage result- ing from submersion of buildings and property, as well as livelihood disruption among the flood-affected population, the study assessed the extent of flood damage for the major affected sectors of the economy, as well as overall potential economic damage. Given the uncertainty surrounding the magnitude and timing of climate change risk, this study considered the following alternative cases. Based on alternative future scenarios, it evaluated the direct and indirect economic damage to the local population from a severe (100-year, return-period) event occurring in 2050; a probability-weighted damage of different return-period rainfall events occur- ring in 2050; and an estimate of total cumulative damage between 2014 and 2050, using random assignment of 1-year to 100-year storms for each year. Results of the hydrological modeling exercise were used to quantify the economic damage—effects on the urban-built environment and infrastructure, human health, and family assets—caused by flooding of the Greater Dhaka detailed study area. The likely increase in damage due to climate change was also separately assessed. Both depth and duration of inundation were analyzed to estimate the extent of flood damage. The study not only assessed readily quantifiable damage; wherever possible, it assigned monetary values to intangible damages, based on analysis of losses gleaned from comparison with similar ones that are more easily quantified. Sector-wise cumulative damage was also examined using random assignment of various return-period, extreme rainfall events. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 xxiv Executive Summary Finally, the adaptation costs of the proposed structural measures were estimated for each modeled zone. These cost estimates were divided into two components: (a) additional investments required to meet the current adaptation deficit with- out climate change and (b) further measures required to meet the 2050 climate change deficit. To estimate the costs of the various recommended measures, a number of assumptions were made, some of which were supplied by DWASA officials. Numerous practitioners were consulted, including officials and engi- neers of DWASA, the Bangladesh Water Development Board (BWDB), the Public Works Department (PWD), IWM, World Bank, and Dhaka North and South City Corporations (DNCC and DSCC), as well as contractors, suppliers, businessmen, and independent consultants. In addition, several field visits were undertaken to physically inspect the situation around Dhaka city. The study also explored non-structural flood-mitigation measures and how key institutional, technical, and financial challenges to their implementation can be effec- tively addressed. The recommended conveyance-centric structural approach to flood mitigation—featuring mainly pipes, khals (canals), and pumps—appears appropriate, given Dhaka’s land scarcity, growing population density, and other prevailing socioeconomic, environmental, and geophysical conditions. However, implementing these structural adaptation measures in isolation cannot be expected to eliminate the flood-damage risk entirely. Therefore, the study explored an array of urgently needed non-structural flood-mitigation measures, including an improved approach to solid waste management and maintenance of the conveyance system, which are designed to complement the structural mea- sures. In addition, it considered alternative ways to shave peak stormwater flow, including delaying, diverting, or detaining runoff. The study’s key findings highlight the importance of implementing the additional investments recommended by the hydrological modeling exercise to close the current adaptation deficit and secure the significant damage savings that would result. The cost of meeting Dhaka’s current adaptation deficit, even without climate change, would total Tk. 2.7 billion, equivalent to just 0.35 percent of the government’s annual development budget expenditure for 2014–15. Of this amount, Central Dhaka would comprise the largest investment, at about Tk. 1.4 billion. The added cost of closing the climate change gap would require the other Tk. 1.3 billion. Implementing the recommended additional investments can result in significant damage savings for Dhaka, given that the expected damage from flooding would be quite significant for the city overall. For example, if an extreme rainfall event like that of 2004 were to occur in 2050, then, without investment to address the current adaptation deficit, the increased damage caused by climate change would amount to Tk. 2.0 billion; however, it would be reduced signifi- cantly (to Tk. 0.9 billion) by investing to close the current adaptation gap. Such savings in damage of Tk. 1.1 billion in just one year reveal how quickly the invest- ment of Tk. 2.7 billion in current adaptation deficit can be paid back. The estimates presented in this study should be considered conservative. Climate science predicts that extreme rainfall events will become more frequent over Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Executive Summary xxv time, suggesting that Dhaka may experience a number of intense rainfall events between now and 2050. Thus, the expected damage from a single storm (which is a proxy for the benefit of adaptation), as presented above, must be considered conservative. In the absence of information on the current condition of all drain- age pipes, this study’s hydrological modeling exercise has assumed that pipes, khals, and box-culverts will perform according to their designated capacity and that stormwater will reach them without undue difficulty. But given Dhaka’s current reality, these assumptions are somewhat optimistic. Accounting for the added expenditures needed to achieve a flood-free Dhaka—including urgently needed non-conveyance flood mitigation measures—would raise the total cost of adaptation substantially. That said, this analytical work will assist Bangladesh’s policymakers to develop effective adaptation to urban flooding in Dhaka and improve the city’s capacities against the impacts of climate change and variability. Equipped with a menu of investment options (the implementation of which would require project-­ specific, cost-benefit analysis) designed to close the current adaptation deficit and further climate-proof urban infrastructure, local decision-makers will be able to formu- late more effective strategies, prioritize interventions, and sequence activities as resources permit. In light of the uncertainties about the magnitude and timing of climate change–related risks, it may be wise to start by addressing the current adaptation deficit that the city’s residents are already facing. Closing the current adaptation gap will provide policymakers a solid foundation on which to develop additional measures to mitigate climate risks. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Abbreviations AD adaptation (to climate change) AI additional investments A1FI IPPC scenario with future based on fossil fuel–intensive development BBS Bangladesh Bureau of Statistics BCCRF Bangladesh Climate Change Resilience Fund BMD Bangladesh Meteorological Department BSCIC Bangladesh Small and Cottage Industries Corporation BUET Bangladesh University of Engineering and Technology BWDB Bangladesh Water Development Board B1 IPPC scenario with future based on a high level of environ- mental and social consciousness combined with a globally coherent approach to a more sustainable development CBO community-based organization CC climate change CCA climate change adaptation CDRI Climate Disaster Resilience Index CER certified emission reduction DALY Disability-Adjusted Life Year DAP Detailed Area Plan DCC Dhaka City Corporation DEM Digital Elevation Model DGHS Directorate General of Health Services DMDP Dhaka Metropolitan Development Plan DNCC Dhaka North City Corporation DND Dhaka-Narayanganj-Demra DRR disaster risk reduction DSCC Dhaka South City Corporation DWASA Dhaka Water Supply and Sewerage Authority EWS economically weaker sections Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7   xxvii   xxviii Abbreviations FGD Focus Group Discussion FVI Flood Vulnerability Index GBM Ganges-Brahmaputra-Meghna (river system) GCM General Circulation Model GHG greenhouse gas GIS Geographic Information System HIG high-income groups IPCC Intergovernmental Panel on Climate Change IUCN International Union for Conservation of Nature IWFM Institute of Water and Flood Management IWM Institute of Water Modelling JICA Japan International Cooperation Agency LGED Local Government Engineering Department MIG middle-income groups NAM Nedbør-afstrømnings (Rainfall-Runoff) Model NCC Narayanganj City Corporation NCRHD North Central Region Hydrodynamic (model) NGO nongovernmental organization NIMBY “not in my back yard” PI planned improvements PWD Public Works Department RAJUK Rajdhani Unnayan Kartripakkha (Capital Development Authority of Bangladesh) SREX Special Report on Managing the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation (by IPCC) UNESCO United Nations Educational, Scientific and Cultural Organization Units of Measure cm centimeter ft3/s cubic foot per second ha hectare K° degree Kelvin km kilometer km2 square kilometer kVA kilovolt-ampere m meter Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Abbreviations xxix m3 cubic meter 3 m /s cubic meter per second mm millimeter Currency Equivalents currency unit = Bangladesh Taka (Tk.) Tk. 78 = US$1 Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Chapter 1 Introduction Overview Dhaka, the capital of Bangladesh and one of the world’s rapidly growing megacities, is an urban hotspot for climate risks. Located in central Bangladesh ­ on the lower reaches of the Ganges-Brahmaputra Delta, the city faces the recur- ring phenomena of urban flooding and waterlogging following intense rainfall nearly every year. As a low-elevation city with a tropical monsoon climate, Dhaka has a long history of river flooding as a natural hazard. Recent major floods have been worse in terms of depth and extent of inundation and duration, especially in fringe areas, where many of the city’s poor reside. Rapid, unplanned urbaniza- tion and the gradual filling up of low-lying flood plains, rivers, canals, and other water bodies traditionally used to drain or retain water during rainfall have exac- erbated the problem. A growing concern is that, in a changing climate, character- ized by heavier and more erratic rainfall in the Ganges-Brahmaputra-Meghna (GBM) Basin during the monsoon season, the situation may worsen. Dhaka’s average annual rainfall is about 2,000 mm, most of it occurring dur- ing the summer monsoon. The rivers surrounding the city are distributaries of the Brahmaputra River. Thus, the water level of the Brahmaputra and rainfall and runoff from the entire Brahmaputra Basin influence the process of river floods in and around the city. The entire drainage system is also influenced by water levels of the Ganges and Meghna rivers. At present, the western part of Dhaka is pro- tected against river floods that formerly resulted from overflow of the Turag, Buriganga, and Balu rivers. However, the western part of the city is now vulner- able to urban flooding resulting from the combined effects of poor drainage of rainfall inside the city’s western (polder) area, inadequate drainage capacity of pump stations, and high water levels of the outfall rivers. The eastern side of the city suffers from river floods, though plans have been developed to build an embankment along the Balu and Lakhya rivers and develop the eastern polder with a number of drainage pumps. When the western polder was completed after the 1988 flood, southeastern Dhaka was a rural setting free from river and urban (drainage congestion) flooding. But rapid industrialization and illegal Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7   1   2 Introduction encroachment along drainage channels in recent years have made this area highly vulnerable to urban flooding (box 1.1). Dhaka has a long history of regular inundation during the monsoon season. Documented incidence of river flooding can be traced back to 1787–88, when floodwaters following heavy monsoons submerged city streets to a depth that required boats for moving around (Hunter 1877). River floods again devastated the city in 1833–34 and 1870. In the past century, the floods of 1954, 1955, 1962, 1966, 1974, 1987, 1988, 1998, and 2004 were of major significance in terms of lost lives and property (Haque 2010; Nishat et al. 2000; Rahman 2006). Today, urban flooding from intense rainfall in the western and southern parts of the metropolis and river floods in the eastern part are recurring phenomena. Box 1.1 Clarification of Key Concepts Throughout this book, unless otherwise noted, the following definitions apply: • River flooding refers to the inundation that occurs when excessive runoff from upstream catchments leads to the river channel’s capacity being exceeded, resulting in water spilling over the bank; thus, the floodplain is flooded by water spilling over from the river channel. • Urban flooding refers to the inundation that occurs after heavy-rainfall events, which causes runoff to exceed the capacity of the local drainage system. As a result, water remains stag- nant in the built-up area for extended periods. This type of flooding is also known as rain flooding or waterlogging; however, these terms can also apply to a rural setting. • Waterlogging. See urban flooding definition. • Return period estimates the time interval between two hydrological events of the same size or intensity. It is a statistical measure of the average recurrence interval over a long period and is the inverse of the probability that the event will be exceeded in a single year. For example, the chances of 10-year, 20-year, and 30-year rainfall amounts occurring in any given year are 10, 5, and 3.3 percent, respectively, regardless of when the last similar event happened. Thus, a 10-year, return-period flood is 10 times more likely to occur, compared to a 100-year, return period event, which has a 1 percent probability. The amount of rainfall, expressed as depth (mm), increases with the return period, meaning that a 100-year rainfall event will be comparatively larger than a 20-year or 30-year event. For design purposes, choice of return period depends on multiple factors, such as size of the  drainage area, risk of failure, importance of at-risk structures, and decision-makers’ desired level of risk aversion. Shorter return periods tend to be based on observed data, where available, while longer ones are estimated from a statistical distribution. For urban areas susceptible to river flooding, the return period of urban flooding depends on the return period for both the rainfall event and water levels of the peripheral rivers. Thus, for unprotected areas, the return period of urban flooding is a joint probability function related to frequency of rainfall events and high river stages. Other factors that contribute to the return period of urban flooding events (e.g., land-cover changes and condition of ­ drainage box continues next page Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Introduction 3 Box 1.1  Clarification of Key Concepts (continued) channels), for which it is difficult to assign probability distributions, are usually kept constant for a given modeling scenario. • Adaptation refers to any activity undertaken to reduce or completely remove potential harm from climate change. It can be “planned” via a deliberate policy decision or “reactive,” after the change has happened. • Adaptation benefits are avoided damage costs or accrued benefits following the implemen- tation of adaptation measures. • Adaptation deficit refers to the gap between the current and desired or targeted prac- tice/performance. For example, if an urban drainage system designed to prevent flood- ing of streets from 100 mm rainfall in six hours can only achieve that when 70 mm falls in six hours, then it has an adaptation deficit of 30 mm in six hours. Before considering climate change adaptation, the adaptation deficit should be addressed; otherwise, the costs (and potentially the benefits) of measures to adapt to climate change will be overestimated. • Detention pond refers to a temporary water body that “detains” or delays the flood peak by storing some stormwater runoff for a certain period. This helps prevent the rest of the drain- age system from being overwhelmed by particularly large rainfall events. Water is removed from the detention pond by a pipe/culvert at the bottom of the pond (lowest point). When there is no rainfall over extended periods of time, the pond area remains dry and can be used for other purposes (e.g., agriculture or sports field). • Retention pond refers to a permanent water body that “retains” stormwater. During rainfall events, the water level increases up to a certain height, after which it is controlled by a pump or outlet structure. Impact of Recent Major Floods In the past few decades, both river and urban flooding have seriously impacted Dhaka in terms of floodwater inundation and duration, especially in fringe areas, mainly affecting slum dwellers. Residents suffered from health-related problems as a result of consuming contaminated drinking water. In addition, communica- tions, livelihoods, and service facilities were severely affected for days.1,2 The 1988 flood inundated about 85 percent of Dhaka (water depths were in a range of 0.3–4.5 m), isolating the city from the rest of the country for about two weeks. Water supplies were severely affected; and road, rail, air transport, and telecommunications were disrupted. A decade later, about 56 percent of Dhaka was inundated by the 1998 flood, which was unprece- dented in terms of duration and damage. Most of the city’s eastern portion suffered from river flooding, and 23 percent of its western part was affected by urban flooding (Hasnat 2006). Nearly 80 percent of the city was inun- dated at water depths of 0.3–3.0 m, which lasted about 10 weeks. Some two- thirds of city residents s­ uffered damage to assets, and daily wage earners (e.g., rickshaw pullers and v ­ endors) suffered direct loss of income. In all, some Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 4 Introduction three-fifths of residents were affected (Hasnat 2006). Rising water levels of the Buriganga, Turag, Balu, and Lakhya rivers affected the eastern and south- eastern portions of the city (Hossain, Azam, and Serajuddin 2006). In central areas, poor discharge capacity of drainage khals (open channels) increased the extent and duration of flooding. Some 66 km of khals and 80 km of storm- water drains were damaged (Huq and Alam 2003). At Narinda, the retention tank overflowed, affecting its only pumping ­ station. The estimated cost of rehabilitating the water supply system was about Tk. 127 million (Huq and Alam 2003). In terms of river levels, the 2004 flood was of shorter duration than the 1988 and 1998 events; however, water took longer to drain from Dhaka city areas and extreme rainfall in September made the situation worse. The flood wreaked havoc on major portions of the city’s population. More than 5 million people— two-fifths of city residents—were affected. Out of the city’s 22 thanas subdistricts), 18 were inundated. The eastern portion of the city was hardest hit. (­ Gulshan, Banani, Baridhara, and Nikunja experienced prolonged inundation, while the Motijheel commercial area, including Arambagh and Gopibagh, was underwater for a few days as a result of drainage congestion. Water pipelines stretching over a few hundred kilometers and many ponds were submerged. The sewerage system broke down, which resulted in contaminated drinking water, in turn, posing a serious threat to public health. The August 2007 flood exceeded the duration of the 1988 and 2004 events; however, the magnitude of peak flow above the danger mark for all rivers sur- rounding Greater Dhaka was less than 1988, 1998, and 2004 levels. Flooding of the Buriganga River receded within 24 hours, but lasted more than 20 days for other major rivers and canals. Eastern Dhaka was especially hard hit due to the absence of a flood-protection embankment, with boats serving as the only means of transport. Sections of major roads and bridges were washed away, and numer- ous drowning deaths were reported. Thousands of stranded residents lacked access to safe drinking water. Hospitals were overwhelmed with patients suffer- ing from water-borne illnesses, including more than 90,000 cases of diarrhea within a single week. Most patients were residents of the more severely inun- dated, low-lying areas. In July 2009, heavy monsoon rains battered the capital city. The national weather office reported more than 333 mm of precipitation within a 12-hour period; more than 290 mm fell within six hours, setting a 60-year record. Downed power lines resulted in at least nine deaths and numerous injuries. Rickshaw pullers struggled to carry passengers through knee-to-waist-deep water. Bus and ferry services were disrupted and major thoroughfares were submerged for hours, leaving thousands of stranded residents. In more vulner- able, low-lying areas, many families were trapped in their flooded homes, and slum dwellers suffered the most. Production was suspended at most industrial units in the southeastern area of Dhaka-Narayanganj-Demra (DND). Businesses and schools were forced to close, and trading on the stock market was delayed. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Introduction 5 Waterlogging and Drainage System Capacity The risk of waterlogging in Dhaka is particularly high during the monsoon season (June–September), when 70 percent of annual rainfall occurs (Zaman 2014). Traditionally, Dhaka’s river system and interconnected canals or khals formed the city’s lifeline. The network of myriad surrounding rivers includes the Tongi Khal to the north, the Balu and Lakhya to the east, the Dhaleswari to the south, and the Buriganga, Bangshai, and Turag to the west. In earlier centuries, the city had dozens of natural khals, which served as arteries for navigation routes within the greater metropolis. Many of these have since disappeared, owing mainly to unplanned urban development, encroachment, dumping of solid waste, poor system maintenance, and lack of coordination between government agencies (Shajahan and Nilufar 2013). The surviving khals, generally in poor condition, continue to serve as the city’s primary drainage system.3 Having to rely on this outdated system means that many urban residents—mainly those in the city’s middle- and lower-class areas— face prolonged inundation after even an hour of normal rainfall. Also, many low- lying areas that previously functioned as rainwater retention ponds are being filled for industrial and residential use, further impeding drainage in many parts of the city. As urban expansion continues, the city’s remaining open spaces and low-lying areas face more intense pressure from private land developers. If urgent measures are not taken to protect these vulnerable lands from further encroach- ment, the situation will likely worsen (Jahan 2014a). Dhaka’s aging drainage infrastructure (e.g., drainage pipes in Old Dhaka and pumps in the DND area) heightens the population’s vulnerability to the disaster risks from waterlogging. The situation is exacerbated by the increased paving of surfaces with impervious materials and a failure to implement activities that reduce flood risk. Waterlogging puts city dwellers and family assets, the urban- built environment, and vital public infrastructure at risk; in turn, industry and trade are harmed, and daily wage earners and temporary workers lose employ- ment and income (Roy 2014). Megacity in a Changing Climate Dhaka has witnessed tremendous population growth in the past century and is now among the world’s largest and fastest growing cities, with an estimated 34,000 people per km2. In 1980, the Greater Dhaka area (1,528 km2) entered the ranks of megacity, with a total population of 6.6 million; by 2011, that number had more than doubled, crossing the 14 million mark (BBS 2012). During 1985–95, its population growth rate averaged 5 percent a year, much higher than for other South Asian megacities. Today, approximately 13 percent of Bangladesh’s population—a staggering 22 million people—call Greater Dhaka their home. The United Nations projects that, by 2025, Dhaka will become the world’s fifth largest city, surpassed only by Tokyo, New Delhi, Mumbai, and São Paolo (UN 2010). According to the Institute of Water Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 6 Introduction Modelling (IWM 2012), Greater Dhaka’s population will have climbed to 29 million by 2035. Much of Dhaka’s current population growth is related to rural-to-urban migration. Since the founding of Bangladesh in 1971, Dhaka has drawn job seek- ers. As the process of industrialization got under way and service sectors became active, job opportunities were created in and around the capital city. The annual flow of migrants includes many poor farmers; since the farming sector cannot accommodate the younger generation entering the job market, many have had to move to the cities and towns to find work. Natural disasters (e.g., river bank erosion) have also contributed to the movement of rural people to Dhaka and other major cities. Urban Flood Vulnerability and Disaster Risk The effects of climate change, including intensified rainfall and greater fre- quency of extreme weather events, coupled with unplanned, rapid urbanization extending into the floodplains, are expected to aggravate Dhaka’s vulnerability to flooding (UNFCCC 2008; World Bank 2010; WWF 2009). The climate of South Asia is changing: the frequency of heavy-rainfall events is increasing, while light-rainfall events are decreasing (IPCC 2013). In the coming decades, all climate models and scenarios are predicting an increase in South Asia’s aver- age and extreme rainfall during the summer monsoon (IPCC 2014). Regarding seasonality, climate model agreement is high on an earlier onset and later retreat and thus longer duration of the summer monsoon. For tropical cyclones making landfall in South Asia, model agreement indicates that rainfall will likely be more extreme near the centers of such systems (IPCC 2013). On the whole, scientific evidence to date predicts high flood risk for low-lying areas in Bangladesh in a changing climate; and Dhaka has been identified as one of the top Asian cities in terms of population exposure to flooding by the 2070s (Hanson et al. 2011; IPCC 2014). Without adequate protection, Dhaka’s residents, particularly slum dwellers, will be continuously exposed to the risks of both river and urban flooding in a changing climate. Thus, policymakers must take targeted steps to better manage Dhaka’s flood infrastructure and integrate climate risks into local planning to strengthen the city’s climate-disaster resilience and adaptive capacity. Building a Sound Strategy for Local Action Preparing for the unfolding impacts of climate-related changes—­ sociodemographic, environmental, economic, and political—is a daunting task. It requires that Dhaka’s policymakers, like those in other megacities of South Asia’s low-lying deltas, understand how climate change is anticipated to affect natural hazards that lead to flooding in the cities. Given that the timing and magnitude of climate change impacts, though uncertain, show a trend, policymakers also have an urgent need to better understand how non-climate factors (e.g., inadequate drainage capacity and unplanned urban development in upstream catchments) have contributed to the severity of recent floods. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Introduction 7 Study Purpose and Approach This study’s overall aim is to provide local decision-makers an effective plan- ning approach for minimizing the damage risk of rainfall-induced urban flood- ing in Dhaka in a changing climate (box 1.1). Specific objectives are to assess the vulnerability of the Greater Dhaka area to urban flooding and waterlogging, estimate probable economic damage due to climate change, develop structural adaptation measures, evaluate the reduction in economic damage resulting from implementing these measures, and estimate their cost. The list of loca- tion-specific interventions and their costing, in turn, can help the city’s local planners to meet their goals of minimizing the total expected disaster costs and reduction measures. Equipped with a menu of options designed the cost of risk-­ to close the current adaptation deficit and further climate-proof urban infra- structure, it is expected that local decision-makers will better evaluate their choices and develop effective strategies that prioritize and sequence activities as resources permit. The study comprises four main activities: (a) hydrological modeling and development of adaptation measures, (b) spatial ranking of flood vulnerability, (c) evaluating expected damage from flooding, and (d) estimating adaptation costs (figure 1.1). The overall objectives and scope of the hydrological modeling exercise were to assess climate change impacts on location-specific waterlogging for various scenarios, focusing on the inundation depth and duration caused by extreme precipitation events for the high-emissions, fossil-fuel intensive (A1FI) scenario in 2050. The baseline selected for the modeling exercise was the September 2004 extreme rainfall event (341 mm in 24 hours), which caused one of the heaviest urban flooding damages in recent history. In order to approximate the Figure 1.1 Overview of Study Approach Review of past studies Review of drainage plans Setup/update of mathematical Data collation and analysis Secondary data/map collection models Limited primary data collection Presentation of flood maps for Simulation of base scenario Interaction with stakeholders study area Simulation of climate change scenarios Summary of inundation by Simulation of adaptation administrative area measures Damage estimation Adaptation measures Interaction with stakeholders Flood vulnerability analysis Cost of adaptation Source: Adapted from Zaman 2014. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 8 Introduction Box 1.2 Expected Increase Factor in Extreme 24-hour Precipitation for Dhaka in 2050  In a changing climate, it is expected that Dhaka will experience a 16 percent increase in extreme 24-hour precipitation by 2050, compared to the 2004 baseline for this study. The statistical downscaling used in Sugiyama (2012) estimated a 16 percent increase factor in ­ extreme 24-hour precipitation for Dhaka in 2050 for the high-emissions, fossil-fuel intensive (A1FI) scenario. To develop the estimates, the methods of Allen and Ingram (2002) and Pall, Allen, and Stone (2007) were used. The study started with the 12 climate models used for the Fourth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC). However, since precipitation extremes increase at different rates by climate model, it was not possible to make a definite conclusion about future precipitation extremes based solely on climate models. For this reason, the study drew on the physical basis that relates change in precipitation extremes with moisture avail- ability in the atmosphere (e.g., Allen and Ingram 2002; Trenberth et al. 2003), and set the ­ fractional change of precipitation extremes to change in precipitable water (i.e., the amount of water in a column of air that could fall as precipitation) at 9 percent per K° for Dhaka. The uncertainty range for estimates of precipitation extremes may be set at 3–28 percent per K°, based on the full range of the model-based statistics. Sources: Allen and Ingram 2002; IPCC 2013; Pall, Allen, and Stone 2007; Sugiyama 2012; Trenberth et al. 2003. future (2050) climate and minimize biases in precipitation estimated from cli- mate models, the study used pattern scaling, a form of statistical downscaling, to convert low-resolution, General Circulation Model (GCM) outputs to regional and city scale (Sugiyama 2012) (box 1.2).4 For hydrological modeling, local engineers were consulted to understand the functioning of each existing drainage system in the detailed study area. In accor- dance with recommendations received from the experts, all future scenarios assumed that all improvements to Dhaka’s drainage infrastructure—both planned and proposed by the relevant drainage-system authorities—would be implemented. In order to identify interventions to close the remaining gap in the existing drainage capacity, flood modeling simulations were conducted with dif- ferent pump capacities and diameters of drainage pipes until resulting flood maps met pre-set acceptable flooding criteria for inundation depth and duration for each modeled zone. Based on these simulations, structural measures beyond current planned improvements were developed to address the current and cli- mate change adaptation deficits. Drawing on this hydrological modeling data, along with information gener- ated from a Climate Disaster Resilience Index (CDRI) analysis for Dhaka (Jahan 2014b),5 the study then developed a flood vulnerability index (FVI) that ranked the population and infrastructure of Dhaka’s major areas and wards, according to their relative exposure, susceptibility, and resilience. Next, it evaluated the direct Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Introduction 9 and indirect economic damage to the local population from a 100-year return- period rainfall event in 2050, a probability-weighted damage of different return- period rainfall events occurring in 2050, and cumulative damage between 2014 and 2050, using random assignment of various return-period extreme rainfall events. Finally, the adaptation costs of the proposed structural measures were estimated for each modeled zone in the detailed study area. The costs of their ­ implementation were assessed against the 2004 baseline (100-year, return- period) extreme rainfall event. These cost estimates were divided into two ­ components: (a) additional investments required to meet the current adaptation deficit without climate change and (b) further measures required to meet the 2050 climate change deficit. The cost estimates should be considered indicative since the extent of expected climate change remains uncertain and the process of change, as viewed from a planning perspective, is slow (Huq 2014). To estimate the costs of the various recommended measures, a number of assumptions were made, some of which were supplied by DWASA officials. Numerous practitioners were consulted, including officials and engineers from DWASA, the Bangladesh Water Development Board (BWDB), the Public Works Department (PWD), the Institute of Water Modelling (IWM), the World Bank, and Dhaka North and South City Corporations (DNCC and DSCC), as well as contractors, suppliers, businessmen, and independent consultants. In addition, several field visits were undertaken to physically inspect the situation around Dhaka city (appendix A). Launched in 2012 and conducted over two-and-one-half years, the study was funded by the Bangladesh Climate Change Resilience Fund (BCCRF), a coordi- nated financing mechanism of the Government of Bangladesh, development partners, and the World Bank, with additional technical support provided by the IWM. Since the project outset, consultative workshops were convened annually, and two Focus Group Discussions (FGDs) were held. The consultative workshops presented study objectives and methods (June 2012); shared preliminary findings on climate change impacts (April 2013); and discussed study results on Dhaka’s flood vulnerability, adaptive capac- ity, and adaptation costs (August 2014). At the first FGD, held in May 2013, experts recommended location-specific, structural adaptation measures, taking into account preliminary findings on climate change impacts. At the second one, held in August 2014, stakeholders agreed on a strategy to mitigate flooding in built-up areas, including ways to reduce peak surface runoff during extreme rainfall events, remove blockage from drainage channels and sewerage pipes, and reduce flood duration in low-lying areas. In addition, meetings and discussions have been held at various stages of the project with drainage system engineers at the DWASA, BWDB, Narayanganj City Corporation (NCC), and IWM. The study also reviewed a range of secondary sources. Previous recommenda- tions on improving drainage-system performance were included in the modeling of future scenarios for Eastern Dhaka (Halcrow Bangladesh Limited 2006), Central Dhaka (RAJUK 2008), the DND area (BWDB 2010), and Kallyanpur Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 10 Introduction (IWM 2011). The study also reiterated recommendations from IWFM (2005, 2008) and UGIIP (2006). RAJUK’s Detailed Area Plan (RAJUK 2010) and drainage and stormwater master plans (IWM 2006; DWASA 2014) were reviewed, as were earlier studies on drainage-system criteria (JICA 1987, 1990) and flood protection of Western Dhaka (Louis Berger International 1991). Defining the Detailed Study Area The general physical setting for this study is a 1,528 km2 area in north-central Bangladesh that corresponds to RAJUK’s Dhaka Metropolitan Development Plan (DMDP). It consists of the Dhaka metropolitan area, the DND area, and the townships of Narayanganj, Savar, Gazipur, and Tongi. For the latter three townships, detailed data on drainage infrastructure, which is required for location-specific mapping of inundation depth and duration, was lacking. ­ Therefore, the detailed study area of 345 km2 was limited to the Dhaka metro- politan area, the DND area, and Narayanganj township. The Dhaka metropolitan area consists of two parts: Eastern Dhaka (121 km2) and Western Dhaka (143 km2). Western Dhaka, in turn, is divided into four areas: Old Dhaka, Central Dhaka, Kallyanpur, and Goranchatbari. The DND area covers 59 km2, while Narayanganj township is a 22 km2 area. Key Features of Modeled Zones Based on the drainage system, the detailed study area was divided into seven models (map 1.1). Each model is briefly described below, while more detailed descriptions are provided in chapter 3. Old Dhaka. A historic area south of Central Dhaka, the Old Dhaka model spans about 12 km2. The area is bounded by Border Guard Bangladesh and Dhaka University campus to the north, Dhaka flood wall and embankment to the west and southwest, and Atish Dipankar-Jatrabari-Mawa Road in the east and southeast. Control structures, known as sluice gates, allow for stormwater to drain into the river; however, they mostly remain closed during the mon- soon season due to the high river-water levels, functioning only when the river- water level drops below the ponded water in front of the sluice gate. Old Dhaka was one of the worst-affected areas during the extreme rainfall events of 2004 and 2009. In 2004, about half of the area’s wastewater contributed to the volume to be drained. Since then, operation and maintenance of the sluices have improved. Even so, recurrent waterlogging persists in low-lying parts of the area. Central Dhaka. Situated in the center of the detailed study area, Central Dhaka is bounded by Goranchatbari to the north, Eastern Dhaka to the east, Old Dhaka to the south, and Kallyanpur to the west. Mainly flat terrain covers about 39 km2, including low-lying areas near Khilgaon, Bashabo, and Kamlapur. Two major river embankments—Western Flood Protection Embankment and Pragati Sharani—protect Central Dhaka from river flooding. Population growth is rapid in such high-density areas as Shantinagar and Malibagh. The area also has pockets Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Introduction 11 Map 1.1  Detailed Study Area, Showing Seven Drainage Models Source: IWM 2014. that are relatively sparsely populated (e.g., Ramna and Gulistan). By 2050, it is anticipated that more than half of all land will be zoned for urban residential and mixed use. Central Dhaka has three major lakes, which serve as retention ponds. All three are connected to the Balu River via the Begunbari Khal. Sluice gates control the lakes’ flow to the khal. Kallyanpur. This densely populated area of about 28 km2 is situated at the western edge of Dhaka along the banks of the Turag River. Kallyanpur shares adjoining borders with Goranchatbari to the north, Central Dhaka to the east, Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 12 Introduction and Old Dhaka to the south. The Western Embankment along the left bank of the Turag River protects the area from river flooding. The area’s several sluice gates, located along the embankment, remain closed during the monsoon sea- son. Its stormwater is pumped from the Kallyanpur Retention Pond, located in the northwest part of the drainage system. Kallyanpur is a key contributor to Dhaka’s economy, featuring a variety of manufacturing and processing fac- tories; commercial activities; and offices, clinics, and schools. By 2050, more than three-fifths of land will have been zoned for either residential use or the transportation network. Goranchatbari. Spanning nearly 64 km2 in northwestern Dhaka, Goranchatbari shares borders with the adjoining areas of Eastern Dhaka to the east, Central Dhaka to the southeast, and Kallyanpur to the south. The Western Embankment— from the Tongi Rail Station to Kellarmor on the city’s west side—protects the area from flooding of the Turag River. The drainage system is predominantly natural. Khals carry stormwater to the Goranchatbari Pumping Station, where it is pumped to the Turag River. In the northeast, stormwater drains to the Tongi Khal through sluice gates. The area varies widely socioeconomically. In planned, densely populated areas, most residents are engaged in commerce. In unplanned, moderately populated areas, residents work mainly in manufactur- ing and ­industrial-related activities. Many women in unplanned areas work in garment and apparel factories. Eastern Dhaka. This 121 km2 area features a mix of urban and rural land- scapes interspersed with numerous wetlands and khals. It is bounded to the north by the Tongi Khal; the Balu River to the east; Demra Road to the south, which forms a boundary with the DND area; and Pragati Sharani to the west, which protects Western Dhaka from river flooding. Land use in Eastern Dhaka is dedicated mainly to agriculture and residential use. Khals are the main component of the drainage system. Outfalls of the khals are mainly to the Balu River; however, some also discharge into Tongi Khal. A road bypass likely to be completed by 2025 will function as embankment along the right bank of the Balu River and Tongi Khal and protect Eastern Dhaka from river flooding. DND (Dhaka-Narayanganj-Demra). The DND is a 59 km2 area in south- east Dhaka originally developed as an irrigation project by the BWDB in 1962. The area is bounded by the Dhaka-Demra-Chittagong Road to the north, the Lakhya River to the east, Narayanganj township to the south, and the Buriganga River to the west. Flood embankments along the Lakhya and Buriganga rivers prevent river flooding. The land is crisscrossed by numerous khals. In the past, the DND area was relatively free of flooding; however, recent haphazard urbanization, as well as encroachments on drainage chan- nels, has resulted in waterlogging during the monsoon season. The 2004 flood inundated 70 percent of the area, highlighting the urgent need for drainage system improvements. Narayanganj. Situated at the southern edge of Dhaka, Narayanganj is one of Bangladesh’s oldest and largest inland ports. The township spans about Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Introduction 13 22 km2. The Lakhya and Dhaleswari rivers form respective natural boundar- ies on the east and west sides. These rivers meet near Munshiganj and con- tinue as the Dhaleswari River until merging with the Meghna River, which eventually flows into the Bay of Bengal. To the north, the Narayanganj Highway separates Narayanganj from the DND area. The Narayanganj- Munshiganj Highway functions as an embankment that protects some of the western portion from flooding of the Dhaleswari River. The drainage system comprises covered street drains considered as box-culverts, a piped network, khals, and large ponds. Most residents are slum dwellers employed as indus- trial laborers and porters. A small portion of the population is engaged in agriculture. Hydrological Features The hydrological setting of the detailed study area consists of climate, river network, topography, and land cover (box 1.3). The tropical monsoon climate comprises four hydrological seasons: pre-monsoon (April–May), monsoon (June–September), post-monsoon (October–December), and dry season (January–March). Most river and urban flooding problems occur during the monsoon season. Annual average rainfall is about 2,000 mm. At the start of the monsoon and post-monsoon seasons, cyclones with strong winds hit Bangladesh, sometimes causing heavy rainfall in Dhaka city with subsequent urban flooding. An intricate river network surrounds the detailed study area. To the north, Tongi Khal (15 km long) is connected to the Turag River at its upstream and to the Balu River at its downstream. The Turag River (75 km long) is connected at its downstream to the Buriganga River, which, in turn is connected to Dhaleswari Box 1.3 Topography of Dhaka at a Glance  Dhaka is situated at the southern tip of the Madhupur Tract, a Pleistocene terrace 1–10 m above the adjacent floodplains. The areas surrounding the city are mostly flat alluvial plain formed during the Holocene age or in recent time. The city is encircled by six rivers and canals. The Turag, Buriganga, and Dhaleswari rivers are located on the western side, while the Balu and Lakhya rivers are situated to the east. Tongi Khal, a manmade canal, connects the Turag and Balu rivers to complete the encirclement. The city’s central area or spine is the highest part, with land sloping to the west, east, and south directions. The high-elevated part continues on the northern side, dividing the city into three major blocks from a flood management perspective. A peripheral dyke, completed after the 1988 flood, forms the western polder, protecting western Dhaka. The eastern side remains unprotected from river flooding, which may occur due to spillage of floodwater from the Balu and Lakhya rivers. The southern part—a low-lying floodplain that was polderized by raising the Dhaka-Narayanganj and Dhaka-Demra roads and connecting Demra and Narayanganj with a flood embankment—is now known as the Dhaka-Narayanganj-Demra (DND) Polder. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 14 Introduction River at its downstream. The Balu River (30 km) is connected to the Lakhya River at its downstream. The Lakhya River (120 km long) is connected to the Dhaleswari River at its downstream. These rivers are primarily distributaries of the Brahmaputra River. The topography of the detailed study area is a flat alluvial plain with irregular contour lines and numerous khals that subdivide the land. About 55 percent of the land is at an elevation of 6–8 m PWD, with just 15 percent above 8 m PWD. The area is dominated by a 5–8 m deep layer of cohesive soils of clay and plastic silt with poor permeability, which overlay a sand layer. Depending on the loca- tion of built-up areas, the soil’s poor permeability has consequences for urban flooding. Structure of the Book This book is divided into two major parts: study methods (part 1) and results (part 2). Part 1 is organized into four chapters. Chapter 2, which will be of inter- est to hydrologists, presents the conceptual framework for the hydrological mod- eling study and the method and processes used to model future scenarios. Chapter 3, of interest to local planners as well as hydrologists, presents the seven model setups for the detailed study area. The following two chapters will be of interest to economists and risk-assessment experts. Chapter 4 describes the method used to assess the modeled areas’ relative vulnerability to urban flooding, while chapter 5 explains the method used to estimate their economic damage. Part 2, also organized into four chapters, is designed mainly for Dhaka’s local poli- cymakers. Chapter 6 provides the study results for each of the seven modeled areas, while chapter 7 summarizes the overall study results for Dhaka city. Chapter 8 turns to the secondary flood-mitigation measures that can comple- ment the recommended conveyance-centric solutions. Finally, chapter 9 suggests a way forward. Notes 1. In the 2007 flood, more than 90,000 people in Dhaka city alone were affected by diarrhea. 2. Nishat et al. (2000) provides a detailed study of the various adverse impacts of the 1998 flood. 3. Seventeen of Dhaka’s 43 natural canals no longer exist. Recently, however, the Dhaka Water Supply and Sewerage Authority (DWASA) has brought the remaining 26 under its control. Rehabilitation is currently under way for 19 of these—11 under the second phase of the government’s Removal of Water-Logging Project and 8 under the World Bank–funded Dhaka Water Supply and Sanitation Project. 4. Sugiyama (2012) constitutes voluntary work conducted by the author. The views expressed in this report are solely those of the author and do not reflect the views of the author’s present or former institutional affiliations. The report describes an exten- sion of an earlier study conducted in 2008. Since then, the climate science community has made significant strides in further developing the pattern-scaling technique, which Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Introduction 15 the report uses extensively. For a summary of recent scientific development, interested readers should consult IPCC (2013, section 12.4.2). 5. The CDRI analysis is based on a planning tool developed under the Climate and Disaster Resilience Initiative of Kyoto University (Joerin and Shaw 2011). To measure climate-disaster resilience, the CDRI considers five dimensions (physical, social, eco- nomic, institutional, and natural), each of which has five parameters; these, in turn, have five variables each, for a total of 125 variables (Jahan 2014b). References Allen, M. R., and W. J. Ingram. 2002. “Constraints on Future Changes in Climate and the Hydrologic Cycle.” Nature 419: 224–32. doi:10.1038/nature0192. BBS (Bangladesh Bureau of Statistics). 2012. Bangladesh Population and Housing Census 2011. National Series, Vol. 4. Dhaka: Bangladesh Bureau of Statistics. BWDB (Bangladesh Water Development Board). 2010. Feasibility Study on Drainage Improvement of Dhaka-Narayanganj-Demra (DND) Project. Dhaka: Bangladesh Water Development Board. DWASA (Dhaka Water Supply and Sewerage Authority). 2014. “Preparation/Updating of Stormwater Drainage Master Plan for Dhaka City.” Prepared by Jurutera Perunding Zaaba Sdn. Bhd. (Malaysia), Farhat Consulting Engineers and Architects Limited, and SARM Associates Limited for the World Bank–funded Dhaka Water Supply and Sanitation Project, Dhaka. Halcrow Bangladesh Limited. 2006. Updating/Upgrading the Feasibility Study of Dhaka Integrated Flood Control Embankment cum Eastern Bypass Road Multipurpose Project. Dhaka: Halcrow Bangladesh Limited. Hanson, S., R. Nicholls, N. Ranger, S. Hallegatte, J. Corfee-Morlot, C. Herweijer, and J. Chateau. 2011. “A Global Ranking of Port Cities with High Exposure to Climate Extremes.” Climate Change 104: 89–111. Haque, M. A. 2010. “Experience of Bangladesh on Flood Disaster Mitigation.” Presentation at UN ESCAP meeting. http://www.unescap.org/idd/events/2010 Pakistan Floods II​ /­Bangladesh.pdf. Hasnat, M. A. 2006. “Flood Disaster Management in Dhaka City.” In Options for Flood Risk and Damage Reduction in Bangladesh, edited by K. U. Siddiqui and A. N. H. A. Hossain. Dhaka: University Press Limited and Institute of Water Modelling. Hossain, A. N. H. A., K. A. Azam, and M. Serajuddin. 2006. “The Impact of Floods on Dhaka City’s Water Supply, Sanitation and Drainage, and Mitigation Options.” In Options for Flood Risk and Damage Reduction in Bangladesh, edited by K. U. Siddiqui and A. N. H. A. Hossain. Dhaka: University Press Limited and Institute of Water Modelling. Hunter, Sir William Wilson. 1877. A Statistical Account of Bengal. Vol. 5. London: Trübner and Company. Huq, Mainul. 2014. Cost of Adaptation to a Changing Climate. Background report pre- pared for this study, World Bank, Washington, DC. Huq, S., and M. Alam. 2003. “Flood Management and Vulnerability in Dhaka City.” In Building Safer Cities: The Future of Disaster Risk, edited by A. Kreimer, M. Arnold, Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 16 Introduction and A. Carlin, chapter 9. Disaster Risk Management Series. Washington, DC: World Bank. IPCC (Intergovernmental Panel on Climate Change). 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, edited by T. F. Stocker, D. Qin, G.-K. Plattner, M. Tignor, S. K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex, and P. M. Midgley, sections 14.6., 14.8.9–12. Cambridge, U.K.: Cambridge University Press. ———. 2014. Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part B: Regional Aspects. Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, edited by V. R. Barros, C. B. Field, D. J. Dokken, M. D. Mastrandrea, K. J. Mach, T. E. Bilir, M. Chatterjee, K. L. Ebi, Y. O. Estrada, R. C. Genova, B. Girma, E. S. Kissel, A. N. Ley, S. MacCracken, P. R. Mastrandrea, and L. L. White, sections 24.3.2.2, 24.4.5.3.1. Cambridge, U.K.: Cambridge University Press. IWFM (Institute of Water and Flood Management). 2005. Investigation of Hydrological Aspects of 2004 Flood, with Special Emphasis on Dhaka City. Dhaka: Institute of Water and Flood Management, Bangladesh University of Engineering and Technology. ———. 2008. Hydrological Aspects of 2007 Flood. Dhaka: Institute of Water and Flood Management, Bangladesh University of Engineering and Technology. IWM (Institute of Water Modelling). 2006. Study on Drainage Master Plan for Dhaka City. Dhaka: Institute of Water Modelling. ———. 2011. Modeling Study for Ascertaining the Area of Regulating Pond for Kallyanpur Pumping Station. Dhaka: Institute of Water Modelling. ———. 2012. Water Supply Master Plan for Dhaka City. Dhaka: Institute of Water Modelling. Jahan, Sarwar. 2014a. Flood Management Strategies for Dhaka: An Evaluation from Institutional Perspective. Background report prepared for this study, World Bank, Washington, DC. ———. 2014b. Assessment of Local Capacity to Deal with Flooding and Water Logging. Background report prepared for this study, World Bank, Washington, DC. JICA (Japan International Cooperation Agency). 1987. Study on Storm Water Drainage System Improvement Project in Dhaka City. Tokyo: JICA. ———. 1990. Updating Study on Storm Water Drainage System Improvement Project in Dhaka City. Tokyo: JICA. Joerin, J., and R. Shaw. 2011. “Climate and Disaster Resilience in Cities.” In Community, Environment and Disaster Risk Management, Vol. 6, edited by R. Shaw and A. Sharma. Bingley, U.K.: Emerald Group Publishing Ltd. Louis Berger International. 1991. Dhaka Integrated Flood Protection. FAP 8B, final report, A. D. B. T. A. No. 1318-BAN. Washington, DC: Louis Berger International. Nishat, A., M. Riazuddin, R. Amin, and A. R. Khan, eds. 2000. The 1998 Flood: Impact on Environment of Dhaka City. Dhaka: IUCN Bangladesh Country Office. Pall, P., M. R. Allen, and D. A. Stone. 2007. “Testing the Clausius-Clapeyron Constraint on Changes in Extreme Precipitation under CO2 Warming.” Climate Dynamics 28: 351–63. doi:10.1007/s00382-006-0180-2. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Introduction 17 Rahman, A. K. M. F. 2006. “The Impact of Floods on Major Highways and Mitigation Options.” In Options for Flood Risk and Damage Reduction in Bangladesh, edited by K. U. Siddiqui and A. N. H. A. Hossain. Dhaka: University Press Limited and Institute of Water Modelling. RAJUK (Rajdhani Unnayan Kartripakkha). 2008. Integrated Development of Hatir Jheel Area. Dhaka: RAJUK. ———. 2010. Preparation of Detailed Area Plan for Dhaka Metropolitan Development Planning (DMDP) Area, GROUP B. Final plan, prepared by Engineering and Planning Consultants Ltd. for RAJUK, Dhaka. Roy, Subhendu. 2014. Urban Flooding of Greater Dhaka Area in a Changing Climate: Vulnerability, Adaptation and Potential Costs. Background report prepared for this study, World Bank, Washington, DC. Shajahan, Amreen, and Farida Nilufar. 2013. “Reviving the Inland Waterways of Dhaka: A Morphological Approach towards an Integrated Transportation System.” In Proceedings of the Ninth International Space Syntax Symposium, edited by Y. O. Kim, H. T. Park, and K. W. Seo. Seoul: Sejong University. Sugiyama, M. 2012. “Study on Future Climate Scenarios for Asian Cities.” Prepared for the World Bank. http://www.bccrf-bd.org/Documents/pdf/Note%20on%20Climate​ %20Downscaling%20by%20Sugiyama.pdf. Trenberth, K. E., A. Dai, R. M. Rasmussen, and D. B. Parsons. 2003. “The Changing Character of Precipitation.” Bulletin of the American Meteorological Society 84: 1205–17. doi:10.1175/BAMS-84-9-1205. UGIIP (Urban Governance and Infrastructure Improvement Project). 2006. Drainage Master Plan of Narayanganj Pourashava. Dhaka: Urban Governance and Infrastructure Improvement Project. UN (United Nations). 2010. World Urbanization Prospects: The 2009 Revision. New York: Department of Economic and Social Affairs, Population Division, UN. UNFCCC (United Nations Framework Convention on Climate Change). 2008. “Physical and Socioeconomic Trends in Climate-Related Risks and Extreme Events, and Their Implications for Sustainable Development.” http://unfccc.int/resource/docs/2008​ /­tp/03.pdf. World Bank. 2010. Climate Risks and Adaptation in Asian Coastal Megacities: A Synthesis Report. Washington, DC: World Bank. WWF (World Wide Fund for Nature). 2009. Mega-Stress for Mega-Cities: A Climate Vulnerability Ranking of Major Coastal Cities in Asia. Gland, Switzerland: World Wide Fund for Nature. http://assets.panda.org/downloads/mega_cities_report.pdf. Zaman, Asif M. 2014. Urban Flooding of Greater Dhaka Area in a Changing Climate: Vulnerability, Adaptation and Potential Costs. Background report prepared for this study, Bangladesh Climate Change Resilience Fund and Institute of Water Modelling, Dhaka. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 P A RT 1 Methods Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7   19   Chapter 2 Hydrological Modeling Introduction Estimating the impact of an extreme rainfall event on the magnitude and dura- tion of flooding requires hydrological modeling. Hydrological models attempt to represent the processes by which various forms of precipitation or rainfall are converted into runoff or overland flow of water, which then drains or flows through depressions or channels into water courses. Depending on the modeling objective, this representation is done over various time and spatial scales. This chapter is divided into three sections. The first one describes the conceptual framework for the hydrological modeling exercise, including the required three levels of modeling. The second reviews the rainfall analyses undertaken, while the third provides a general overview of the process used to model future scenarios for the detailed study area. Modeling Framework As mentioned in chapter 1, Dhaka is located on the lower reaches of the Ganges- Brahmaputra Delta, surrounded by six rivers and canals. In addition, flooding in the study area of Greater Dhaka is affected by flows in the Brahmaputra (Jamuna) River and its distributaries in north-central Bangladesh. Since the study needed to assess flooding changes in the Dhaka Metropolitan Development Plan (DMDP) area, which is mainly influenced by river flooding, a three-step model- ing approach was adopted (figure 2.1). In the first step, basin-level hydrological modeling simulated flows from the Brahmaputra River in the Ganges-Brahmaputra-Meghna (GBM) system since they strongly influence monsoon-season river flows and water levels in the extended study area. In the second step, the effects of climate change on the regional rivers were simulated. Finally, in the third step, detailed modeling of the drainage system in and around Dhaka city was simulated. Although river levels affect the drainage system within the city area, the focus of this study was the worst-case scenario, meaning that river levels are high and all sluice gates are closed; that is, there is no gravity drainage out of the city and the drainage system is primarily dependent on Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7   21   22 Hydrological Modeling Figure 2.1 Conceptual Framework for the Modeling Exercise Step 1 Other input Basin level hydrological data modeling Flows of DEM of major rivers DMDP area Step 2 Indicative flood Other input Regional hydrological and maps of extended data hydrodynamic modeling study area Flows and DEM of water levels of Dhaka area regional rivers Step 3 Other input Flood maps of Detailed hydrological and data detailed study area hydrodynamic modeling Data set Model elements Data set Source: Zaman 2014. Note: DEM = Digital Elevation Model; DMDP = Dhaka Metropolitan Development Plan. the performance efficiency of drainage pumps. Thus, changes in river levels due to climate change were not considered in the analyses of the detailed study area.1 Basin-level model output was used to create boundary conditions for region- level hydrological and hydrodynamic modeling. As input, the region-level step simulated flows from the basin-level modeling, along with data from the Digital Elevation Models (DEMs) of RAJUK’s DMDP area and climate and river-­ network information. Region-level model outputs were river flows and water levels of key rivers in north-central Bangladesh. This simulated data was used to generate indicative flood maps for the DMDP area. Outputs from region-level simulations were used as boundary conditions for modeling the detailed study area. The urban-area modeling step involved a more detailed level of hydrologi- cal and hydrodynamic modeling of the urban drainage system and catchments. Model outputs were stormwater flows and flood levels, which were used to generate location-specific depth-and-duration maps. The subsections below pro- vide more detailed descriptions of the three modeling levels. Basin-Level Modeling The basin-level model for the GBM system developed by the Institute of Water Modelling (IWM) was used to estimate future flow changes of the Brahmaputra (Jamuna) River due to climate change. This model was selected because it had already been calibrated and validated for large flood events and had been used for several climate change studies (e.g., IWM and CCC 2008). This is a water mass-balance model that runs at a daily time-step, based on the MIKE Basin software package (Nishat and Rahman 2009). Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Hydrological Modeling 23 The model was used to simulate daily river flows for the base period and future monsoon river flows of the Brahmaputra (Jamuna) River with climate change–adjusted catchment rainfall and temperature series and without climate change. The selected base period was 1998–2008, which covers extreme, normal, and less than normal monsoon flows. The catchment rainfall and temperature series for climate change were based on the estimates of Masahiro Sugiyama for the Brahmaputra Basin (Sugiyama 2012). Monsoon-season flows of the Brahmaputra River were simulated for future (2050) low- and high-emissions scenarios. Selection of the year 2050 was considered appropriate in terms of deriving a clear climate signal and being close enough to be relevant for today’s planners and decision-makers. Modeling input obtained from the IWM included meteorological (e.g., rainfall, evapotranspiration, and temperature), river-channel (e.g., alignment, catchment area, and distributary channels), and catchment-properties (e.g., runoff coeffi- cient, baseflow parameters, and snowmelt characteristics) data. The 543,467 km2 Brahmaputra Basin was modeled with a total of 34 catchments, including 4 snowmelt ones. The Nedbør-afstrømnings (Rainfall-Runoff) Model (NAM) was used to represent the hydrological processes. The 1998–2008 base simulation period was calibrated at Bahadurabad Station. Figure 2.2 shows that agreement between the simulated and observed flows was good for some years and poor for others. The model performance was measured using the Nash Sutcliffe Coefficient  (E) for each year and monsoon season (June–September). Whole-year performance Figure 2.2  Basin Model Calibration Plot 120,000 100,000 80,000 Discharge (m3/s) 60,000 40,000 20,000 0 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 Observed Q Simulated Q Source: Zaman 2014. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 24 Hydrological Modeling varied from 0.85 in 1999 to 0.05 in 2001. The model performed better (0.76) when only monsoon-season flows were considered. Based on the calibration results, the model performance was deemed acceptable for the purposes of the extended study area. In keeping with the objective of assessing changes in monsoon flows, key assumptions were made for future basin scenarios. It was considered that flood- ing of the Greater Dhaka area would be dominated by flows from the Brahmaputra (Jamuna) River. The water levels of the Ganges and the Meghna were not considered important as they only influence the outflow boundary condition. It may be of some importance when the peaks of the major rivers coincide, as happened in 1988, and drainage from upstream rivers is impeded. It was assumed that future land use and land cover would not change dramatically (i.e., catchment runoff coefficients would be similar).2 It was further assumed that future basin water losses through evapotranspiration would not differ signifi- cantly in the monsoon season. Additional assumptions were that peak monsoon flows would be minimally affected by the building of future dams, hydrological regimes would not be significantly altered by water demands, a lumped calcu- lated method would be used to model snowmelt, and snowmelt modeling procedures would include glacier contribution to runoff. ­ Region-Level Modeling To model the Greater Dhaka region, the study used IWM’s North Central Region Hydrodynamic (NCRHD) model to simulate monsoon seasonal flows and water levels of the river network in the extended study area (map 2.1). Developed in 1991, the NCRHD model has been validated for 14 hydrological years (up to 2005–06). It is linked to MIKE 11 RR, a lumped conceptual rainfall- runoff model that generates flows from catchments. Surface runoff generated from rainfall is estimated for each catchment. The flows are then used as an input to a hydrodynamic model that simulates the flow of water through drainage channels and surrounding rivers of Dhaka city. The hydrodynamic model also simulates the operation of infrastructure at the end of the drainage system. For this study, the NCRHD model was used to simulate daily flood flows and water levels for the base year (2004 monsoon).3 Then, using adjusted catchment rainfall series for climate change based on Masahiro Sugiyama’s estimates for catchments in the north-central region (Sugiyama 2012) and adjusted flows from Chilmari (Jamuna River), the NCRHD model was used to simulate future (2050 monsoon season) river flows and water levels for the A1FI scenario.4 Based on the simulated maximum water levels, flood depths were calculated and mapped using GIS for each scenario. The NCRHD model covers a 14,223 km2 area divided into 20 catchments. The total length of rivers and khals is about 3,006 km. Calibration and validation of the model are based on available observed discharge and water-level data. The detailed input data required for hydrodynamic modeling, already avail- able at IWM, included meteorological (e.g., rainfall and evapotranspiration), river- water level and discharge, and hydraulic structures data, as well as information on Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Hydrological Modeling 25 Map 2.1 Regional Rivers Surrounding Dhaka City Source: IWM 2014. river channels (e.g., alignment, slope, width, bank, and bed friction) and ­catchment properties (e.g., runoff coefficient and baseflow parameters). ­Daily-point rainfall data from manual rain gauges at 31 rainfall stations maintained by Bangladesh Water Development Board (BWDB) in north-central Bangladesh was used to calculate the mean area rainfall for each catchment in the region (map 2.2). Pan evaporation data was collected from 6 evaporation stations m ­ aintained by BWDB.5 Observed river-water level data were available at 41 ­ stations in the north-central area; water-level, time-series data were also required for model calibration and validation. Similarly, river discharge data was required as both an input into the hydrodynamic model and for model calibration and validation. For the base scenario, measured discharge data were collected from BWDB stations.6 The rainfall-runoff model comprises 20 subcatchments—19 in Bangladesh and 1 in India—which drain into the Old Brahmaputra River. The rainfall-runoff model receives precipitation and evaporation from 31 and 6 stations, respectively, within the areas where measurements are available. The NAM model usually uses abstraction data processed by the Water Resources Planning Organisation, based on the census of the National Minor Irrigation Development Project. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 26 Hydrological Modeling Map 2.2 River System and Hydrological Stations of the North Central Region Hydrodynamic Model Source: IWM 2006. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Hydrological Modeling 27 Land-Use and Land-Level Data Analysis of 2004 and future land-cover data was based on RAJUK’s Detailed Area Plan (DAP) in GIS format and IWM’s projected population data (IWM 2012). Land levels were a key input for flood mapping of the extended and detailed study areas. The study used an existing DEM available from IWM, based on topographic sheets from the Survey of Bangladesh (Government of Bangladesh 1979) and past IWM surveys. For future changes in terrain, the DEM was adjusted according to a relationship between population density and land levels in proposed built-up areas in the DAP (e.g., urban residential, institutional, and mixed use) (Jahan 2014) (table 2.1).7 Land levels for other land-use zones in the original DEM (e.g., rural settlements, open spaces, water bodies, flood flows, and agriculture) remained unchanged. Thana-wise population projections were obtained from the Dhaka City Water Supply Master Plan (IWM 2012). Boundary Conditions The model boundaries comprised 10 upstream and 6 downstream locations. The change in upstream flows were simulated in the GBM Basin model and used as input in the boundary condition at Chilmari Station. From the basin model simulations, it was observed that the monsoon flows increased due to climate change for the A1FI emissions scenario in 2050. Other considerations included wastewater inflows, sea-level rise, and subsidence. It was found that wastewater flows during the monsoon are negligi- ­ ble compared to high river flows; thus, the NCRHD model did not consider them. Potential changes in sea-level rise can affect river-water levels around Dhaka during the monsoon season due to possible backwater effects from the Padma and Meghna rivers. Based on a review of the Special Report on Managing the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation (SREX) by the Intergovernmental Panel on Climate Change (IPCC 2012), it was determined that the effect of sea-level rise in the study area by 2050 would be negligible.8 The study also reviewed subsidence rates for Dhaka Table 2.1 Relationship between Population Density and Land Levels Density (thousand people/km2) Land level (m PWD) 50 and above 7.5 45–50 7.5 40–45 7.0 35–40 7.0 30–35 6.5 25–30 6.5 20–25 6.0 15–20 5.5 10–15 5.0 Less than 10 4.5 Source: Jahan 2014. Note: PWD = Public Works Department. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 28 Hydrological Modeling using secondary sources, notably Alam (1996) and Hoque and Alam (1997). The former study gave a rate of 0.65 mm per year, while the latter reported 1.88 mm a year. Using the higher rate, this study calculated that total subsidence by 2050 would be less than 10 cm, which was considered negligible. Model Calibration and Validation The model was calibrated based on a comparison of simulated and observed data at 34 locations where measured water level and discharge data were available. The calibrated model was validated against known events without ­ changing model parameters. From the comparison plots, it was evident that the model-generated water levels matched reasonably well with the observed data (figure 2.3). Therefore, the model was considered suitable for application runs to study climate change scenarios. Key Assumptions Key model assumptions were that land use and land cover would change in accordance with the DAP, land levels in built-up areas would increase as popula- tion densities increased, and land levels for other land-use categories (e.g., rural settlements, water bodies, and agriculture) would not increase. These assump- tions were confirmed through various workshops and stakeholder consultations. Figure 2.3 Comparison of Simulated and Measured Water Levels (m PWD) at Selected BWDB Stations, 2004 a. Balu River at Demra b. Buriganga River at Mill Barrack 7 7 6 6 5 5 4 4 3 3 2 2 May June July August September October May June July August September October c. Turag River at Mirpur d. Dhsleswari River at Savar 7 7 6 6 5 5 4 4 3 3 2 2 May June July August September October May June July August September October Measured Simulated Source: Zaman 2014. Note: BWDB = Bangladesh Water Development Board; PWD = Public Works Department. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Hydrological Modeling 29 Furthermore, they assumed that the DAP would be implemented and enforced, which would prevent changes in land uses that could lead to land-level changes. It was also assumed that water losses through evapotranspiration would not dif- fer significantly in the future, no major changes in the river network would occur in the future, and water demands would not significantly alter the hydrological regime in the monsoon. Urban-Area Modeling For modeling of the detailed study area, two types of models were used: MIKE 11 and MIKE Urban. MIKE 11 is a one-dimensional model that estimates water flows and levels by solving non-linear, one-dimensional Saint Venant equations that mathematically describe the motion of fluids. This model can also simulate the performance of various drainage structures, such as pipes, khals (open chan- nels), box-culverts, bridges, pumps, and sluice gates. MIKE Urban is a coupled one- and two-dimensional model that links the capabilities of MIKE 11 in a GIS framework so that DEMs can be used to simulate the flow of water on the ground surface. Chapter 3 provides details on the modeling requirements. Rainfall Analyses The study reviewed previous research that investigated trends and changes in rainfall frequencies in and around Dhaka City for various time periods. It then analyzed and compared historical rainfall data from the BWDB and the Bangladesh Meteorological Department (BMD). Results of these rainfall analy- ses, along with a review of frequency analyses done by JICA (2000), are pre- sented in appendix B, while appendix C summarizes the findings from previous research. Process for Modeling Future Scenarios Figure 2.4 shows the general process used to model future scenarios in the detailed study area. Design rainfall events were adjusted by climate change fac- tors for A1FI future scenarios (box 1.2). These were inputted into the MIKE Urban A rainfall-runoff model with calculated changes in catchment impervious area due to the densification driven by population growth. Next, the resulting simulated stormwater runoff was inputted into the one-dimensional hydrody- namic model, along with planned changes in the drainage infrastructure (e.g., increased pump capacity and re-sectioning of khals). The one-dimensional model provided estimates of water levels at various points for each time-step. For the areas modeled with MIKE 11, the data on peak-water level was brought into GIS format and analyzed with reference to the DEM of the model area to produce flood maps. Changes in land levels due to land filling or local subsidence were represented by changing the DEM accordingly. For the areas modeled with MIKE Urban, the flooding process was done automatically due to its one- and two-dimensional coupling. That is, when pipes and khals overflow, Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 30 Hydrological Modeling Figure 2.4 Process for Modeling the Detailed Study Area Rainfall changes, A1FI (2025 and 2050) Expected land cover changes (2025 and 2050) Rainfall-runo modeling Sea-level rise estimates Drainage infrastructure (2025 and 2050) scenarios (2025 and 2050): Business as usual 1-D hydrodynamic modeling Planned development Subsidence estimates Full adaptation (2025 and 2050) DEM (2025 and 2050) Map showing waterlogging depths and duration (2025 and 2050) for each scenario and at di erent probabilities of occurrence Source: Zaman 2014. Note: A1FI = Intergovernmental Panel on Climate Change scenario with future based on fossil fuel–intensive development; DEM = Digital Elevation Model. the water is spread across the DEM; after the peak runoff passes, the water follows the natural slope of the DEM to re-enter the drainage system. As a result, ­ MIKE Urban outputs a time series of inundation maps. Simulations of 2050 Scenarios The models described above were used to simulate various future scenarios, whose outputs were flood depth-and-duration maps. As mentioned in chapter 1, the 100-year, extreme rainfall event that occurred in September 2004 (341 mm in 24 hours) was taken as the baseline to simulate future scenarios in 2050. The selection of this baseline event was based on simulations and analyses that showed a clear trend of increasing frequency of intense rainfall events. The underlying assumption was that such extreme events will become more frequent due to climate change processes. Also, most of the available datasets coincided at that time. These included land-use data from the DAP, infrastructure data from existing IWM models, and DEM data. The effects of the 2004 event were simulated for five future scenarios (table 2.2). For all scenarios, detailed drainage models incorporated projected population growth and densities, planned DAP land-use changes, estimated land-level changes, and planned drainage improvements. The resultant flood- ­ ing patterns in the future scenarios were compared to specified acceptable flood- ing criteria. These were set through consultations with experts and relevant Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Hydrological Modeling 31 stakeholders in Bangladesh, taking into consideration that residents of Dhaka, where flooding is a recurrent phenomenon, have become accustomed to even ankle-deep water, and the vertical accuracy of the current DEM does not allow one to go below 0.10 m. The criteria set were that, in the peak flood situation, 90 percent of each administrative area should have less than 0.25 m flood depth—with the exception of Central Dhaka, where the threshold was set at 0.10 m—and the duration of inundation should be less than 12 hours. Modeling Assumptions: Planned Improvements Development of the five future scenarios described in table 2.2 assumed that all improvements to Dhaka’s drainage infrastructure—both planned and proposed by the relevant drainage-system authorities—would be implemented. These included improvements covered in RAJUK’s DAP (RAJUK 2010), DWASA’s Sewerage Master Plan, and Narayanganj City Corporation’s Concept Vision Plan. Previous recommendations on improving drainage-system performance were included in the modeling of future scenarios for Eastern Dhaka (Halcrow Bangladesh Limited 2006), Central Dhaka (RAJUK 2008), the DND area (BWDB 2010), Kallyanpur (IWM 2011), and Narayanganj (UGIIP 2006). Modeling assumptions also included recommendations of DWASA (2014), IWM (2006), and IWFM (2005, 2008). Table 2.2 Model Simulations for the Detailed Study Area Scenario Data sources/remarks Base run, 2004 100-year return- DAP’s land-cover, land-use, and topography data; IWM’s drainage infrastructure period storm; 1 in 10-, 20-, data and Dhaka DEM. and 30-year events 2050 SEa + PIb + rain events DAP’s proposed land-cover and land-use data; changes in land levels due to with no CCc expansion of built-up areas. 2050 SEa + PIb + CCc Drainage improvements based on published reports and discussions with relevant agencies. A1FI climate change factors were derived from Sugiyama (2012) for increased rainfall (based on statistical downscaling of GCM results) and NCRHD model simulations for changes in peripheral river-water levels. 2050 SEa + PIb + AId + Additional investments to meet the flooding criteria set through consultations: rain events with no CCc 90% of administrative area has less than 0.25 m flood depth (0.10 m for Central 2050 SEa +PIb + AId + CCc Dhaka) and duration of less than 12 hours. 2050 SEa + PIb + AId + CCc + ADe Design parameters of adaptation measures based on consultations with relevant stakeholders and experts. Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; AD = adaptation to climate change; DAP = Detailed Area Plan. a. Socioeconomic changes refer to changes in land use and DEMs. b. Planned improvements refer to drainage-system improvements by relevant agencies. c. Climate change refers to the effects of the IPCC’s A1FI scenario (i.e., future based on fossil fuel–intensive development) applied to the design storm events and river levels. d. Additional investments refer to improvements to address the current climate adaptation deficit without climate change to meet the acceptable flooding criteria set for the extreme rainfall event. e. Adaptation refers to the adaptation measures modeled to mitigate the impacts of climate change to satisfy the acceptable level of flooding in each administrative area. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 32 Hydrological Modeling Computing Flood Damage in a Changing Climate The computations for estimating flood damage and adaptation costs were based on the estimates of extent, depth, and duration of flooding, as described above. The avoided damage from adaptation was based on the 100-year, return-period flood occurring in 2050. Because of the likelihood that such extreme events may become more frequent in the future due to climate change processes, the study also examined the flood damage that would result if the 100-year, return-period storm were to hit with greater frequency (i.e., once every 50, 33, or 20 years). Notes 1. All three levels of modeling were undertaken by the Institute of Water Modelling (IWM). 2. It was beyond the scope of this study to assess how land use and land cover may change in the basin area with and without climate change. 3. The year 2004 was chosen as the base year as key datasets for drainage modeling were available for this flood year. 4. The A1FI scenario was chosen since it provides an upper bound for the possible effects of climate change. 5. Potential evapotranspiration data from direct measurements were not available. Pan evaporation expresses the volume of water that evaporates from a standard evapo- ration pan over a given period of time. This is the most widely used method by hydrologists for calculating actual evaporation rates of large water bodies. 6. IWM and BWDB are the main sources of river discharge data in the north-central region of Bangladesh. 7. Prior experience suggests that the conversion into built-up areas involves land filling. 8. The SREX report indicates that sea-level rise above 2 m by 2100 would be implau- sible and that “an estimate of 0.8 m by 2100 that included increased ice dynamics was considered most plausible” (IPCC 2012, p. 179). References Alam, M. 1996. “Subsidence of the Ganges-Brahmaputra Delta of Bangladesh and Associated Drainage, Sedimentation and Salinity Problems.” In Sea-Level Rise and Coastal Subsidence, edited by J. D. Milliman and B. U. Haq, 169–92. Dordrecht, Netherlands: Kluwer Academic Publishers. BWDB (Bangladesh Water Development Board). 2010. Feasibility Study on Drainage Improvement of Dhaka-Narayanganj-Demra (DND) Project. Dhaka: Bangladesh Water Development Board. DWASA (Dhaka Water Supply and Sewerage Authority). 2014. “Preparation/Updating of Stormwater Drainage Master Plan for Dhaka City.” Prepared by Jurutera Perunding Zaaba Sdn. Bhd. (Malaysia), Farhat Consulting Engineers and Architects Limited, and SARM Associates Limited for the World Bank–funded Dhaka Water Supply and Sanitation Project, Dhaka. Government of Bangladesh. 1979. Survey of Bangladesh. Topographic maps of Bangladesh compiled between 1952 and 1964. Dhaka: Government of Bangladesh. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Hydrological Modeling 33 Halcrow Bangladesh Limited. 2006. Updating/Upgrading the Feasibility Study of Dhaka Integrated Flood Control Embankment cum Eastern Bypass Road Multipurpose Project. Dhaka: Halcrow Bangladesh Limited. Hoque, M., and M. Alam. 1997. “Subsidence in the Lower Deltaic Areas of Bangladesh.” Marine Geodesy 20: 105–20. IPCC (Intergovernmental Panel on Climate Change). 2012. Managing the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation. A Special Report of Working Groups I and II of the Intergovernmental Panel on Climate Change, edited by C. B. Field, V. Barros, T. F. Stocker, D. Qin, D. J. Dokken, K. L. Ebi, M. D. Mastrandrea, K. J. Mach, G.-K. Plattner, S. K. Allen, M. Tignor, and P. M. Midgley. Cambridge, U.K.: Cambridge University Press. IWFM (Institute of Water and Flood Management). 2005. Investigation of Hydrological Aspects of 2004 Flood, with Special Emphasis on Dhaka City. Dhaka: Institute of Water and Flood Management, Bangladesh University of Engineering and Technology. ———. 2008. Hydrological Aspects of 2007 Flood. Dhaka: Institute of Water and Flood Management, Bangladesh University of Engineering and Technology. IWM (Institute of Water Modelling). 2006. Study on Drainage Master Plan for Dhaka City. Dhaka: Institute of Water Modelling. ———. 2011. Modeling Study for Ascertaining the Area of Regulating Pond for Kallyanpur Pumping Station. Dhaka: Institute of Water Modelling. ———. 2012. Water Supply Master Plan for Dhaka City. Dhaka: Institute of Water Modelling. IWM and CCC (Institute of Water Modelling and Climate Change Cell). 2008. Impact Assessment of Climate Change and Sea Level Rise on Monsoon Flooding. Dhaka: Institute of Water Modelling and Climate Change Cell, Department of Environment, Ministry of Environment and Forests, Government of Bangladesh. Jahan, Sarwar. 2014. Climate and Disaster Resilience of Greater Dhaka Area: A Micro-Level Analysis. Background report prepared for this study, World Bank, Dhaka. JICA (Japan International Cooperation Agency). 2000. Baseline Information Study of the Dhaka Combined Flood Control cum Eastern Bypass Road Project: Final Report. Dhaka: Bangladesh Water Development Board, Ministry of Water Resources, Government of Bangladesh. Nishat, B., and S. M. M. Rahman. 2009. “Water Resources Modeling of the Ganges- Brahmaputra-Meghna River Basins Using Satellite Remote Sensing Data.” Journal of the American Water Resources Association 45 (6): 1313–27. RAJUK (Rajdhani Unnayan Kartripakkha). 2008. Integrated Development of Hatir Jheel Area. Dhaka: Rajdhani Unnayan Kartripakkha. ———. 2010. Preparation of Detailed Area Plan for Dhaka Metropolitan Development Planning (DMDP) Area, GROUP B. Final plan, prepared by Engineering and Planning Consultants Ltd. for Rajdhani Unnayan Kartripakkha, Dhaka. Sugiyama, M. 2012. “Study on Future Climate Scenarios for Asian Cities.” Prepared for the World Bank. UGIIP (Urban Governance and Infrastructure Improvement Project). 2006. Drainage Master Plan of Narayanganj Pourashava. Dhaka: UGIIP. Zaman, Asif M. 2014. Urban Flooding of Greater Dhaka Area in a Changing Climate: Vulnerability, Adaptation and Potential Costs. Background report prepared for this study, Bangladesh Climate Change Resilience Fund and Institute of Water Modelling, Dhaka. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Chapter 3 Detailed Area Modeling Introduction The detailed study area was divided into seven areas for modeling: Old Dhaka, Central Dhaka, Kallyanpur, Goranchatbari, Eastern Dhaka, DND, and Narayanganj. Map 3.1 shows the model boundaries, respective areas modeled, and types of models used. Deciding on which model to apply depended on the drainage-system features of each area. For example, in Eastern Dhaka and the DND area, where khals constitute the primary drainage system, MIKE 11 was used to assess future river flood situations. In Western Dhaka and Central Dhaka, where the primary drainage system consists of pipes and khals, MIKE Urban was used to predict urban flood problems. Modeling Setup Data Requirements and Sources To provide location-specific depth and duration of inundation, model analyses required detailed spatial and temporal resolution data. This included detailed information on topography, meteorology, drainage infrastructure,1 land cover, and population density. Topographical data included a Digital Elevation Model (DEM) (25 m2 grid) prepared by the Institute of Water Modelling (IWM), using information from their various projects, as well as the Survey of Bangladesh (Government of Bangladesh 1979). Meteorological data included historical daily rainfall provided by the Bangladesh Water Development Board (BWDB) and historical three-hourly rainfall provided by the Bangladesh Meteorological Department (BMD). Sedimentation measurements for drainage pipes and ­ box-culverts were estimated by IWM via field inspections and discussions with Dhaka Water Supply and Sewerage Authority (DWASA) engineers. The Demographic Census of Bangladesh for 2001 and 2011, conducted by the Bangladesh Bureau of Statistics (BBS), provided the source for estimating population density (table 3.1). ­ Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7   35   36 Detailed Area Modeling Map 3.1 Model Boundaries of Detailed Study Area Source: IWM 2014. Table 3.1 Summary of Data Requirements and Availability Data type Source Topography (DEM) IWM, Survey of Bangladesh Meteorological information BMD and BWDB Drainage infrastructure DWASA Land cover BUET and RAJUK River stage BWDB River flow BWDB River sections, slope, and alignments IWM Population density BBS Note: BBS = Bangladesh Bureau of Statistics; BMD = Bangladesh Meteorological Department; BUET = Bangladesh University of Engineering and Technology; BWDB = Bangladesh Water Development Board; DEM = Digital Elevation Model; DWASA = Dhaka Water Supply and Sewerage Authority; IWM = Institute of Water Modelling; RAJUK = Rajdhani Unnayan Kartripakkha (Capital Development Authority of Bangladesh). Design of Base Model and Future Scenarios Each base model was calibrated against a schematic inundation map of Dhaka city for the year 2004 (IWFM 2005). For this purpose, temporally concentrated rain- fall data was required; thus, the BMD three-hourly rainfall data was used. Design rainfall patterns were made by segregating one-day rainfall data from JICA (2000) into six hourly rainfalls following the pattern recommended by JICA (1987). To simulate the 2050 A1FI emissions scenario for Dhaka, the design rainfalls were Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Detailed Area Modeling 37 adjusted using the 16 percent increase factor in extreme 24-hour precipitation estimated by Sugiyama (2012) (box 1.2). Appendix B provides further details. MIKE Urban A, a lumped hydrological model that uses the time-area method, was selected as the most appropriate method to represent generation of runoff from local rainfall. MIKE Urban A uses various parameters to control the amount of surface runoff. These include size of the contributing (impervious) area, initial loss (wetting and filling of local depressions/storages), and reduction factor (for evapotranspiration and imperfect imperviousness). The average daily evaporation in Dhaka for the month of September was estimated at 4 mm per day (based on a rate of 118 mm per month). Total rainfall within the simulation period (eight- day event of September 2004) was 660.5 mm. The percentage of evaporation thus computed was 5 percent of total rainfall (32/660.5). It was assumed that another 5 percent of rainfall is lost due to evapotranspiration and depression stor- age (not contributing to runoff). Therefore, a value of 0.9 was used as the reduc- tion factor (table 3.2). Estimation of Impervious Area The 2004 and proposed land-cover information was based on Detailed Area Plan (DAP) data. Based on the various types of land cover identified in each catchment, the percentages of pervious and impervious area were estimated for the detailed study area, using typical values in the field (i.e., the actual percent values corresponding to a particular land cover) (table 3.3). For example, for resi- dential land use, DND (a low-density area), Central Dhaka (a medium-density area), and Old Dhaka (a high-density area) have impervious areas of 17 percent, 38 percent, and 70 percent or higher, respectively.2 For example, in the residential-use category, the percentage of impervious area in low-density areas varies from 18 percent to 34 percent as the population den- sity increases from 10,000 people per km2 to 20,000 per km2. In high-density areas, the percentage of impervious area varies from 60 percent to 70 percent as the population density increases from 40,000 people per km2 to 50,000 per km2 (table 3.3). In this way, the relationship captures the time dimension for future (2050) scenarios. Modeling Steps Each model setup included hydrological and hydraulic models. The first step was to develop the hydrological model. Input data, including catchment properties Table 3.2 Typical Parameters and Values for MIKE Urban Model A Parameter Value used Impervious area (%) 70 Time of concentration (minutes) 120 Initial loss (mm) 10 Reduction factor 0.9 Time area curve no. Varies with catchment shape Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 38 Detailed Area Modeling Table 3.3 Impervious Percentage for Various Land Uses Population density (thousand people/km2) 10 15 20 25 30 35 40 45 50 Low density Moderate density High density DAP land type Impervious area by land type (%) Agriculture 0 1 1 1 1 1 1 1 1 Circulation network 2 3 4 4 5 6 6 7 7 Commercial activity 19 27 35 44 50 56 63 68 73 Community service 21 30 39 48 55 62 69 74 80 Diplomatic 9 14 18 22 25 28 31 34 36 Education and research 11 16 21 26 30 34 38 41 44 Governmental services 9 14 18 22 25 28 31 34 36 Manufacturing and processing activity 15 22 28 35 40 45 50 54 58 Mixed use 21 30 39 48 55 62 69 74 80 Recreational facilities 4 5 7 9 10 11 13 14 15 Residential 18 26 34 42 48 54 60 65 70 Restricted area 2 3 4 4 5 6 6 7 7 Service activity 9 14 18 22 25 28 31 34 36 Transport and communication 8 11 14 18 20 23 25 27 29 Vacant land 2 3 4 4 5 6 6 7 7 Water body not connected to drains 0 1 1 1 1 1 1 1 1 Water body connected to drains 100 100 100 100 100 100 100 100 100 Source: Jahan 2014. Note: DAP = Detailed Area Plan. and available climate information, were used to simulate surface runoff. These data, along with a DEM of the study area and information on the drainage network, sedimentation depth, and pump discharge, were used as inputs in the ­ detailed-area hydraulic models to generate time-series inundation data. These simulated data, in turn, were used to generate location-specific flood maps and an inundation depth and duration table. Old Dhaka Model Setup The model selected for Old Dhaka was MIKE Urban, and the area modeled covered about 12 km2. The Old Dhaka area is protected by embankments along the western boundary. The stormwater networks drain into the Buriganga River through control structures (sluice gates) and Dholaikhal Pump Station. Embankments in the west (Western Flood Protection Embankment) protect Old Dhaka from river flooding (map 3.2). Future Scenario: Socioeconomic Changes In 2011, Old Dhaka had an estimated population of some 1.1 million, with an extremely high density of about 94,000 people per km2. Data from the cen- sus for 2001 and 2011 was linearly interpolated to estimate the 2004 base-­ scenario density (92,305 per km2). Using the 2011 census data, future changes Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Detailed Area Modeling 39 Map 3.2 Physical Setting of Old Dhaka Model with Adjoining Boundaries Source: IWM 2014. in population densities up to 2050 were projected for Old Dhaka thanas (129,435 per km2 in 2050) (IWM 2012). Only 23 percent of the study area is below 6 m in elevation, including some low-lying areas below 5 m in elevation along the banks of the Buriganga River. The total area above 8 m in elevation accounts for 18 percent of the study area. Another 59 percent is in a range of 6–8 m in elevation. The study area also fea- tures a few small water bodies. For flood modeling, the DEM grid provided by IWM was used for the 2004 base model; for future scenarios, it was adjusted based on projected population densities and DAP-proposed land uses, as described in chapter 2 (map 3.3). In 2004, about 94 percent of the study area was dedicated mainly to four uses: mixed use (48 percent), residential use (26 percent), education and research (11 percent), and transportation network (9 percent). The remaining 6 percent was used for government services and commercial activities, restricted and historical zones, water bodies, and scant vacant land. By 2050, pronounced changes in land- use patterns are expected for residential areas in the east and southeast, where respective plans include mixed use and dedicated industrial use. For example, Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 40 Detailed Area Modeling Map 3.3 Change in Land Elevation over the Study Period, Old Dhaka Model Source: IWM 2014. Map 3.4 Changes in Proposed Land Use over the Study Period, Old Dhaka Model Source: IWM 2014. some residential areas near Wari and Mirhazaribag have been declared mixed-use zones, and area near Shyampur has been declared a general industrial, instead of a residential, zone (map 3.4). Drainage System: Base Scenario and Planned Improvements Pipe networks dominate Old Dhaka’s drainage system. For the 2004 base sce- nario, the system includes some 24 km of stormwater pipes and 3.64 km of box- culverts and open intake canal of Dholaikhal Pump Station. For future scenarios, planned improvements include another 6 km of pipe networks, along with reha- bilitated and diverted pipes. All pipes and box-culverts were assumed to drain into the Buriganga River (map 3.5, photo 3.1). Other Parameters and Features The area’s only permanent pumping station, at Dholaikhal, has a 22.4 m3/s capacity, with three pumps (each with a 7.47 m3/s capacity). Since installation, Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Detailed Area Modeling 41 Map 3.5 Planned Drainage System Improvements for Old Dhaka Source: IWM 2014. only two of the three pumps have been run simultaneously, with the third kept on stand-by (photo 3.2). But during the 2004 flood, only one pump was operational, and 22 temporary pumps (each with a 0.14 m3/s capacity) were ­ installed at various locations. These pumps were also considered for future sce- narios. For permanent pumps, an efficiency level of 80 percent was considered, while 75 percent was considered for temporary pumps. As flood-control struc- tures, the six sluice gates were included in the base model. For all scenarios, these gates were kept closed, as the high anticipated river-water level would block natural drainage, depicting the worst possible scenario for urban flooding. For the 2004 base scenario, the model was divided into 262 subcatchments. These were delineated according to the distribution of drainage networks and land terrain. For future scenarios, some of the subcatchments were split further due to installation of new drainage pipes, resulting in a total of 275 subcatch- ments (table 3.4). Sedimentation depth was estimated by IWM via field inspections and discus- sions with DWASA engineers. For pipes, sedimentation depth was estimated as 20 percent of pipe diameter and, for box-culverts, as 33 percent of section height (table 3.4). The 2004 base model included 50 percent contribution of wastewater to the drainage system. A study that compared wastewater and runoff contribution for the 2004 base case concluded that the ratio between peak runoff (from the 1-in- 30-year event, with a peak intensity of 93 mm per hour) and the wastewater rate was less than 1 percent for 255 out of the 262 subcatchments. For the other 7, the ratio was within 5 percent. The study also calculated the ratio of total accumulated Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 42 Detailed Area Modeling Photo 3.1 Site investigation to measure sediment depth in box-culvert Credit: © Asif Zaman / World Bank. Used with permission. Further permission required for reuse. wastewater volume to total accumulated runoff (from the 1-in-30-year design storm, at 211 mm per day). It was seen that 247 out of 262 subcatchments had a ratio below 20 percent, suggesting that wastewater flow has a minimal effect on flooding. Thus, wastewater was not included in the 2050 scenarios (table 3.4). Key Assumptions The following key assumptions were made for the 2050 simulations for the Old Dhaka model: • All stormwater from the model area drains out to the Buriganga River through the drainage networks. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Detailed Area Modeling 43 Photo 3.2  Drainage pumps at Dholaikhal Pump Station, Old Dhaka Credit: © Asif Zaman / World Bank. Used with permission. Further permission required for reuse. Table 3.4  Development Baseline: Key Assumptions Comparison between 2004 and 2050, Old Dhaka Model Parameter or key feature 2004 2050 Population density (persons/km2, weighted average) 92,305 129,435a Land level (m PWD, average) 7.24 7.68 Subcatchments (number) 262 275 Imperviousness (%, weighted average) 71 76 Time of concentration (minutes, weighted average) 88.26 86.53 Initial loss (m) 0.01 0.01 Reduction factor 0.9 0.9 Total length of pipes, box-culverts, and khals (km) 27.77 33.88 Wastewater (% of total domestic load) 50 0 Sedimentation (%) Pipes (diameter) 20 20 Box-culverts (section height) 33 33 Ponding area (km2) 0.011 0.011 Pumps (number) Permanent 3 3 Temporary 22 22 Total pump capacity (m3/s) 25.48 25.48 Working efficiency of pumps (%) Permanent 80 80 Temporary 75 75 Sluice gates (number) 6 6 a. This projected value is based on past trends; densification is presently continuing, but at a comparatively lower rate. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 44 Detailed Area Modeling • All sluice gates/regulators remain closed during the monsoon season; that is, there is no natural gravity drainage to the river (worst-case flooding). • Embankments in the west protect the area from river flooding.3 • Operating efficiency of permanent pumps at Dholaikhal Station is considered as 80 percent; temporary pumps are considered to have 75 percent efficiency. • Sediment depth remains 20 percent of the diameter of drainage pipes and 33 percent of the section height of box-culverts. • Initial water level at Dholaikhal is considered as 4 m Public Works Department (PWD). • In 2004–12, significant changes were made in the drainage system of BUET and in Islambag areas. In other areas, the drainage system remains almost the same, except for construction of some new drainage lines and a few rehabilitation works. DWASA has no major plans for improving Old Dhaka’s drainage system. • Wastewater will not contribute to the stormwater drainage system in the future as a result of DWASA’s implementation of the Sewerage Master Plan for Dhaka (DWASA 2012). Boundary Conditions For all scenarios, the water level of the Dohlaikhal Pump Station (4 m PWD) was used as an initial boundary condition. For future scenarios, initial boundary con- ditions also included discharge data from the Central Dhaka model. None of the scenarios used the water level of the Buriganga River as an external boundary condition because all six flood control structures (sluice gates) are expected to remain closed during the peak of the monsoon season. Central Dhaka Model Setup The MIKE Urban platform was selected for Central Dhaka, and the area modeled covered about 39.2 km2 located in the center of the detailed study area. ­ The model area has four adjoining models: Goranchatbari to the north, Eastern Dhaka to the east, Old Dhaka to the south, and Kallyanpur to the west (map 3.6). Several wide roads serve as transportation arteries in more densely populated areas, such as Malibag and Shantinagar, while narrow roads are com- mon in areas of relatively low-population density, such as Ramna and Gulistan. Embankments in the west and east (Western Flood Protection Embankment and Pragati Sharani, respectively) protect Central Dhaka from river flooding. Future Scenario: Socioeconomic Changes Central Dhaka has a rapidly growing population, whose overall density is projected to reach about 95,000 per km2 by 2050—58,000 per km2 more than ­ in 2011 (IWM 2012). Population densities above 60,000 per km2 are projected for all but two thanas (Cantonment and Demra). For three-fourths of the area, land elevation is in a range of 6.5–8.0 m PWD; another 13 percent is above 8.5 m PWD, and the other 12 percent—with the exception of three lakes—is below 6.5 m PWD (2004 figures). A uniform land level above 7 m PWD is expected to have been achieved by 2050 (map 3.7). Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Detailed Area Modeling 45 Map 3.6 Physical Setting of Central Dhaka Model with Adjoining Boundaries Source: IWM 2014. A substantial portion of land has been zoned for either urban residential (40.17 percent) or mixed use (12.32 percent); by 2050, these two categories will still account for more than half of all land use (map 3.8). New developments will be more pronounced in eastern areas, and southeastern residential areas near Motijheel will be re-zoned for mixed use. By 2050, impervious land will have reached 82 percent. Preservation of the area’s three lakes—located at Hatir Jheel, Gulshan, and Banai—is expected. All three serve as retention ponds, with a total ponding area of 1.91 km2. The lakes are connected to the Balu River via the Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 46 Detailed Area Modeling Map 3.7 Change in Land Elevation over the Study Period, Central Dhaka Model Source: IWM 2014. Begunbari Khal. Sluice gates and drainage pumps at the Rampura Bridge control the lakes’ flow to the khal. Drainage System: Base Scenario and Planned Improvements Central Dhaka’s drainage system consists of pipes, box-culverts, and khals (map 3.9). The 2004 base scenario includes about 104.3 km of stormwater pipes, 9.7 km of box-culverts, and 10 km of open canals. Additional drainage pipeline (1,500 mm diameter) has been recently laid along Gulshan Lake to protect it from wastewater pollution; 34 km2 at the center of the area drains to Hatir Jheel, while the rest drains to Manda Khal. The area has two temporary pump stations, Rampura and Maniknagar, whose respective pump capacities are 4.5 m3/s and 2.5 m3/s (photo 3.3). In 2004, the total length of the drainage network was about 124 km, with a total pump capacity of 7 m3/s. Future scenarios assumed a total drainage-­ network length of 139 km, with a total pump capacity of 25–40 m3/s at Rampura and 15 m3/s at Maniknagar. The start-and-stop levels for permanent pumps were considered as 5.25 m and 5.0 m, respectively.4 Future scenarios assumed efficiency levels of 90 percent for permanent pumps (table 3.5). ­ Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Detailed Area Modeling 47 Map 3.8 Changes in Proposed Land Use over the Study Period, Central Dhaka Model Source: IWM 2014. Other Parameters and Features The 2004 model area was divided into 315 subcatchments, delineated according to the distribution of drainage networks and land terrain (table 3.5). In future scenarios, the impervious area of subcatchments will increase from 10–96 per- cent (2004) to 14–99 percent (2050). The Hatir Jheel cross-section will be implemented in accordance with the RAJUK proposal (RAJUK 2008). Sedimentation depth was estimated by IWM via field inspections and discussions with DWASA engineers. For pipes, sedimentation depth was estimated as 20 percent of pipe diameter and, for box-culverts, as 20 percent of the section height (table 3.5). All scenarios assumed that wastewater flow is insignificant compared to the runoff volume generated by the extreme rainfall events being simulated. Key Assumptions The following key assumptions were made for the 2050 simulations for the Central Dhaka model: • All stormwater is pumped out to the Begunbari and Manda Khals (open chan- nels of the Eastern Dhaka model), and eventually falls to the Balu River. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 48 Detailed Area Modeling Map 3.9 Planned Drainage System Improvements, Central Dhaka Source: IWM 2014. • A portion of the runoff from ward 75 drains to Dholaikhal in the Old Dhaka model. • All sluice gates, located at Rampura, Kamalapur, and Bashabo, remain closed during the monsoon season; this means there is no gravity drainage to the river, depicting the worst case of urban flooding. • Embankments in the west and east protect the area from river flooding. • The two temporary pump stations at Rampura and Maniknagar will become permanent pump stations with a total capacity of 40 m3/s; permanent pumps will have a 90-percent operating efficiency. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Detailed Area Modeling 49 Photo 3.3 Pumping floodwater out of Hathir Jheel using temporary pumps, Central Dhaka Credit: © Asif Zaman / World Bank. Used with permission. Further permission required for reuse. Table 3.5  Development Baseline: Key Assumptions Comparison between 2004 and 2050, Central Dhaka Model Parameter or key feature 2004 2050 Population density (persons/km2, weighted average) 33,160 95,160 Land level (m PWD, average) 7.03 7.54 Subcatchments (number) 315 315 Imperviousness (%, weighted average) 68 82 Time of concentration (minutes, weighted average) 24 22 Initial loss (m) 0.01 0.01 Reduction factor 0.9 0.9 Total length of pipes, box-culverts, and khals (km) 124 139 Sedimentation (%) Pipes (diameter) 20 20 Box-culverts (section height) 20 20 Ponding area (km2) 1.91 1.91 Pump stations (number) Permanent 0 2 Temporary 2 0 Total pump capacity (m3/s) 7 40 Working efficiency of pumps (%) Permanent n.a. 90 Temporary 60 n.a. Sluice gates (number) 3 3 Note: n.a. = not applicable. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 50 Detailed Area Modeling • Sediment depth remains 20 percent of the diameter of the drainage pipes and 20 percent of the section height of box-culverts. • Wastewater volume will be insignificant compared to runoff volume and will not enter the drainage system. • The initial water level at Hatir Jheel, Banani Lake, and Gulshan Lake is consid- ered as 5 m PWD. Boundary Conditions Outside river-water levels were not used as external boundary conditions since they do not influence flooding, given that sluice gates remain closed for all sce- narios. An inflow boundary was used at the upstream end of the Panthapath Box-Culvert to represent the contribution from a small portion of ward 49 from the Kallyanpur model. Kallyanpur System Model Setup The MIKE Urban platform was selected for the Kallyanpur system model, which covered about 27.8 km2. Kallyanpur is a densely populated, urban area on the west side of Dhaka City, situated along the Turag River. The river’s Western Embankment (left bank) protects Kallyanpur from river flooding. The model area is adjoined by three models: Goranchatbari to the north, Central Dhaka to the east, and Old Dhaka to the south (map 3.10). Future Scenario: Socioeconomic Changes A key contributor to Dhaka’s economy, Kallyanpur has steadily increasing ­ population densities in both high-density (e.g., Lalbagh) and low-density (e.g., Cantonment) thanas. The area’s overall population density is projected to reach 92,946 per km2 by 2050, nearly twice the 2011 level (IWM 2012). The Kallyanpur system model features generally flat terrain, consisting mainly of low-lying areas along the Western Embankment. In 2004, about 93 percent of the area was above 6 m PWD in elevation. By 2050, it is expected that all land below 6 m PWD will have reached a uniform level above 7.96 m PWD on average, with only designated water bodies remaining in low-lying areas (map 3.11). In 2004, the following categories comprised about 90 percent of land use: residential, manufacturing and processing activity, restricted area (Old Airport and Peelkhana Cantonment), transportation network, education and research, and water bodies. By 2050, more than three-fifths of land will have been zoned for either residential use or the transportation network. Vacant land and water bodies are expected to increase, mainly replacing restricted and agricultural areas. Catchment impervious areas are projected to grow by 15 percent (from 51 per- cent in 2004 to 66 percent in 2050) (map 3.12). Drainage System: Base Scenario and Planned Improvements The Kallyanpur system model’s drainage system consists mainly of pipes, box-culverts, khals, and lakes. The main khals are Ramchandrapur, Katasur, and ­ Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Detailed Area Modeling 51 Map 3.10 Physical Setting of Kallyanpur System Model with Adjoining Boundaries Source: IWM 2014. various branches of Kallyanpur in the northern part of system. Most of the drains and khals carry stormwater to the Kallyanpur Retention Pond, a one-km2 (100 ha) area located in the northwest part of the system (photo 3.4).5 Stormwater is pumped from the pond to the Turag River via the Western Embankment. For 2004, the system model setup included about 55.5 km of stormwater pipes, drains, and khals (map 3.13). The main features were a permanent pump station at Kallyanpur with a 10 m3/s capacity and 80 percent efficiency and four temporary pump stations (one located at Rayerbazar and one near Sikder Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 52 Detailed Area Modeling Map 3.11 Change in Land Elevation over the Study Period, Kallyanpur System Model Source: IWM 2014. Medical College with respective capacities of 0.14 m3/s and 0.56 m3/s and two in Hazaribag, each with a 0.35 m3/s capacity), all of which had a working ­ efficiency of 60 percent (map 3.13, photo 3.5). For 2050, recent capacity improvements at Kallyanpur Pump Station (to 20 m3/s) were incorporated into the drainage model. Efficiency levels were con- sidered as 80 percent for permanent pumps and 60 percent for temporary pumps. The drainage system also included new pipes, for a total length of about 75 km, and some changes to khal cross-sections. The number of subcatchments increased to 374, 38 more than in 2004 (table 3.6). Other Parameters and Features Sedimentation depth was estimated by IWM via field inspections and discussions with DWASA engineers. For pipes, sedimentation depth was estimated as 20 ­percent of pipe diameter (table 3.6). For the 2004 base scenario, it was assumed that 100 percent of domestic wastewater load enters the stormwater drainage system. However, for all future scenarios, it was assumed that wastewa- ter flow is separate from the stormwater system, as proposed by the Dhaka Sewerage Master Plan (DWASA 2012). Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Detailed Area Modeling 53 Map 3.12 Changes in Proposed Land Use over the Study Period, Kallyanpur System Model Source: IWM 2014. Photo 3.4  Kallyanpur Retention Pond in the dry season Credit: © Subhendu Roy / World Bank. Used with permission. Further permission required for reuse. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 54 Detailed Area Modeling Map 3.13  Kallyanpur System Model Drainage Network and Subcatchments, 2004 Source: IWM 2014. Key Assumptions The following key assumptions were made for the 2050 simulations for the Kallyanpur system model: • All sluice gates, which are located along the Western Embankment, remain closed for all scenarios, depicting the worst possible urban flood situations. • The Western Flood Embankment protects the area from river flooding. • Operating efficiency is 80 percent for permanent pumps at Kallyanpur Station and 60 percent for temporary pumps. • Sediment depth remains 20 percent of the diameter of drainage pipes. • Wastewater volume is insignificant compared to the runoff volume generated by the extreme rainfall events simulated as a result of DWASA’s imple­ mentation of the Sewerage Master Plan for Dhaka (DWASA 2012); thus, Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Detailed Area Modeling 55 Photo 3.5 Outlet of temporary drainage pumps on the Western Embankment near Rayerbazar, Kallyanpur system Credit: © Asif Zaman / World Bank. Used with permission. Further permission required for reuse. Table 3.6  Development Baseline: Key Assumptions Comparison between 2004 and 2050, Kallyanpur System Model Parameter or key feature 2004 2050 2 Population density (persons/km , weighted average) 40,509 92,946 Land level (m PWD, average) 7.26 7.97 Subcatchments (number) 336 374 Imperviousness (%, weighted average) 51 66 Time of concentration (minutes, weighted average) 37 35 Initial loss (m) 0.01 0.01 Reduction factor 0.9 0.9 Total length of pipes, box-culverts, and khals (km) 55.51 74.94 Wastewater (% of total domestic load) 100 0 Sedimentation for pipes (diameter) (%) 20 20 Ponding area (km2) 1 1 Pump stations (number) Permanent 1 1 Temporary 4 4 Total pump capacity (m3/s) 11.4 21.4 Working efficiency of pumps (percent) Permanent 80 80 Temporary 60 60 Sluice gates (number) 5 5 Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 56 Detailed Area Modeling wastewater will not contribute to the stormwater drainage system in the future. • The initial water level for Kallyanpur Retention Pond is considered as 4 m PWD. Boundary Conditions The water levels of outside rivers were not used as external boundaries because they have no influence on flooding, given that all sluice gates remain closed for all scenarios. A small part of ward 49 drains to Panthapath Box-Culvert, which is used as an inflow boundary for the Central Dhaka model. Goranchatbari System Model Setup The MIKE Urban platform was selected for the Goranchatbari system model, and the area modeled spans about 63.8 km2 in northwestern Dhaka. The drain- age system model area shares borders with the adjoining areas of Eastern Dhaka to the east, Central Dhaka to the southeast, and Kallyanpur to the south. The Western Embankment—from the Tongi Rail Station to Kellarmor on the city’s west side—protects the area from flooding of the Turag River (map 3.14). Future Scenario: Socioeconomic Changes Goranchatbari has a steadily increasing population density, which is projected to reach about 92,000 people per km2 by 2050, nearly three-and-one-half times the 2011 density (IWM 2012). The area varies widely socioeconomically. Most ­ residents in planned, densely populated urban areas are engaged in commerce, while those in moderately populated, unplanned areas are involved in manufac- turing and industrial-related activities; many women in unplanned areas work in garment and apparel factories. Goranchatbari lies on a flat alluvial plain. In 2004, approximately three- fifths of the total study area was above 6 m PWD, while about one-fifth was above 8 m PWD. Two-fifths was in a range of 6–8 m PWD. The northwestern part of the study area comprises quite low-lying areas, including the Goranchatbari Retention Pond, situated along the left bank of the Turag River. Many low-lying areas have been raised and are being developed for urban resi- dential use. By 2050, all land below 6 m PWD will have been raised to a uni- form level above 7 m PWD, with an average level of 7.68 m PWD (map 3.15). Only designated water bodies will remain in low-lying areas. By 2050, it is expected that 66 percent of the study area will be impervious, compared to 51 percent in 2004. In 2004, the vast majority of the study area (about 91 percent) was dedicated to four land-use categories: residential (40.11 percent), vacant land (21.53 per- cent), restricted area (15.17 percent), and water bodies (14.44 percent). The remaining areas functioned as transportation network, agricultural and educa- tional areas, and recreational zones, with a small portion designated as a mixed- use zone; few areas were zoned for commercial, manufacturing and processing, and service activities (map 3.16). Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Detailed Area Modeling 57 Map 3.14 Physical Setting of Goranchatbari System Model with Adjoining Boundaries Source: IWM 2014. By 2050, nearly 59 percent of the land will be zoned for urban residential use, while the transportation network will cover more than 7 percent. Vacant land and water bodies are expected to increase, primarily at the expense of restricted and agricultural areas (map 3.16). Drainage System: Base Scenario and Planned Improvements Goranchatbari has a predominantly natural drainage system, featuring khals that carry stormwater to the Goranchatbari Pump Station, where it is pumped Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 58 Detailed Area Modeling Map 3.15 Change in Land Elevation over the Study Period, Goranchatbari System Model Source: IWM 2014. to the Turag River via the Western Embankment. In 2004, the 2.43 km2 retention pond area had a total pump capacity of 22 m3/s, consisting of three ­ pumps, each with a 7.33 m3/s capacity. Based on observation of the pump log- book at the site, all pumps were run simultaneously during various extreme events. Currently, construction is under way to double the station’s total pump- ing capacity to 44 m3/s, with the addition of three pumps of equal capacity. Start-and-stop levels were considered as 4 m and 3.5 m PWD, respectively (IWFM 2005). For both 2004 and future scenarios, all pumps were considered to have an efficiency level of 85 percent. In the northeastern part of the study area, some stormwater also drained to Tongi Khal through sluice gates. For all model simulations, these gates were kept closed due to high river levels, representing the worst possible case for urban flooding. The 2004 base scenario included about 44.6 km of drainage khals and storm sewers (map 3.17). The model area was divided into 134 subcatchments, which were delineated according to the distribution of drainage networks and land terrain. By 2050, subcatchments will be split further due to the addition of drain- age pipes or changes in land terrain, resulting in a total of 244 subcatchments, Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Detailed Area Modeling 59 Map 3.16 Changes in Proposed Land Use over the Study Period, Goranchatbari System Model Source: IWM 2014. 110 more than in 2004. An additional 21.8 km of pipe networks were included as planned improvements, based on reports from DWASA and RAJUK, giving a total length of approximately 66 km. All pipes and canals were assumed to drain into the Turag River via the Goranchatbari Pump Station and Tongi Khal through sluice gates. The current capacity improvements at the Goranchatbari Pump Station were incorporated into the drainage model (table 3.7). Other Parameters and Features Because the Goranchatbari study area consists of a relatively new storm-sewer network, it was assumed that sedimentation in those pipes is not prominent; thus, all pipes were considered free of sediment for all scenarios (table 3.7). For the 2004 model setup, it was assumed that 100 percent of domestic wastewater load enters the stormwater drainage system. However, for all future scenarios, it was assumed that wastewater flow is separate from the stormwater system, as proposed by the Dhaka Sewerage Master Plan (DWASA 2012), and thus will not contribute to the drainage system (photo 3.6). Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 60 Detailed Area Modeling Map 3.17  Goranchatbari System Model Drainage Network and Subcatchments, 2004 Source: IWM 2014. Key Assumptions The following key assumptions were made for the 2050 model simulations for the Goranchatbari system: • There are no water inflows into the Goranchatbari area from other drainage areas. • All stormwater drains out to the Turag River and Tongi Khal through the ­drainage networks. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Detailed Area Modeling 61 Table 3.7  Development Baseline: Key Assumptions Comparison between 2004 and 2050, Goranchatbari System Model Parameter or key feature 2004 2050 Population density (persons/km2, weighted average) 20,066 88,776 Land level (m PWD, average) 6.19 7.68 Subcatchments (number) 134 244 Imperviousness (%, weighted average) 42 62 Time of concentration (minutes, weighted average) 85 68 Initial loss (m) 0.01 0.01 Reduction factor 0.9 0.9 Total length of pipes, box-culverts, and khals (km) 44.6 66.4 Wastewater (% of total domestic load) 100 0 Sedimentation for pipes (diameter) (%) 0 0 Ponding area (km2) 2.43 2.43 Pumps (number) Permanent 3 6 Temporary 0 0 Total pump capacity (m3/s) 22 44 Working efficiency of pumps (%) Permanent 85 90 Sluice gates (number) 4 4 Photo 3.6  Water inlet after garbage filtering at Goranchatbari Pump Station Credit: © Subhendu Roy / World Bank. Used with permission. Further permission required for reuse. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 62 Detailed Area Modeling • All sluice gates remain closed during the monsoon season. • The Western Embankment protects the area from river flooding. • The operating efficiency of permanent pumps at the Goranchatbari Station is 90 percent (compared to 85 percent in the 2004 setup). • Sedimentation in drainage pipes is minimal to none. • Wastewater volume is insignificant compared to runoff volume. • Initial water level at Goranchatbari Retention Pond is 4 m PWD. Boundary Conditions No upstream boundaries were used since there are no external inflows into the Goranchatbari system. For the downstream external boundary, the river water level was not used since it will not influence flooding, given that sluice gates remain closed for all model scenarios. The water level of the Goranchatbari Pump Station, 4 m PWD, was used as the initial condition for all scenarios. Eastern Dhaka Model Setup The MIKE 11 platform was selected for the Eastern Dhaka model. This part of the city still remains unprotected from river flooding caused by overflow of the Balu and Lakhya rivers. Natural khals that fall into these two rivers comprise Eastern Dhaka’s primary drainage system. The 121 km2 area modeled features a mix of urban and rural landscapes interspersed with numerous wetlands and khals (photo 3.7). It is bounded by Tongi Khal to the north, the Balu River to the east, Photo 3.7 Tin-shed house built on stilts in low-lying area of Khilgaon, Eastern Dhaka Credit: © Asif Zaman / World Bank. Used with permission. Further permission required for reuse. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Detailed Area Modeling 63 Map 3.18 Physical Setting of Eastern Dhaka Model with Adjoining Boundaries Source: IWM 2014. Demra Road to the south (which forms a boundary with the DND area), and Pragati Sarani—the high ridge that passes through the central part of the city—to the west. This western boundary forms the Eastern Embankment, and protects Western Dhaka from river flooding (map 3.18). Future Scenario: Socioeconomic Changes Eastern Dhaka has a steadily rising population density, estimated at about 17,000 people per km2 in 2011. By 2050, it is expected to reach nearly 81,000 people per km2 (IWM 2012). Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 64 Detailed Area Modeling Map 3.19 Change in Land Elevation over the Study Period, Eastern Dhaka Model Source: IWM 2014. Eastern Dhaka’s land elevation is mainly within a 2–5 m PWD range, reaching up to 12 m PWD and as low as 0 m PWD near the Balu River. The development of low-lying areas for urban residential use is under way. Owing to land filling to accommodate increasing urbanization and population, all lower-elevation areas are expected to be raised up to 7 m PWD by 2050, with only designated water bodies remaining as low-lying areas. Impervious land is expected to reach about 71 percent by 2050, up from 19 percent in 2004 (map 3.19). Major land-use categories in 2004 consisted of agriculture (34.4 percent) and residential use (34.2 percent), followed by vacant land (14.0 percent) and water bodies (13.2 percent) (RAJUK 2010). By 2050, the residential category is expected to account for 58.28 percent of land use, and residential/commercial mixed use area will comprise 8.11 percent. The water retention area will cover 11.99 percent, while the proposed road network will comprise 8.53 percent (map 3.20). Drainage System: Base Scenario and Planned Improvements Khals are the main component of Eastern Dhaka’s drainage system. Major khals are Babur, Barail, Begunbari, Manda, Mangakhali, Satarkul, and Senoti. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Detailed Area Modeling 65 Map 3.20 Changes in Proposed Land Use over the Study Period, Eastern Dhaka Model Source: IWM 2014. The Senoti Khal receives drainage water from Dhaka International Airport and the Nikunja residential area via sluice gates, while the Begunbari and Manda khals receive drainage water from Central Dhaka through sluice gates/pumps at Rampura and Maniknagar, respectively. The outfalls of the khals are mainly the Balu River; however, some also discharge into Tongi Khal. A bypass expected to be completed by 2025 will function as an embankment along the right bank of the Balu River and Tongi Khal. A dedicated sewerage system is planned for Eastern Dhaka under DWASA’s 2012 Sewerage Master Plan (DWASA 2012). Eastern Dhaka is divided into three large subcatchments (map 3.21). In the 2004 model setup, the total river and khal length was 164 km. The existing cross- section data for the khals was collected from the 2004 IWM survey and DWASA drawings. The Balu River is the main outfall for these khals. Some khals also discharge into Tongi Khal. The Eastern Dhaka Bypass, expected to be completed by 2025, is designed to serve as an embankment along the right side of the Balu River and Tongi Khal. The parameters for the project’s regulators and pump stations to control urban drainage congestion and river flooding have been adopted from a separate 2006 Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 66 Detailed Area Modeling Map 3.21 Eastern Dhaka Drainage Network and Subcatchments, 2004 Source: IWM 2014. modeling study conducted by Halcrow Bangladesh Limited. It is expected that all of the khals are to be resectioned as the design cross-section, according to Halcrow Bangladesh Limited (2006). Other Parameters and Features For future scenarios, it was assumed that the full cross-section, as conceived by Halcrow Bangladesh Limited (2006), will be available for conveying stormwater. It was also assumed that the dedicated sewerage system planned for the study area under DWASA’s Sewerage Master Plan means that wastewater will not contribute to the drainage system in the future (table 3.8). Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Detailed Area Modeling 67 Table 3.8  Development Baseline: Key Assumptions Comparison between 2004 and 2050, Eastern Dhaka Model Parameter or key feature 2004 2050 Population density (persons/km2, weighted average) 12,523 73,516 Land level (m PWD, average) 3.4 7.45 Subcatchments (number) 3 3 Imperviousness (percent, weighted average) 19 71 Time of concentration (minutes, weighted average) 1,200 1,200 Initial loss (m) 0.01 0.01 Reduction factor 0.6 0.8 Total length of pipes, box-culverts, and khals (km) 157 157 Wastewater (% of total domestic load) 100 0 Pump stations (number) Permanent 0 3 Temporary 0 0 Total pump capacity (m3/s) 0 245 Sluice gates (number) 0 6 Key Assumptions The following key assumptions were made for the 2050 model simulations for Eastern Dhaka: • All stormwater drains out to the Balu River and Tongi Khal through the ­drainage networks. • The Eastern Dhaka Bypass Project will be completed. • The proposed seven regulators and three pump stations will be operational. • All stormwater drains out to the Balu River and Tongi Khal through the ­drainage networks. • Initial loss and the reduction factor remain the same in the future. • All sluice gates remain closed during the monsoon season. • All khals will be resectioned as the design cross-sections provided by Halcrow Bangladesh Limited (2006). • Retention ponds and water bodies identified in the DAP will be preserved; thus, catchment areas will not change in the future. • The contribution of wastewater volume to the drainage system is insignificant compared to runoff volume. Boundary Conditions The Eastern Dhaka model has 4 river and 18 khal boundary conditions. In the 2004 model setup, the average monsoon water-level values were used at the river boundaries. Zero discharge was used as the upstream boundary for the khals, except for those receiving water from the Western Dhaka side (i.e., Begunbari and Manda khals). The boundary flows for these khals were obtained from the Central Dhaka model. Table 3.9 shows the distribution of runoff contributions from the three catchments among the khals. ­ Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 68 Detailed Area Modeling Table 3.9 Runoff Contribution into Khals, Eastern Dhaka Model Catchment compartment (number) Contributing area (km2) Khal (name) 1 3.19 Tongi_K 1 2.65 AD3 1 22.70 Senoti 1 10.50 AD8 1 1.00 AD1 1 6.45 Babur 1 2.53 AD4 2 10.03 Satarkul_Barail 2 6.90 Shahzadpur 2 8.41 Mangakhali_Upper 2 1.85 Batara 2 3.11 Mangakhali_Lower 2 2.92 Meruler 2 7.91 Barail 3 0.97 CD9 3 1.57 Banabil 3 2.30 Gazaria 3 15.88 Manda 3 10.01 Joaira 3 5.41 Kajla 3 2.73 Khilgaon_Bashabo 3 37.34 Begunbari Model Calibration The original model of Eastern Dhaka was developed by IWM by truncating the NCRHD model and including the internal khal system as calibrated at Demra. Figure 2.3 (chapter 2) shows the comparison plots of the simulated and observed water levels. The internal khal system has no other water level or discharge data; thus, more detailed calibration is not possible. The results for Demra indicate that the model performance is sufficiently sound for the modeling objective. DND Model Setup The MIKE 11 platform was selected for the Dhaka-Narayanganj-Demra (DND) model. This 59 km2 area, located in southeast Dhaka, is subdivided by numerous khals. The area is bounded by the Dhaka-Chittagong Highway to the north, the Lakhya River to the east, Narayanganj township to the south, and the Buriganga River to the west (map 3.22). Flood embankments along the Lakhya and Buriganga rivers prevent fluvial flooding. In the past, DND land was of compara- tively low value and relatively free of urban flooding and waterlogging. Today, however, waterlogging is a common occurrence during the monsoon season, Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Detailed Area Modeling 69 Map 3.22 Physical Setting of DND Model with Adjoining Boundaries Source: IWM 2014. owing to accelerated unplanned development and encroachment on drainage channels, despite embargos (RAJUK 2010). Future Scenario: Socioeconomic Changes The next few decades will witness a nearly fivefold increase in the DND area’s population density, which is projected to reach 65,144 per km2 by 2050. Among the area’s four thanas, Shyampur and Demra are expected to exhibit the highest and lowest densities, respectively (IWM 2012). Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 70 Detailed Area Modeling Map 3.23 Change in Land Elevation over the Study Period, DND Model Source: IWM 2014. About half of the DND’s land is at an elevation of 3–5 m PWD; another one- fifth is above 5 m PWD, and the remainder is in a range of 1.5–3 m PWD. By 2050, it is expected that most land will be raised above 7 m PWD (map 3.23). Land levels of proposed water bodies, open lands, recreational areas, flood- flow zones, and canals/khals in the DAP will not change in the future. Haphazard urban development without adherence to standards has resulted in waterlogging during the monsoon season, often for prolonged periods. The flood risk is exacerbated by the poor permeability of soils, characterized by plastic silts and clay on a sand overlay. The percentage of impervious land will rise sharply in the future, significantly increasing runoff volume. In 2004, about 64 percent of the DND area was zoned for residential use. By 2050, this percentage will decrease by about 7 percent, owing to the proposed increase in mixed-use areas. Other proposed land-use changes include an increase in road network and a reduction in free-flow zones (e.g., water bodies and water- retention areas) (map 3.24). Drainage System: Base Scenario and Planned Improvements The drainage system includes some 85 km of khals, which carry both stormwater and wastewater toward Shimrail Pump Station, where water is pumped out into Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Detailed Area Modeling 71 Map 3.24 Changes in Proposed Land Use over the Study Period, DND Model Source: IWM 2014. the Lakhya River. Major khals include MDC-02 (15–20 m width, running from Shyampur to Shimrail Pump Station), Pagla (15 m width, running from west to east in the middle of DND), and MDC-01 (10–15 m width, running from south DND toward Shimrail Pump Station). The area has 19 subcatchments (map 3.25). The 2004 urban flood inundated 70 percent of the area, highlighting the urgent need for drainage system improve- ments. Planned drainage improvements include design of cross-sections for khals to increase conveyance capacity and installation of new pump stations at Shimrail, Adamjinagar, Pagla, and Fatulla (BWDB 2010) (table 3.10). In addition, separate drainage and sewerage systems will be developed under DWASA’s 2012 Sewerage Master Plan. Other Parameters and Features For the 2004 model setup, historical khal cross-sections from the BWDB feasibil- ity study were used (BWDB 2010). The proposed cross-sections from the same study were used for the planned improvements. Additional conveyance losses due to sedimentation were not taken into consideration. The estimated 2004 wastewater inflows into the drainage system were based on population estimates and per capita water consumption. Given BWDB’s plans to implement separate Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 72 Detailed Area Modeling Map 3.25  DND Drainage Network and Subcatchments, 2004 Source: IWM 2014. drainage and sewerage systems, wastewater flow into the drainage system was not considered for future scenarios. Key Assumptions The following key assumptions were made for the 2050 DND model simulations: • There are no stormwater inflows into the DND area. • Highways do not act as drainage barriers due to structures allowing water to flow under the roads. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Detailed Area Modeling 73 Table 3.10  Development Baseline: Key Assumptions Comparison between 2004 and 2050, DND Model Parameter or key feature 2004 2050 Imperviousness (%) 39–92 62–96 Time of concentration (minutes) 2,170 2,170 Initial loss (mm) 6 6 Reduction factor 0.9 0.9 Time area curve (no.) 1 1 Total length of khal (km) 85 85 Pump stations (number) 1 4 Total pump capacity (m3/s) 14.51 74.69 Planned drainage pumps Number of Pump capacity Start-and-stop levels Station location pumps (m3/s) (m, PWD) Shimrail 7 38.41 2.5, 2 Adamjinagar 5 30.21 2.5, 2 Pagla 9 2.82 2.6, 2.5 Fatulla 10 3.25 2.6, 2.5 Note: DND = Dhaka-Narayanganj-Demra; PWD = Public Works Department. • Four permanent pump stations and resectioning of khals, as proposed in the BWDB feasibility study, will be implemented (BWDB 2010). • Wastewater does not contribute to the drainage system. • Land levels of the proposed wetland/pond/flood-flow zone in the DAP remain at 2004 levels. • All pumps have an 80-percent operating efficiency. Boundary Conditions The DND model area is bounded by embankments. Upstream of all khals is a dead end; therefore, the MIKE 11 one-dimensional model considers the dis- charge at all upstream boundaries of the drainage khals as zero. Because a canal downstream of Shimrail Pump Station connects the inside khals with Lakhya River, pump discharge at Shimrail Pump Station is the only boundary used for the 2004 simulations (photo 3.8). Although other drainage pumps are used in the DND area, these were considered as insignificant and their operational data for 2004 were not available. Model Calibration The model was calibrated for the 2004 period. Figure 3.1 shows the comparison plots of simulated and observed water levels at Shimrail Pump Station. Comparison of water levels indicated that the model performance was accept- able for the modeling objective, especially during peak monsoon period. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 74 Detailed Area Modeling Photo 3.8 Intake pipes of drainage pumps at Shimrail Pump Station, DND system Credit: © Asif Zaman / World Bank. Used with permission. Further permission required for reuse. Figure 3.1 Comparison of Simulated and Observed Water Levels at Shimrail Pump Station 4.0 3.5 3.0 Simulated Meters (PWD) 2.5 Observed 2.0 1.5 Aug 19, Aug 29, Sept 8, Sept 18, Sept 28, Oct 8, Oct 18, 2004 2004 2004 2004 2004 2004 2004 Source: Zaman 2014. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Detailed Area Modeling 75 Narayanganj Model Setup The model setup for Narayanganj used both the MIKE Urban and MIKE 11 platforms, spanning 12.5 km2 and 9.8 km2, respectively, for a total coverage of 22.3 km2. Situated at the southern edge of Dhaka, the township of Narayanganj is one of Bangladesh’s oldest and largest inland ports. The Lakhya and Dhaleswari rivers form respective natural boundaries on the east and west sides of the study area (map 3.26). These rivers meet near Munshiganj and continue as the Map 3.26 Physical Setting of Narayanganj Model with Adjoining Boundaries Source: IWM 2014. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 76 Detailed Area Modeling Map 3.27 Change in Land Elevation over the Study Period, Narayanganj Model Source: IWM 2014. Dhaleswari River until merging with the Meghna River, which eventually flows into the Bay of Bengal. To the north, the Narayanganj Highway separates Narayanganj from the DND area. The Narayanganj-Munshiganj Highway func- tions as an embankment that protects some of the western portion of the area from flooding by the Dhaleswari River. About 12 km2 suffers mainly from urban flooding, while the other 10 km2 is affected by river flooding. Future Scenario: Socioeconomic Changes Narayanganj features both densely and sparsely populated areas at varying land elevations. The overall population density, estimated at 94,000 per km2 in 2011, is steadily increasing and will likely reach 129,000 per km2 by 2050 (IWM 2012). Owing to shortage of space for construction, multistoried buildings are on the increase in the core area. Land elevation averages about 6.4 m PWD in built-up areas and 4.9 m PWD in outlying ones. Planned land-use changes include development of some resi- dential areas into mixed-use zones and protection of major canals and large ponds. It is expected that low-lying areas and small ponds will be filled and con- verted into urban residential zones. By 2050, the land will be raised to an average elevation of 7.54 m PWD (map 3.27). Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Detailed Area Modeling 77 Map 3.28 Changes in Proposed Land Use over the Study Period, Narayanganj Model Source: IWM 2014. In 2004, the vast majority of land area in Narayanganj was zoned for residen- tial use, manufacturing and processing activities, and commercial use, as well as education and research and vacant land. The remaining land was classified as government services, restricted, and service activity areas (map 3.28). By 2050, changes in land-use patterns will be pronounced, particularly in the western portion of the study area, which is considered a flood-flow zone (map 3.28). For example, some residential zones alongside the Lakhya River near Gonje Ali Khal and B. B. Road and near Hariharpar by the Buriganga River are planned for mixed commercial and residential use. Drainage System: Base Scenario and Planned Improvements The drainage system of Narayanganj comprises covered street drains considered as box-culverts, a piped network, khals, and large ponds. A rail line and Bangabandhu Road divide the system into two parts, which drain eastward to the Lakhya River and westward to the Dhaleswari River. Several khals—Baburail, Boalia, and Gonje Ali—convey excess water to these rivers, and some large ponds serve as retention ponds. The flow from these water bodies to the river is unregu- lated and works as gravity drainage. Because most of the area is at a high eleva- tion, there are no sluice gates at drainage outfalls (photo 3.9). Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 78 Detailed Area Modeling Photo 3.9  Drainage outfalls to Lakhya River in Narayanganj township at end of RK Mitra Road, eastern area (above) and end of BK Road, southeastern area (below) Credit: © Asif Zaman / World Bank. Used with permission. Further permission required for reuse. The 2004 base scenario included about 1.7 km of stormwater pipes, 40 km of box-culverts, and 4.6 km of open canals. An additional 930 m of covered street drains (box-culverts) will be laid near the Bangladesh Small and Cottage Industries Corporation (BSCIC) area to drain excess rainfall runoff to the BSCIC Khal (map 3.29). Other Parameters and Features The modeled area was divided into 547 subcatchments, delineated according to the distribution of drainage networks and land terrain. No sediment depth was considered in the simulations since, according to officials of the Narayanganj City Corporation (NCC), siltation is minimal since most networks are relatively new and regular cleaning and maintenance are observed. The 2004 model setup included 100 percent of total generated wastewater since no dedicated disposal system had been implemented at that time. Given that DWASA’s 2012 Sewerage Master Plan includes a dedicated sewerage system for the study area (DWASA 2012), it was assumed that no wastewater will contribute to the drainage system for future scenarios (table 3.11). Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Detailed Area Modeling 79 Map 3.29  Drainage System of Narayanganj, 2004 Source: IWM 2014. Key Assumptions The following key assumptions were made for the 2050 model simulations for Narayanganj: • Stormwater drains from part of Narayanganj township to Lakhya River through several outlets along the river bank and, from the other part, to Dhaleswari River through several canals. • The railway track functions as drainage boundary. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 80 Detailed Area Modeling Table 3.11  Development Baseline: Key Assumptions Comparison between 2004 and 2050, Narayanganj Model Parameter or key feature 2004 2050 Population density (persons/km2, weighted average) 92,000 129,000 Land level (m PWD, average) 7.03 7.54 Subcatchments (number) 547 547 Imperviousness (%, weighted average) 26 55 Time of concentration (minutes, weighted average) 32 28 Initial loss (m) 0.01 0.01 Reduction factor 0.9 0.9 Total length of pipes, box-culverts, and khals (km) 42.8 43.6 Wastewater (percent of total domestic load) 100 0 Sedimentation of pipes (%) 0 0 Ponding area (km2) 2.11 2.11 Pumps (number) 0 0 Note: PWD = Public Works Department. • The drainage system is based on gravity; thus, there are no pumps or sluice gates. • All of the network consists of box-culverts/covered drains, and no sedimenta- tion is considered as the network is relatively new. • The building footprints are elevated to represent actual flood-flow path. • A network will be constructed along the road near the Panir Tanki area to convey stormwater to the nearest canal, which currently drains to low land being filled, causing drainage congestion in that area. Boundary Conditions Because there are no sluice gates at the outfalls, the water levels of the Shitalakhya and Dhaleswari rivers govern the drainage system. Therefore, the boundary con- ditions were considered as river water levels at the outfalls, which were extracted using the NCRHD model. A local channel situated in the southern part of the Narayanganj model area connects the Shitalakhya and Dhaleswari rivers. The water level governs the direction of flow of that channel and subsequently the drainage system of the southern part of the modeled area. Notes 1. Because insufficient data were available on the drainage infrastructure of Savar, Gazipur, and Tongi, rigorous urban-flood modeling could not be undertaken for these townships; thus, they were omitted from the detailed area study. 2. The typical imperviousness values were estimated from analyses of aerial images and Geographic Information System datasets of existing land uses in various parts of Dhaka with different population densities. The imperviousness of a water body that does not contribute to stormwater runoff into the drainage system is considered quite low. However, the imperviousness is quite high for a water body that acts like a Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Detailed Area Modeling 81 retention pond, meaning that all rainfall on the water-body area contributes to the drainage system flow (i.e., linked to outfalls via sluice gates or pumps). 3. These embankments have been designed for 100-year, return period events; they have never been overtopped in the past, and the modeling exercise assumed this will still hold true in the future. 4. Start-and-stop levels for additional improvements were considered as 4.25 m and 4.0 m, respectively. 5. The Kallyanpur Retention Pond area is threatened by encroachers. References BWDB (Bangladesh Water Development Board). 2010. Feasibility Study on Drainage Improvement of Dhaka-Narayanganj-Demra (DND) Project. Dhaka: Bangladesh Water Development Board. DWASA (Dhaka Water Supply and Sewerage Authority). 2012. “Dhaka Sewerage Master Plan.” Prepared by Grontmij and Institute of Water Modelling for Dhaka Water Supply and Sewerage Authority, Dhaka. Government of Bangladesh. 1979. Survey of Bangladesh. Topographic maps of Bangladesh, compiled between 1952 and 1964. Dhaka: Government of Bangladesh. Halcrow Bangladesh Limited. 2006. Updating/Upgrading the Feasibility Study of Dhaka Integrated Flood Control Embankment cum Eastern Bypass Road Multipurpose Project. Dhaka: Halcrow Bangladesh Limited. IWFM (Institute of Water and Flood Management). 2005. Investigation of Hydrological Aspects of 2004 Flood, with Special Emphasis on Dhaka City. Dhaka: Institute of Water and Flood Management, Bangladesh University of Engineering and Technology. IWM (Institute of Water Modelling). 2012. Water Supply Master Plan for Dhaka City. Dhaka: Institute of Water Modelling. Jahan, Sarwar. 2014. Climate and Disaster Resilience of Greater Dhaka Area: A Micro-Level Analysis. Background report prepared for this study, World Bank, Dhaka. JICA (Japan International Cooperation Agency). 1987. Study on Storm Water Drainage System Improvement Project in Dhaka City. Tokyo: JICA. ———. 2000. Baseline Information Study of the Dhaka Combined Flood Control cum Eastern Bypass Road Project: Final Report. Dhaka: Bangladesh Water Development Board, Ministry of Water Resources, Government of Bangladesh. RAJUK (Rajdhani Unnayan Kartripakkha). 2008. Integrated Development of Hatir Jheel Area. Dhaka: Rajdhani Unnayan Kartripakkha. ———. 2010. Preparation of Detailed Area Plan for Dhaka Metropolitan Development Planning (DMDP) Area, GROUP B. Final plan, prepared by Engineering and Planning Consultants Ltd. for Rajdhani Unnayan Kartripakkha, Dhaka. Sugiyama, M. 2012. “Study on Future Climate Scenarios for Asian Cities.” Prepared for the World Bank. Zaman, Asif M. 2014. Urban Flooding of Greater Dhaka Area in a Changing Climate: Vulnerability, Adaptation and Potential Costs. Background report prepared for this study, Bangladesh Climate Change Resilience Fund and Institute of Water Modelling, Dhaka. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Chapter 4 Vulnerability Assessment Introduction The hydrological modeling method, discussed in chapters 2 and 3, enables the estimation and assessment of the depth, duration, and extent of flooding in Dhaka due to extreme weather events, along with the additional flooding that may arise from the effects of climate change (Zaman 2014). While hydrological modeling results provide valuable information on the depth, extent, and duration of flooding due to extreme rainfall events, they do not offer a complete picture of how the various areas of Dhaka may be affected. The impact of inundation of even the same depth and duration on the population and infrastructure will dif- fer significantly, depending on an affected area’s vulnerability, determined by its physical, social, economic, and institutional characteristics. Therefore, estimating overall vulnerability requires a more detailed vulnerability impact assessment, taking into account all other factors that, when combined with flooding, may exacerbate that impact. This study conducted a flood vulnerability assessment to recognize what appropriate actions can be taken to reduce vulnerability before possible harm is realized (Roy 2014). Key Concepts and Approach In the scientific literature, the concept of vulnerability analysis covers a variety of dimensions, including natural, physical, social, economic, and institutional vulnera- bility. But the relative emphasis on these sectors of vulnerability can differ, depend- ing on the nature of the cause that increases vulnerability. In the context of flood vulnerability, it is important to understand what can be used to define it. For the purposes of this study, the approach adopted by the United Nations Educational, Scientific and Cultural Organization (UNESCO) for assessing flood vulnerability was followed (UNESCO 2012). This approach focuses not only on factors that aggravate vulnerability, but also on those that can help us understand the best ways to reduce it. According to this approach, vulnerability is considered as the extent of harm that can be expected under certain conditions of exposure, susceptibility, and resilience. The first two factors heighten vulnerability while the third lessens it. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7   83   84 Vulnerability Assessment Exposure In this analysis, exposure is defined as the predisposition of a system to be ­ disrupted by a flooding event due to its location (UNESCO 2012). The expo- sure arises from what elements (e.g., people, property, goods and other infra- structure) are present at the location that may be impacted by floods that may occur. To capture this aspect, the indicators for this component must include physical aspects that take into account the exposure of elements at risk and also the general characteristics of the flood. In terms of the physical dimension, examples of vulnerability exposure include topography, geology, and proximity to river. Other examples include population density and population below the poverty line (social dimension), land use and percent urbanized area (eco- nomic dimension), and types of vegetation and degraded land (environmental dimension). Susceptibility Susceptibility is generally defined as the elements exposed within the system, which influence the probabilities of being harmed at times of hazardous floods (UNESCO 2012). For evaluating susceptibility, it is important to examine physi- cal aspects at the location, as well as social and economic factors among the population at risk. The focus here is not only on the advance preparation that exists in the area to face the flood risk but also on awareness and preparedness of people regarding the risk they live with. It also includes the presence of institu- tions, mechanisms, and systems that can minimize the impact caused during the floods. Examples of susceptibility include dams and storage capacity and roads (physical dimension), warning systems and evacuation routes (social dimension), unemployment and inequality (economic dimension), and nature reservations and human health (environmental dimension). Resilience Resilience refers to the capacity of a system to endure any perturbation, like floods, maintaining significant levels of efficiency in its social, economic, environ- mental, and physical components (UNESCO 2012). Resilience to flood damage must be considered both during and after floods. The social, economic, and insti- tutional factors in the location can determine how quickly the affected area can recover from the flood impact. Examples of resilience include building codes (physical dimension), literary rates and public awareness (social dimension), investment in counter-measures and flood insurance (economic dimension), and flood recovery time (environmental dimension). Development and Application of the FVI Capturing these aspects of exposure, susceptibility, and resilience requires ­ appropriate indicators that account for the various dimensions involved. To develop a composite Flood Vulnerability Index (FVI) based on these dimensions, this study modified the Climate Disaster Resilience Index (CDRI), a planning Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Vulnerability Assessment 85 tool developed by the Climate and Disaster Resilience Initiative of Kyoto University (Joerin and Shaw 2011) (box 4.1).1 The CDRI is a widely used indicator for examining the vulnerability of an area facing the effects arising out of climate change. Because this study focused on flood vulnerability only and not vulnerability from all natural disasters, only the four CDRI dimensions that contribute to flood vulnerability (physical, social, economic, and institutional) were used; the fifth CDRI dimension (natural) was replaced with a flood dimension. Each CDRI dimension used has five parame- ters, which, in turn, have five variables each. The selected parameters and vari- ables adequately reflect those likely to be most affected by the effects of climate change. The flood dimension has two parameters, each of which is defined by one variable (table 4.1). These vulnerability indicators can provide additional information for setting more precise and quantitative targets for vulnerability reduction and facilitate analysis of the relative conditions in the flood-affected area. Identifying areas with high flood vulnerability can guide the decision-making process toward help- ing societies deal with floods in a better way. In short, FVI is a powerful tool that policymakers can utilize to prioritize investments and make the decision-making process more transparent. Box 4.1 CDRI: Assessing Adaptive Capacity at the Local Level Traditional frameworks for assessing local-level adaptive capacity have focused mainly on asset-oriented approaches. While useful in understanding a system’s capacity to cope with and adapt to changing environments, such approaches overlook the processes and functions that can enhance its adaptive capacity. Dhaka City Corporations (DCCs) are divided into numerous wards that vary significantly in terms of physical features, land development levels, land-use patterns, socioeconomic features, and exposure to various types of hazards. These differences underscore the need for a multidimensional approach that considers micro-level variations in the city’s adaptive capacity to climate change. The Climate Disaster Resilience Index (CDRI) is a useful tool for filling this gap. To compute Dhaka’s CDRI, a micro-level analysis was conducted for this study (Jahan 2014). The CDRI’s 125 variables (five dimensions × five parameters × five variables) were used to develop a survey questionnaire to assess the local situations of wards and thanas in the seven drainage zones of the detailed study area. Respondents used a five-point rating scale (from lowest to highest) to assess the status of each variable. In addition, they were asked to assign weights to the variables and parameters to reflect the relevance of the indicators to their local situations. Using data collected from the questionnaires, a weighted-mean-index method was used to compute the scores for each parameter. The resulting CDRI is the simple average of the indexes for the five dimensions (physical, social, economic, institutional, and natural). The index ranges from 1 to 5, with higher CDRI values equal to higher preparedness to cope with climate change and disasters. Source: Jahan 2014. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 86 Vulnerability Assessment Table 4.1 Indicators Used to Develop the Flood Vulnerability Index Dimension Parameters Variables Physical Electricity Access, availability, supply capacity, dependence on external supply, and alternative capacity Water Access, availability, supply capacity, dependence on external supply, and alternative capacity Sanitation and Access to sanitation, toilets, waste collection, waste treatment, and recycling solid-waste disposal Road accessibility Percentage of land transportation network, paved roads, accessibility during flooding, status of interruption after intense rainfall, and roadside-covered drain Housing and Building code, buildings with nonpermanent structure, buildings above land use waterlogging, ownership, and population living close to polluted industries Social Population Population growth, population under 14 years of age and above 64 years old, population of informal settlers, and population density Health Population suffering from waterborne/vectorborne diseases, population suffering from waterborne diseases after a disaster, access to primary health facilities, capacity of health facilities during a disaster, and disaster preparedness Education and Literacy rate, population’s awareness of disasters, availability of public-awareness awareness programs/disaster drills, Internet access, and functionality of schools after disaster Social capital Population’s participation in community activities and clubs, acceptance level of community leader (in ward), ability of communities to build consensus and participate in city’s decision-making process (level of democracy), and mixing and interlinking of social class Community Preparedness (logistics, materials, and management), provision of shelter for preparedness affected people, support from NGOs/CBOs, population evacuating voluntarily, during a disaster and population participating in relief works Economic Income Population below poverty line, number of income sources per household, income derived from informal sector, income disparity, and percentage of households with reduced income due to a disaster Employment Formal sector: percentage unemployed labor, unemployed youth, employed women, employees who come from outside the city, and child labor in city Household assets Households with television, mobile phone, motorized vehicle, non-motorized vehicle, and basic furniture Finance and savings Availability of credit facility to prevent disaster, credit access, credit access for urban poor, household savings practice, and household’s insured properties Budget and subsidy City’s annual budget for DRR and CCA, availability of subsidies/incentives for residents to rebuild houses, alternative livelihood, and health care after a disaster Institutional Mainstreaming of Mainstreaming of DRR and CCA in city development plans, housing and disaster risk transport policies; ability (manpower) and technical capacity to reduction (DRR) and produce development plans, extent of community participation in climate change development-plan preparation process, and implementation of disaster adaptation (CCA) management plan Effectiveness of city’s Existence of disaster management plan, existence and effectiveness of an crisis management emergency team during a disaster: leadership, availability of evacuation framework centers, efficiency of trained emergency workers during a disaster, and existence of alternative decision-making personnel table continues next page Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Vulnerability Assessment 87 Table 4.1  Indicators Used to Develop the Flood Vulnerability Index (continued) Dimension Parameters Variables Knowledge Effectiveness to learn from previous disasters, availability of disaster-training dissemination and programs for emergency workers, existence of disaster-awareness programs management for communities, capacity (e.g., books and leaflets) to disseminate disaster awareness-raising and education programs, and extent of community satisfaction with disaster awareness programs Institutional City’s dependence on external institutions/support, collaboration and collaboration with interconnectedness with neighboring cities, city’s cooperation (support) with other organizations central municipal department for emergency management, cooperation of and stakeholders city’s ward officials for emergency management, and city’s institutional during a disaster collaboration with NGOs and private organizations Good governance Effectiveness of early warning systems, accountability and transparency of city government, implementation of building codes, existence of disaster drills, promptness of city body to disseminate emergency information during a disaster to communities, and capability of city body to lead recovery process Floodinga Extent Area flooded >0.25 m (percentage) Duration Duration of flood >0.25 m (hours) Source: Adapted from Jahan 2014. a. For the flood dimension, depth of inundation was not used as a separate parameter since depth varies widely in a flooded area and average flood depth is highly correlated with the flood duration parameter. Data Collection and FVI Calculation For the four dimensions of the CDRI, data were collected using a survey ques- tionnaire and secondary sources for each ward and some outlying areas of the detailed study area (Jahan 2014). The survey was conducted among planners involved in the 2009 preparation of the Detailed Area Plan (DAP) for the Dhaka metropolitan area. The planners belonged to various consultancy firms, each of which was responsible for preparing the plan of a particular area of the city. Data on some variables, especially related to physical and social aspects, were also collected from such secondary sources as the 2011 population census and DAP reports. Each dimension’s respective five parameters could take vulnerability values ranging from 0 (indicating high vulnerability) to 5 (indicating low vulnerabil- ity). For the flood dimension, data was obtained from the results generated by the hydrological modeling exercise (Zaman 2014). This information was avail- able for five future scenarios based on alternative assumptions. The flood dimension’s two parameters were measured in percentages and hours, respec- tively. But unlike the CDRI parameter values, higher values indicated greater vulnerability. Since the parameters used in the four CDRI dimensions and flood dimension differed in units of measure, they first had to be normalized—they were assigned values ranging from 0 (indicating high vulnerability) to 1 (indicating low vulnerability)—before being combined into one FVI. For the four dimensions of ­ Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 88 Vulnerability Assessment the CDRI, the normalized value Yij for parameter i in ward j was calculated using the following formula: ( X ij − Min X ij ) Yij = , (4.1) ( Max X ij − Min Xij ) where Xij represents indicator i for ward j. For the normalization, another formula was needed for the flood parameters because a higher value in those parameters indicated greater vulnerability. Thus, to ensure normalization with assigned values between 0 and 1, with indicating high vulnerability and 1 indicating low vulnerability, the following 0 ­ formula was used: ( Max X ij − X ij ) Yij = , (4.2) ( Max X ij − Min X ij ) where Xij represents indicator i for ward j. For each ward in the detailed study area, the vulnerability indexes in each of the five dimensions were separately determined based on data from each ward for the normalized values of the chosen parameters. Thus, Y j k , the vulnerability index of dimension k in ward j, was computed as the weighted sum of all the normalized values of the parameters included in dimension k for that ward, expressed as follows: N Y jk = ∑w Y i =1 i ijk , (4.3) where wi is the weight for indicator i, and N is the number of indicators in the index of type k. The weights were assigned based on the variability of each indi- cator among all wards in Dhaka since a greater variability poses higher risk.2 To ensure that the weights added up to 1, they were defined as follows: 1 w i = k Var (Yi ) where k= N , (4.4) ∑ i =1 Var (Yi ) where Var(Yi  ) is the variance of indicator i normalized values among all of Dhaka’s wards. Finally, the composite FVI was computed as a weighted average by combining the indexes developed for each of the five dimensions. Separate FVIj for ward j is created for each of the five flooding scenarios as follows: 5 FVI i = ∑ w Y , (4.5) k =1 k jk where wk represents the weight for dimension k. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Vulnerability Assessment 89 The FVI computation is done under two alternative assumptions about the importance of flooding parameter in the composite FVI index. The first assumes equal weight for each of the five dimensions and thus wk = 0.2 for all five dimen- sions. Under the second assumption, the weight of the combined four CDRI dimensions is considered the same as the weight for the flood dimension. Since all CDRI dimensions are given the same weights, wk = 0.125 for each of those four dimensions and wk = 0.5 for the flood dimension. The second assumption thus assigns a much greater weightage to flooding effects. In all, 10 FVIs are generated for each study area, one for each of the five scenarios under the two alternative assumptions about the importance of flooding in the FVI. Finally, the comparison of vulnerability is done at two area levels: (a) a broad, aggregated study-area level by comparing the FVI among the specific study areas and (b) a ward-wise FVI to compare the relative vulnerability among the various wards in Dhaka.3 The ward-wise comparison allows development of vulnerabil- ity ranking to determine a list of Dhaka’s most vulnerable wards. Notes 1. This study drew specifically on the CDRI analysis conducted by Sarwar Jahan for Dhaka (Jahan 2014). To develop the FVI, the study used a portion of the data col- lected in that study, along with data generated by the hydrological modeling exercise (Zaman 2014). 2. A greater variability of data means a greater uncertainty, and uncertainty directly cor- relates with risk. For example, if two areas have the same average rainfall but ­ different variance in rainfall, then the area with the higher variance is likely to be more vulnerable as the range of rainfall will be much wider. Thus, the risk being faced by the area with higher vulnerability is likely to be higher. The index created is trying to capture that. 3. The vulnerability comparison excludes the Narayanganj area for lack of complete detailed data and difficulty in matching of the areas between CDRI rankings and flooding data. References Jahan, Sarwar. 2014. Climate and Disaster Resilience of Greater Dhaka Area: A Micro-Level Analysis. Background report prepared for this study, World Bank, Dhaka. Joerin, J., and R. Shaw. 2011. “Climate and Disaster Resilience in Cities.” In Community, Environment and Disaster Risk Management, Vol. 6, edited by R. Shaw and A. Sharma. Bingley, U.K.: Emerald Group Publishing Ltd. Roy, Subhendu. 2014. Urban Flooding of Greater Dhaka Area in a Changing Climate: Vulnerability, Adaptation and Potential Costs. Background report prepared for this study, World Bank, Washington, DC. UNESCO (United Nations Educational, Scientific and Cultural Organization). 2012. “Flood Vulnerability Indices.” http://unescoihefvi.free.fr/vulnerability.php (accessed June 18). Zaman, Asif M. 2014. Urban Flooding of Greater Dhaka Area in a Changing Climate: Vulnerability, Adaptation and Potential Costs. Background report prepared for this study, Bangladesh Climate Change Resilience Fund and Institute of Water Modelling, Dhaka. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Chapter 5 Estimating Flood Damage Introduction Results of the hydrological modeling exercise were also used to quantify the economic damage—effects on the urban-built environment and infrastructure, human health, and family assets—caused by flooding of the Greater Dhaka detailed study area in the current climate.1 The likely increase in damage due to climate change was also separately assessed. Both depth and duration of inunda- tion were analyzed to estimate the extent of flood damage. The study assessed not only property damage, income loss, and other readily quantifiable damages; wherever possible, it assigned monetary values to intangible damages, based on analysis of losses gleaned from comparison with similar ones that are more easily quantified (Roy 2014). Alternative Cases and Assumptions for Future Scenarios The analysis centered on the likely economic damage that may occur in 2050 and cumulatively between 2014 and 2050. For each of the five future scenarios covered in the hydrological modeling exercise (table 2.2), three cases were computed, as follows: ­ • The damage resulting from a 100-year return-period storm occurring in 2050. This is the most severe weather event used in the hydrological modeling exer- cise; thus, the estimate provides an upper limit of the annual damage that may occur in 2050. • A weighted average estimate of damage, based on the probability of each return-period storm from 1 year to 100 years occurring in 2050. The probabil- ity of each return-period event is given by 1/N, where N equals the return period. • The cumulative damage between 2014 and 2050, based on random assign- ment of 1-year to 100-year storms during each year. This estimate helps to determine the expected total damage between 2014 and 2050, using the prob- ability of each outcome. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7   91   92 Estimating Flood Damage The increased precipitation likely to result from climate change effects has already been factored into the hydrological modeling exercise by incorporating the 16 percent expected increase in precipitation from the pattern-scaling exer- cise by Sugiyama (2012) (box 1.2). In a changing climate, the frequency of heavy-rainfall events in South Asia is also increasing (IPCC 2013). In order to account for the increased frequency of more severe storms in estimating eco- nomic damage, a sensitivity analysis was conducted under three assumptions: • Low increased frequency. The 100-year, return-period storm (1 percent ­ probability), would occur once every 50 years (2 percent probability). • Medium increased frequency. The 100-year, return-period storm (1 percent probability) would occur once every 33 years (3.3 percent probability). • High increased frequency. The 100-year, return-period storm (1 percent probability) would occur once every 20 years (5 percent probability). ­ Data to Project Future Outcomes The analysis extrapolated Dhaka’s predicted population growth between 2014 and 2050 based on past decadal growth rates adjusted for likely future changes.2 Expected changes in housing characteristics, roads, and other physical infra- structure were estimated based on existing plans. To assess the damage in real terms, all economic data were based on 2013 prices, thus ignoring any inflation that could occur by 2050. Any increase in per capita income in 2014–50 was excluded to compensate for the need to discount future damage.3 It was assumed that excluding inflation and increased per capita income would obvi- ate the need for converting future damage estimates into present values using a discount factor based on time preference for money. This approach was adopted because of the inherent controversy over the choice of a discount rate in environmental ­economics when examining outcomes long into the future; even catastrophic outcomes show negligible effects when discounted over a long time horizon, and the estimates included here are looking into outcomes well into the future.4 Direct and Indirect Damage Direct damage is primarily concerned with the economic loss resulting from damage to physical capital. The main determinant is depth of inundation; how- ever, the degree of damage can also increase with flood duration of a given depth. Thus, any analysis of direct damage must start with depth of flooding and then be adjusted upward using duration of flooding. Damaged physical capital com- prises both private and public resources. Damaged private resources commonly include residential, commercial, and industrial buildings and property; industrial factories also face damage to machinery and inventory. Other private-property losses include damage to transport vehicles and telecommunication and electric infrastructure. Common examples of publicly-owned damaged resources are Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Estimating Flood Damage 93 roads, rails, ports, water and sewerage facilities, hospitals, and government-owned buildings and property. Indirect damage is mainly concerned with loss of income due to the inability of residents, businesses, and industries to carry out their daily economic activities. The main determinant is duration of flooding above a certain depth, which can lead to disruption of day-to-day activities over a period of time. Most economic losses occur during the flooding, but some may continue after the water recedes. The direct damage of public and private resources may prolong the total or par- tial loss of earning capacity, even after floodwaters recede. Indirect damage also results from direct damage to roads and communication infrastructure, which leads to future disruption of economic activity. Indirect damage to public health arises from added health-care costs and the effects of increased morbidity on livelihoods. It has been found that a water depth level of 0–0.1 m produces little or no damage. In all of the affected areas, people have learned to adapt to such levels of flooding as a common annual occurrence. With the exception of Central Dhaka, a water depth level of 0.1–0.25 m produces slight damage in a few areas. Again, such flooding tends to be perceived as an annual feature during the rainy season. However, with water depths above 0.25 m, some level of economic damage is unavoidable. The extent of direct damage depends mainly on the flooding depth. At a water depth range of 0.25–0.75 m, the extent of damage is low for all areas. At 0.75–1.50 m, all areas experience moderate damage. Above 1.50 m, damage is extensive in all areas and continues to rise the longer flooding persists. To examine the damage resulting from duration of flooding, the study analyzed the extent of flooding above 0.25 m for up to 10 days. The selection of ­ 10 days as the cut-off was based on the finding that floodwater recedes from most areas of Dhaka within that period, even in the worst case (i.e., A1FI sce- nario for the 100-year, return-period storm). Indirect damage depends mainly on how long flooding exceeds the critical 0.25 m depth level; however, it may be exacerbated by higher levels of inundation since indirect damage may arise as the result of loss of physical capital and other resources resulting from direct damage from a given water depth. Key Sectors The sectors most affected by the economic damage from flooding are residential (buildings and property), commerce and industry, health care, roads and railways, and public transport. The subsections below describe how the direct and indirect damages were assessed for these sectors. Residential Buildings Repair and clean-up costs were used to estimate the direct damage to residential buildings. The study analysis was based on Dhaka’s categorization of building structures (jhupri, kutchha, semi-pucca, and pucca) and floor space.5 Pucca struc- tures were further divided into three broad housing categories: economically Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 94 Estimating Flood Damage weaker sections (EWS), middle-income groups (MIG), and high-income groups (HIG). For multi-storied buildings, it was assumed that damage would be limited to the first floor.6 Based on past experience, the housing composition for 2050 was assessed. It was assumed that building upgrades would result in a large reduc- tion of jhupri and kutchha structures and a gradual transformation of a percent- age of semi-pucca buildings into pucca structures. It was further assumed that the average jhupri, kutchha, or semi-pucca structure is a one-story building, while the average pucca structure consists of two (EWS), three (MIG), or four (HIG) stories (table 5.1). Repair Costs The calculation of repair costs was based on building construction costs and the percentage of construction cost needed for repair (2013 figures). For jhupri, kutchha, semi-pucca, and pucca buildings, repair costs were estimated at 50, 25, 12, and 6 percent, respectively, of building construction costs. For EWS, MIG, and HIG structures, building construction costs per square meter of floor area were estimated at Tk. 8,000, 9,300, and 12,800, respectively.7 The average costs of jhupri, kutchha, and semi-pucca houses were estimated at Tk. 50,000, 90,000, and, 150,000, respectively. Determining the percentage of buildings incurring repair costs was based on damage factors developed using flooding depth and duration for each ward in Dhaka. It was assumed that any inundation greater than 0.25 m depth lasting more than one day would cause a minimum-threshold,8 proportional damage requiring repair that varies by building type. The minimum-damage threshold for EWS, MIG, and HIG housing differs because of the nature of the building and quality of construction materials used (table 5.1). It was also assumed that any building submerged for 10 days or longer at an inundation depth greater than 1.5 m would always require repair. The proportion of buildings needing repair decreases linearly for fewer days of inundation. For depth ranges of 0.1–0.25 m, 0.25–0.75 m, and 0.75–1.5 m, the damage proportions were 25, 50, Table 5.1 Residential Housing Data to Estimate Direct Damage Stories, Building cost Repair costs Minimum- Clean-up costs Floor area, average First floor for all floors per building damage per building Building type average (m2) (no.) area (%) a (million Tk.) (million Tk.) b threshold (%) (million Tk.) b Jhupri Not used 1 100 0.05 0.025 33 0.0005 Kutchha Not used 1 100 0.09 0.022 20 0.0005 Semi-pucca Not used 1 100 0.14 0.017 10 0.0005 Pucca: EWS 25 2 50 0.40 0.012 5 0.001 MIG 75 3 33 2.09 0.042 5 0.002 HIG 150 4 25 7.68 0.115 5 0.004 Source: World Bank 2010. Note: EWS = economically weaker sections; HIG = high-income groups; MIG = middle-income groups. a. This estimate is based on the average number of floors for each building type. b. 2013 figures. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Estimating Flood Damage 95 and 75 percent, respectively, of the proportion for a depth greater than 1.5 m. Since the depth of inundation continues to change as water recedes from a given area, the proportion of damaged building requiring repair (Rf) in an area is the maximum proportion found from the inundation duration for the four depths over a 10-day period, expressed as follows: Rf = Mt + MAX (D1 * 0.25, D2 * 0.5, D3 * 0.75, D4 * 1) * {(1 – Mt /10)}, (5.1) where Mt represents the minimum threshold per building type, and D1, D2, D3, and D4 equal the number of days at inundation depths of 0.1–0.25 m, ­ 0.25–0.75 m, 0.75–1.5 m, and greater than 1.5 m, respectively. Clean-Up Costs The analysis assumed that all inundated buildings not requiring repair would incur clean-up costs. Based on field surveys, the main clean-up components were identified as whitewash, fixed inventories, and labor. Their total costs were estimated at Tk. 500 for jhupri, kutchha, and semi-pucca structures and Tk. 1,000, 2,000, and 4,000, respectively, for EWS, MIG, and HIG buildings. For each building category, the total damage to the residential building (DRB) can be expressed by the following equation: DRB = HH * I * P * S * [(Cb * Rf * Dh) + 1 – Rf) * Cc], (5.2) where HH equals the total number of households in the affected area by ­category, I equals the percentage of area inundated in the affected area, P is the percentage of composition of various categories, S is the proportion of house- holds on the first floor, Cb equals building construction costs, Rf is the propor- tion of damaged buildings requiring repair costs, Dh is the damage factor of a building needing repair (assumed at 0.06), and CC equals clean-up cost. Finally, total costs are extrapolated to 2050 based on the predicted housing structures for that year. Residential Property Sudden and prolonged flooding causes extensive damage to residential property. Submerged vehicles owned by households, situated in ground-floor garages, com- prise the largest component of such direct damage. Other household property, including appliances, electronics, and furniture, can be damaged to varying degrees, especially if they are difficult to move to upper floors. The direct correla- tion between household income and value of household property makes it pos- sible to estimate damage to residential property based on household income for each ward in Dhaka. The analysis extrapolated household-income data to 2050 using an average GDP growth of 4 percent per capita over the period. For the middle three household-income groups, the midpoint in each income range was used as the average income for that group. For the lowest and highest income groups, which are characterized by skewed income distribution, income data was Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 96 Estimating Flood Damage Table 5.2 Household Income Distribution and Affected Property Percent Monthly income group (thousand Tk.) Household factor <5 5–10 10–15 15–20 >20 Savings rate 5 10 15 20 25 Proportion on first floor 100 50 33 33 25 Source: World Bank 2010. derived location-wise to determine the number of households in each income category (table 5.2) (BBS 2011). The value of the property owned depends on the savings accumulated in each income category over a five-year period, with annual savings rate differing by income category. The property-damage factor is based on the depth and duration of flooding. For household property submerged 10 or more days at an inundation depth greater than 1.5 m, with proportional decrease for fewer days of inundation, the assumed maximum damage is 33 percent. For depth ranges of 0.25–0.75 m and 0.75–1.5 m, the corresponding maximum damages are 20 percent and 25 percent, respectively. The property damage factor (DP) can be expressed as follows: DP = MAX (D3 * .033, D2 * 0.025, D1 * 0.02), (5.3) where D3, D2, and D1 represent the number of days of inundation at depths of more than 1.5 m, 0.75–1.5 m, and 0.25–0.75 m, respectively. For each income group, residential property damage (DRP) is expressed as follows: DRP = HH * Y * C * S * DP * I, (5.4) where HH represents the total number of households in each income category in the affected area, Y equals the average income in a household income group in 2050, C equals the savings rate in an income group for five years, S represents the proportion of total households located on the first floor, DP is the property damage factor, and I is the percentage of area inundated in the affected area. Commerce and Industry As mentioned above, both commercial and industrial sectors face direct damage to buildings and property (e.g., inventory); industry also faces direct damage to machinery. In the absence of data identifying business hubs, it was not possible to assess damage to business interruptions. A damage factor was developed, based on the average depth, duration, and extent of flooding in the affected area. The damage factor was applied to the value of the affected holdings to assess such damage, expressed as follows: DXP = V * DC, (5.5) Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Estimating Flood Damage 97 where DXP represents the direct damage, V equals the total value of the holding, and DC is the holding’s property damage factor. The analysis assumed that 50 percent of commercial establishments will be damaged since flooding affects only the first floor. For industries, fixed capital is used to estimate the value of buildings and machinery (BBS 2003). Because ward-wise distribution of commercial and industrial data was unavail- able, the average extent of flooding for the 10 days under various depths for the whole of Dhaka was used to develop the damage factor. The analysis assumed the extent of damage for depth ranges of 0.25–0.75 m, 0.75–1.5 m, and > 1.5 m as 3, 6, and 9 percent of the value of the holdings, respectively. Thus, DC = [0.03 * (FL − FM − FH) + 0.06 * (FM − FH) + 0.09 * FH], (5.6) where DC represents the property damage factor and FL, FM, and FH represent the average proportion of flooding in the Dhaka area over the 10 days for the three depths in increasing order. The damage to property was assessed as the difference in the daily value added for the number of days a business is affected. The number of lost days of business was based on the extent of average flooding and its duration. This indi- rect damage (DXF) is expressed as follows (box 5.1): DXF = N * DT, (5.7) where N equals the value added per day and DT represents the number of lost days. In turn, the number of lost days is expressed as follows: DT = [0.5 * (FL − FM − FH) + 0.67 * (FM − FH) + FH ]/100 * 10, (5.8) where FL, FM, and FH represent the average extent of flooding in the Dhaka area over the 10 days for the three depths in increasing order. Health Care Increased cases of diarrhea and dengue fever constitute most of the economic losses from flooding suffered in the health-care sector. This analysis assumed that, with increased prosperity, current rates of these diseases will likely decline, but that economic impacts will be higher because of the increased value of life. These two factors were assumed to offset each other; thus, the current dam- age valuation was used as the future estimates between 2014 and 2050. The effect of changes caused in the future scenarios was based on the percentage of area flooded. Current damage was used as the baseline, including only planned improvements. Health damage costs were computed using data on the incidence and mortal- ity from diarrhea and dengue fever.9 Productivity loss from morbidity and loss of productive life from mortality were assessed using Disability Adjusted Life Years (DALYs) data (WHO 2009).10 Since urban areas like Dhaka witness about Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 98 Estimating Flood Damage Box 5.1 Calculating Indirect Damage from Income Loss The disruption caused by urban flooding results in income loss not only for Dhaka residents, but also for daily migrant workers who commute to the Dhaka area for their livelihood. To ­estimate the indirect damage from income loss, this analysis extrapolated the population of Dhaka and the number of migrant workers to each year between 2014 and 2050, using an estimated annual population growth rate of 3 percent. For Dhaka residents, income loss was based on household income data since all workers in a household are likely to face similar impact from flooding. The computation was limited to workers in the unorganized sector since (a) most workers in the organized sector are paid on a monthly basis and thus are unaffected by daily disruptions and (b) most of the loss for those affected is captured in the loss in value added under the commerce and industry portion of the analysis. The analysis assumed that all workers in the top monthly income bracket (> Tk. 20,000) are in the organized sector and all workers in the lowest monthly income bracket (< Tk. 5,000) are in the unorganized sector. For the middle three brackets (Tk. 15,000–20,000, Tk. 10,000–15,000, and Tk. 5,000–10,000 per month), 10, 50, and 90 percent, respectively, were assumed to work in the unorganized sector. Income loss was computed based on lost work days multiplied by the average income in each category. The analysis further assumed that 25 percent of migrant workers are in the unorganized sector. Since these workers tend to have lower skills, it was assumed that they earn an average of 33 percent less than the average urban worker in Dhaka. Income loss was computed based on lost work days, developed using the average flooding for the whole of Dhaka and its duration. For each income group, income loss (DI) is expressed as DI = I * DT, where I equals the income per day in each income category and DT equals the number of lost work days due to flooding in each ward for Dhaka residents and the average for the Dhaka area overall for migrant workers. Source: Roy 2014. one-quarter of the overall average for the entire country, the DALY estimates were lowered to 25 percent of the overall average for Bangladesh. Roads and Railway Flooding directly damages roads and railways and also causes indirect damage through disruption of economic activity. For the purposes of this analysis, only the direct damage was estimated, based on repair costs. Depth of flood- ing was a major factor in estimating the extent of road damage. Repair costs were estimated based on the per-meter of inundation approach used by Kok (2001). Accordingly, damage factors of 0.1, 0.225, and 0.4, respectively, were used for water depths of 0.25–0.75 m, 0.75–1.5 m, and > 1.5 m. The maxi- mum extent of inundation in the Dhaka area over the 10 days of flooding Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Estimating Flood Damage 99 under each depth was used to determine the length of damaged roads (LD), expressed as follows: LD = LT * [AH * 0.4 + (AM − AH) * 0.225 + (AL − AM − AH) * 0.1], (5.9) where LT equals the total length of roads and AL, AM, and AH represent the maxi- mum proportions of flooded area over the 10 days under the three respective flood depths in increasing order. The average cost of road repair was assessed at Tk. 24 million per km, based on World Bank estimates (World Bank 2010). For the final analysis, the average road lengths were extrapolated to 2050, and the road repair cost from flooding was then obtained by multiplying the length of damaged roads by the per-­ kilometer road repair cost. Similarly, property damage for railways was estimated based on the length of track in each ward of Dhaka. An average rail repair cost of Tk. 240 million per km was based on World Bank estimates.11 Other Surface Transport The analysis covered the direct damage from flooding to public-transport vehi- cles, including buses, taxis, auto-rickshaws, and goods vehicles. The estimated number of public-transport vehicles in 2050 was extrapolated from current data for such vehicles. Their damage repair costs were obtained using sample surveys from auto-repair shops. The average estimated cost of repairs was in a range of Tk. 10,000–30,000, based on depth of water submersion and vehicular type. To determine the proportion of vehicles damaged, assumptions were made by depth category, as follows: > 1.5 m, 50 percent of vehicles; 0.75–1.5 m, 33 per- cent of vehicles; and 0.25–0.75 m, 25 percent of vehicles. Thus, the transport damage factor (TP) can be expressed as follows: TP = [0.25 * (FL − FM − FH) + 0.33 * (FM − FH) + 0.5 * FH]/100, (5.10) where FL, FM, and FH represent the average extent of flooding in the Dhaka area over the 10 days for the three respective depths in increasing order. Notes 1. Damage estimates are based mainly on the Dhaka Municipal Corporation area (both North and South) owing to the availability of detailed data on its population and physical assets. 2. Details are available at http://www.prb.org/Publications/Articles/2001/Urbanization​ TakesonNewDimensions inAsiasPopulationGiants.aspx. 3. In computing damage, current income/output data was used. In 2050, income/output will certainly be higher (current annual GDP growth in Bangladesh exceeds 6 percent). One reason for discounting future values is to account for a higher future income level in the economy. Since no discounting of future values was done, the income/output data was not extrapolated with estimated growth rates from the present to 2050. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 100 Estimating Flood Damage Thus, instead of trying to separately estimate GDP growth and likely discounting factors (an exercise difficult to do with accuracy) an assumption was made that the ­ two will counterbalance each other. More information is available at http://csc.noaa.gov/archived/coastal/economics​ 4. /­discounting.htm. 5. Jhupri structures are temporary houses made of such removable materials as grass, thatch, bamboo, and wood. Kutcha houses have walls and roofs made of unburnt brick, bamboo, mud, reed, thatch, or loosely packed stones. Semi-pucca houses have fixed walls made of pucca materials (see below), with non-pucca materials used for roofing. Pucca houses have walls and roofs made of such materials as burnt bricks, stones (packed with lime or cement), cement concrete, timber, and ekra. More information is available at http://mospi.nic.in/Mospi_New/upload/statistical_year_book_2011​ /­SECTOR​-4-SERVICE%20SECTOR/CH-28-HOUSING/HOUSING​­-­​WRITEUP.pdf. 6. The 2011 population census was used to estimate the number of houses in each building category. For the pucca subcategories, data was extrapolated from a similar study conducted for Calcutta (Kolkata), which used Google Earth data to separate the composition of city residential buildings into EWS, MIG, and HIG. 7. These figures are based on the ACC Help Home Building Calculator (https://www​ .acchelp.in/cost-calculators.asp). 8. For Central Dhaka, a threshold of 0.1 m was used. 9. Data was provided by the Director for Disease Control, Directorate General of Health Services (DGHS), Ministry of Health and Family Welfare, Mohakhali, Dhaka. 10. The estimates for Southeast Asia were applied to Bangladesh as a whole. 11. Details are available at http://www.chinaafricarealstory.com/2010/08/real-cost-of​ -chinese-railway.html. References BBS (Bangladesh Bureau of Statistics). 2003. Economic Census 2003. Dhaka: Bangladesh Bureau of Statistics. ———. 2011. Population Census. Dhaka: Bangladesh Bureau of Statistics. IPCC (Intergovernmental Panel on Climate Change). 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, edited by T. F. Stocker, D. Qin, G.-K. Plattner, M. Tignor, S. K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex, and P. M. Midgley, sections 14.6., 14.8.9–12. Cambridge, U.K.: Cambridge University Press. Kok, M. 2001. Damage Functions for the Meuse River Floodplain. Internal report. Ispra, Italy: Joint Research Center, European Commission. Roy, Subhendu. 2014. Urban Flooding of Greater Dhaka Area in a Changing Climate: Vulnerability, Adaptation and Potential Costs. Background report prepared for this study, World Bank, Washington, DC. Sugiyama, M. 2012. “Study on Future Climate Scenarios for Asian Cities.” Prepared for the World Bank. WHO (World Health Organization). 2009. Global Burden of Disease and Risk Factors. Geneva: WHO. World Bank. 2010. Climate Change Impact and Adaptation in Kolkata Metropolitan Area. Washington, DC: World Bank. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Pa r t 2 Results Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7   101   Chapter 6 Assessment of Dhaka’s Modeled Zones Introduction Having estimated the regional impacts of climate change, land development, and land-use changes, the impact of urban (as well as river) flooding on the Dhaka detailed study area can be investigated. This chapter is organized into seven major sections corresponding to this study’s modeled drainage zones. Each sec- tion begins by estimating the location-specific inundation depth and duration for 100-year, return-period rainfall with and without climate change. The flood vulnerability indices of constituent wards are then presented, taking into account each ward’s physical, social, economic, and institutional characteristics, together with the inundation depths and duration of the alternative scenarios considered for hydrological modeling with different weights assigned to flooding (­chapter 4). Next, the expected damage and loss from flooding are provided. Finally, the estimated costs of addressing the current adaptation deficit and climate change adaptation are presented and compared with the benefits of making these investments. Old Dhaka Results Hydrological Modeling Output The study generated maps simulating the inundation depth and duration of urban flooding for the historic 100-year, return-period rainfall event of 2004 and alternative scenarios for 2050 for a similar rainfall event with and without cli- mate change. For 2050, alternative scenarios were simulated using flood criteria of 0.1 m and 0.25 m. The maps generated for the 100-year rainfall event in 2050 show that, with improvements already planned, even without climate change, the extent of shal- low flooding (0.1–0.25 m depth) is significantly greater than in 2004. Some deeper pockets of waterlogging (>0.75 m depth) persist, and flooding duration Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7   103   104 Assessment of Dhaka’s Modeled Zones Table 6.1 Comparison of Flooding Results with 0.1 m Flood Criteria, Old Dhaka Model Maximum flooded area (km2) Average duration Extreme storm Flood free above 0.1 m depth (341 mm/24 hour) (0–0.1 m) 0.1–0.25 m 0.25–0.75 m 0.75–1.5 m >1.5 m (hours to recede) 2004 (actual storm) 7.8 1.2 2.0 0.5 0.3 19.2 2050 (SE + PI) 3.7 7.0 0.6 0.3 0.2 20.6 Source: Zaman 2014. Note: SE = socioeconomic changes; PI = planned improvements. Map 6.1 Extreme Rainfall Scenarios in 2050 with 0.25 m Flood Criteria, Old Dhaka Source: IWM 2014. Note: PI = planned improvements; AI = additional investments. increases slightly (table 6.1). The changes in depth and duration are attributable to increases in land elevation and imperviousness. Achieving an inundation-free Old Dhaka, especially the deep pockets, will be a difficult task from a practical perspective. Since Old Dhaka residents have adopted various strategies for coping with periodic episodes of modest flooding, the flood criteria for 0.1 m depth appeared too stringent, as pointed out by local experts. As a planning goal, it was later agreed that flood criteria up to a depth of 0.25 m would be acceptable, and this was the inundation depth later used to simulate scenarios for rainfall extremes (341 mm per 24 hours) in 2050. The maps generated showed that there is an adaptation deficit (map 6.1), meaning that the infrastructure in place or the improvements already planned are insuf- ficient to cope with such extreme events. Thus, even without climate change, additional investments (AI) in the drainage system would be needed to meet acceptable criteria: 90 percent of administrative areas with a maximum flood depth less than 0.25 m and duration less than 12 hours. Table 6.2 compares the simulated results for the extreme rainfall event in 2050 with planned improve- ments (PI) only, as well as with the incorporation of AI required to meet the current adaptation deficit. The AI required to meet the current adaptation deficit included increasing the capacity of temporary pumps located along the flood embankment. To remove Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Assessment of Dhaka’s Modeled Zones 105 Table 6.2 Comparison of 2050 Flooding Results with 0.25 m Flood Criteria, Old Dhaka Model Extreme storm in Maximum flooded area (km2) Average duration 2050 (without Flood free above 0.25 m depth climate change) (0–0.25 m) 0.25–0.75 m 0.75–1.5 m >1.5 m (hours to recede) Future SE + PI (341 mm/24 hour) 10.6 0.6 0.3 0.2 5 Future SE + PI + AI (341 mm/24 hour) 11.3 0.2 0.1 0.1 1 Source: Zaman 2014. Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments. floodwater faster, all three pumps at the Dholaikhal Station were operated simul- taneously in the model. It was assumed that sedimentation would be removed from key box-culverts (Debdulai and Dholaikhal) and that khals would be pro- tected from encroachment and solid waste. Specifically, the analysis identified the following measures to address the current adaptation deficit: ­ • At Shahid Nagar—Increase pump capacity from 0.7 m3/s to 1.82 m3/s. • At Swarighat—Increase pump capacity from 0.14 m3/s to 1.12 m3/s. • At Islambag—Increase pump capacity from 1.26 m3/s to 3.08 m3/s. • Near Islambag—Increase pump capacity from 0.28 m3/s to 1.26 m3/s. • Near Farashganj—Increase pump capacity from 0.28 m3/s to 0.56 m3/s. • Near Faridabad—Increase pump capacity from 0.14 m3/s to 1.12 m3/s. • At Postogola—Increase pump capacity from 0.28 m3/s to 0.84 m3/s. • At Postogola—Install additional pumps of 0.42 m3/s capacity. • At Islambag—Install additional pumps of 0.98 m3/s capacity. • At Imamganj—Install additional pumps of 0.56 m3/s capacity. • At Dholaikhal—Operate all three pumps together for extreme rainfall events. • At Debdulai—Remove sediment from the box-culvert (2.5 m × 2.8 m): total length = 2,385 m; total volume of sediment to be cleaned = 5,500 m3 (assum- ing sedimentation depth is 33 percent of culvert height). • At Dholaikhal—Remove sediment from the box-culvert (3.8 m × 4.0 m): total length = 940 m; total volume of sediment to be cleaned = 4,700 m3 (assuming sedimentation depth is 33 percent of culvert height). These measures were first designed based on an understanding of the drainage-­ system behavior and then simulated in the drainage model. This process was repeated until the acceptable flooding criteria were met in all administrative areas. The results highlight how AI significantly reduces the extent and duration of flooding. Simulation with climate change increased the extent and duration of flooding (table 6.3). Several localities experienced more than 0.25 m inundation over 12 hours; thus, further adaptation measures were required. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 106 Assessment of Dhaka’s Modeled Zones Table 6.3  2050 Rainfall Event in a Changing Climate, Old Dhaka Maximum flooded area (km2) Average duration Flood free above 0.25 m depth Extreme storm in 2050 (0–0.25 m) 0.25–0.75 m 0.75–1.5 m >1.5 m (hours to recede) Future SE + PI (341 mm/24 hour) 10.6 0.6 0.3 0.2 5 Future SE + PI + CC (396 mm/24 hour) 9.8 1.4 0.3 0.2 6 Future SE + PI + AI (341 mm/24 hour) 11.3 0.2 0.1 0.1 1 Future SE + PI + AI + CC (396 mm/24 hour) 11.0 0.3 0.2 0.2 2 Future SE + PI + AI + CC + AD (396 mm/24 hour) 11.3 0.2 0.1 0.1 1 Source: Zaman 2014. Note: SE = socioeconomic changes; PI = planned improvements; CC = climate change; AI = additional investments; AD = adaptation to climate change. Once again, alternative measures—increasing the capacity of existing pumps, adding pumps, and removing sediment from drainage pipes—were designed, and their likely impacts were simulated in the flood model. The process was repeated until the acceptable flooding criteria were met. Further adaptation measures required to cope with climate change were identified as follows: • At Shahid Nagar—Increase pump capacity from 1.82 m3/s to 2.24 m3/s. • At Swarighat—Increase pump capacity from 1.12 m3/s to 1.26 m3/s. • At Imamganj—Increase pump capacity from 0.56 m3/s to 1.26 m3/s. • Near Farashganj—Increase pump capacity from 0.56 m3/s to 1.68 m3/s. • Near Faridabad—Increase pump capacity from 1.12 m3/s to 1.54 m3/s. • At Postogola—Increase pump capacity from 0.84 m3/s to 0.98 m3/s. • At Islambag—Install additional pumps of 1.12 m3/s capacity. • At Faridabad—Install additional pumps of 0.84 m3/s capacity. • Toward Islambag—Remove sediment from the main pipe draining toward Islambag: total length of pipe = 2,834 m; volume of sediment to be cleaned = 500 m3 (based on sedimentation depth of 20 percent of diameter). Additional simulations were conducted with the drainage model to investi- gate the effectiveness of other solutions to the urban flooding problem. For example, it was seen that converting the existing Dhopkhola playground into a 2-ha dry pond (2.5 m deep) would be effective in reducing shallow (0.1–0.25 m) flooding in that vicinity (Zaman 2014). Vulnerability from Urban Flooding It should be noted that only two of Old Dhaka’s 19 wards—wards 82 and 83—figure among Dhaka’s top 20 wards for urban flood vulnerability, indicat- ing this area’s comparatively low overall vulnerability. With socioeconomic changes and planned improvements only (SE + PI), ward 82 ranks third when Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Assessment of Dhaka’s Modeled Zones 107 equal weight (20 percent) is assigned to all five dimensions including flooding and fourth when flooding is assigned 50 percent weight. When the effect of climate change is added (SE + PI + CC), ward 82 ranks second for both equal weight and 50 percent flood weight, indicating that ward’s high vulnerability. However, if the envisaged additional investments (AI) are implemented, that ward’s vulnerability drops sharply under AI scenarios, falling, for the most part, below the top 20 for both equal weight and 50 percent flood weight. With only PI, ward 83 barely ranks in the top 20 list of vulnerable wards for both equal weight and flood weight, even when the effect of climate change is added. Under the AI scenarios, that ward’s vulnerability falls below the top 20 ranking (table 6.4). As most wards in Old Dhaka fall below the top 20 list, the area’s overall vul- nerability is relatively low, and these vulnerability rankings are further reduced when higher weightage is assigned to flooding, indicating that the area has rela- tively fewer problems from urban flooding and climate change. In addition, the suggested investments under the AI scenarios appear quite effective in further reducing these vulnerabilities. Table 6.4  FVI Rankings of Old Dhaka Wards under Alternative Flooding Scenarios Future scenarios a 50% flood weight a Equal weight  SE + PI SE + PI SE + PI SE + PI + AI SE + PI SE + PI SE + PI SE + PI + AI Ward SE + PI + CC + AI + AI + CC + CC + AD SE + PI + CC + AI + AI + CC + CC + AD 82 3 2 22 19 31 04 2 14 13 27 83 20 17 28 29 37 19 15 28 21 33 60 27 27 50 49 50 13 11 33 23 32 65 30 9 52 52 53 12 1 59 44 64 80 46 46 29 34 27 33 35 24 24 20 81 47 47 46 45 44 46 42 53 43 43 66 53 57 70 66 71 6 13 42 33 63 63 55 53 51 51 51 49 32 29 25 23 59 56 55 54 54 54 66 63 68 66 65 64 57 56 55 55 55 72 66 69 70 69 61 58 58 57 56 56 73 68 71 71 71 79 66 66 75 72 77 27 22 65 57 67 73 83 86 83 82 82 79 78 75 76 76 71 88 91 88 88 88 90 91 89 89 89 72 89 92 89 89 89 91 92 90 90 90 67 90 87 90 90 90 92 83 91 91 91 68 91 80 91 91 91 89 67 88 88 88 69 92 64 92 92 92 93 23 92 92 92 78 93 93 93 93 93 94 93 93 93 93 Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; AD = adaptation to climate change; FVI = Flood Vulnerability Index. a. Lower numbers indicate higher vulnerability. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 108 Assessment of Dhaka’s Modeled Zones Damage Estimates The total damage for Old Dhaka in 2050 was estimated for the five future sce- narios, using alternative assumptions about how frequently a 100-year, return- period storm would occur (chapter 5). Figure 6.1 shows the total monetary value of the damage that would result from the actual 100-year event in 2050 under the five future scenarios, followed by expected damage estimates using alternative probabilities of the 100-year event’s recurrence (i.e., every 20, 33, 50, or 100 years). Figure 6.2 shows the expected cumulative damage over the 2014–50 period for the five future scenarios, with random assignment of various return-period storms under alternative assumptions of increased frequency of the 100-year event. As described in chapter 5, alternative probabilities of the recurrence of the 100-year event are used to compute the expected value of the total cumulative damage between 2014 and 2050 if the event were to occur every 20, 33, 50, or 100 years. The results show how the damage that would occur under the scenario with planned improvements (PI) alone is reduced when additional investments (AI) are made to address the current adaptation deficit and how climate change exacerbates the extent of damage for each scenario. Comparison of the dam- age in a changing climate with and without AI reveals that increased damage due to climate change is less if AI are made and the current adaptation defi- cit is addressed. With adaptation measures for climate change, the damage in Figure 6.1 Total Estimated Damage for Old Dhaka in 2050 1,400 1,200 1,000 Million Tk. 800 600 400 200 0 100-year event RP20 RP33 RP50 RP100 Assumptiona SE + PI SE + PI + Al SE + PI + Al + CC + AD SE + PI + CC SE + Al + PI + CC Source: Roy 2014. Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; AD = adaptation to climate change. RP = return period (years). a. 100-year event = actual damage from a 100-year event occurring in 2050 with certainty; RP20, 33, 50, and 100 = damage from a 100-year event occurring once every 20, 33, 50, or 100 years, respectively. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Assessment of Dhaka’s Modeled Zones 109 Figure 6.2 Total Cumulative Damage for Old Dhaka, 2014–50 25,000 20,000 15,000 Million Tk. 10,000 5,000 0 RP20 RP33 RP50 RP100 Assumptiona SE + PI SE + PI + Al SE + PI + Al + CC + AD SE + PI + CC SE + Al + PI + CC Source: Roy 2014. Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; AD = adaptation to climate change. RP = return period (years). a. 100-year event = actual damage from a 100-year event occurring in 2050 with certainty; RP20, 33, 50, and 100 = damage from a 100-year event occurring once every 20, 33, 50, or 100 years, respectively. a changing climate decreases somewhat, but is still larger than in the AI scenario without climate change. Table 6.5 captures the increased damage caused by climate change under the planned improvements (PI) and additional investments (AI) scenarios for the damage occurring in 2050 and the cumulative damage in 2014–50 using various alternative assumptions. In 2050, the increased damage from a 100- year event due to climate change would amount to Tk. 608 million if only PI were made and would be reduced to Tk. 95 million with AI made to address the adaptation deficit in the current climate. Table 6.5 also indicates that the same result holds for various expected return-period storms in 2050; that is, increased damage from climate change is higher unless AI are made. The cumulative damage from climate change in 2014–50, with random assign- ment of various return-period storms over the years, is also higher when the current adaptation deficit is not addressed. In terms of sector-wise compari- sons, the largest damage for Old Dhaka occurs in residential property (i.e., moveable goods inside the house, including furniture, electronic goods, cloth- ing, and any other stored personal possessions), followed by residential build- ings and industry. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 110 Assessment of Dhaka’s Modeled Zones Cost of Adaptation Results of the hydrological modeling exercise show that most of Old Dhaka’s needed structural measures against a 100-year return-period rainfall event in 2050 beyond planned improvements involve increasing temporary pump capac- ity and sludge cleaning (Zaman 2014). For the temporary pumps, two discharge capacities are recommended: 5 ft3/s (0.14 m3/s) and 25 ft3/s (0.70 m3/s). Limiting DWASA’s inventory to two types of stock is expected to ease inventory maintenance. Pump operation requires both pipes and transformers. Keeping DWASA’s current inventory in mind, this study recommends 100 kVA and 200 kVA transformers. Each 100 kVA transformer can run three pumps of 5 ft3/s capacity, while each 200 kVA transformer can run one 25 ft3/s pump, plus one 5 ft3/s pump; alternatively, it could run six 5 ft3/s pumps. The transformers also require cables. Table 6.6 breaks down the total unit costs for the recommended capacity pumps and transformers. Table 6.5 Increased Damage in Old Dhaka from Climate Change for Future Scenarios under Alternative Assumptions Increased damage occurring in 2050 (Tk. million) Return period based on probability Scenario 100 year RP20 RP33 RP50 RP100 SE + PI + CC 608 426 378 352 317 SE + PI + AI + CC 95 83 70 62 52 Increased cumulative damage, 2014–50 (Tk. million) Using random numbers for each return period Scenario RP20 RP33 RP50 RP100 SE + PI + CC 9,746 9,489 8,746 8,409 SE + PI + AI + CC 1,554 1,518 1,404 1,357 Source: Roy 2014. Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; RP = return period (years). Table 6.6  Unit Costs for Temporary Pumps, Old Dhaka Component Cost Total unit cost (million Tk.) Pump, 5 ft3/s discharge 0.5 0.8 Pipes 0.3 Pump, 25 ft3/s discharge 1.4 2.6 Pipes 1.2 Transformer, 100 kVA 0.7 1.2 Cables 0.5 Transformer, 200 kVA 1.0 2.0 Cables 1.0 Source: DWASA officials. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Table 6.7 Estimated Cost of Meeting the Current Adaptation Deficit, Old Dhaka Additional temporary pump capacity (Tk. million) Added pumps (5 ft3 /s) and transformers Added pumps (25 ft3/s) and transformers Cost of Cost of pumps 100 kVA transformers Cost of pumps 200 kVA Cost of Pump location, added and pipes transformers and cables and pipes transformers transformers and discharge capacity (m3/s)a Pumps (no.) (0.8/unit) (no.) (1.2/unit) Pumps (no.) (2.6/unit) (no.) cables (2.0/unit) Total cost Shahid Nagar, 1.12 3 2.4 1 1.2 1 2.6 1 2.0 8.2 Islambag, 2.80 0 0 0 0 4 10.4 4 8.0 18.4 Swarighat, 1.12 2 1.6 1 1.2 1 2.6 1 2.0 7.4 Imamganj, 0.56 4 3.2 0 0 0 0 1 2.0 5.2 Farashganj, 0.28 2 1.6 1 1.2 0 0 0 0 2.8 Faridabad, 0.98 2 1.6 1 1.2 1 2.6 1 2.0 7.4 Postogola, 0.56 4 3.2 0 0 0 0 1 2.0 5.2 Subtotal 13.6 4.8 18.2 18.0 54.6 Sludge cleaning (Tk. million) Removal Length to Sediment removal from box-culverts Removal rate/m3 volume (m3) clean (m) Total cost Debdulai (2.5 m × 2.8 m) 1,000 5,500 2,385 5.5 Dholaikhal (3.8 m × 4 m) 1,000 4,700 940 4.7 Subtotal 10.2 Total cost: 64.8 Sources: DWASA officials and authors’ calculations. a. Pumping head = 10 m. 111 112 Table 6.8 Estimated Cost of Meeting the Climate Change Adaptation Deficit, Old Dhaka Additional temporary pump capacity (Tk. million) Added pumps (5 ft3/s) and transformers Added pumps (25 ft 3/s) and transformers Cost of Cost of Pump location, Cost of pumps 100 kVA transformers Cost of pumps 200 kVA transformers added discharge and pipes transformers and cables and pipes transformers and cables capacity (m3/s)a Pumps (no.) (0.8/unit) (no.) (1.2/unit) Pumps (no.) (2.6/unit) (no.) (2.0/unit) Total cost Shahid Nagar, 0.42 3 2.4 1 1.2 0 0 0 0 3.6 Islambag, 1.12 3 2.4 1 1.2 1 2.6 1 2.0 8.2 Swarighat, 0.14 1 0.8 1 1.2 0 0 0 0 2.0 Imamganj, 0.70 0 0 0 0 1 2.6 1 2.0 4.6 Farashganj, 1.12 3 2.4 1 1.2 1 2.6 1 2.0 8.2 Faridabad, 0.84 1 0.8 0 0 1 2.6 1 2.0 5.4 Postogola, 0.14 1 0.8 1 1.2 0 0 0 0 2.0 Subtotal 9.6 6.0 10.4 8.0 34.0 Additional sludge cleaning (Tk. million) Removal Removal Length to Description rate/m3 quantity (m3) clean (m) Total cost Main drainage pipe toward Islambag 600 500 2,834 0.3 Subtotal 0.3 Total cost: 34.3 Sources: DWASA officials and authors’ calculations. a. Pumping head = 10 m. Assessment of Dhaka’s Modeled Zones 113 As previously discussed, addressing the current structural adaptation deficit will require temporary pumps at the following points: Shahid Nagar, Islambag, Swarighat, Imamganj, Farashganj, Faridabad, and Postogola. In addition, it will require regular sediment removal from the Debdulai and Dholaikhal box-­ culverts. Table 6.7 breaks down the cost estimates for meeting the current adap- tation deficit. As shown, the cost of added temporary pump capacity and sludge removal are Tk. 54.6 million and Tk. 10.2 million, respectively, for a total esti- mated adaptation cost of Tk. 64.8 million. It is argued that nearly 1 m of sludge is deposited in box-culverts each year. The total cost of removing 1 m3 of sludge (including the costs of dewatering, dam construction, and other related costs) is estimated at Tk. 1,000. Since the average width of a box-culvert is estimated at 5 m, the cost of removing sludge from 1 m of box-culvert is Tk. 5,000. Pipe cleaning is less expensive, at Tk. 600 per m3. This figure slightly overestimates the cost of cleaning a meter of pipe since the average pipe diameter is less than 1 m. Adapting to climate change would mean increasing the temporary pump capacity at all points. For sludge cleaning, it would mean cleaning the drainage pipe toward Islambag in addition to removing sediment from the box-culverts at Debdulai and Dholaikhal. Table 6.8 breaks down the total added cost estimates for addressing climate change adaptation. The estimated added expense for pumps, pipes, and transformers is Tk. 34 million, while the cost of additional sludge cleaning is Tk. 0.3 million, for a total additional cost of Tk. 34.3 million. Finally, table 6.9 provides a summary comparison of the total costs to meet the current adaptation deficit and the added cost of climate change adapta- tion. As shown, addressing the current deficit totals Tk. 64.8 million, while the cost of meeting the added climate change deficit totals Tk. 34.3 million, for a total combined cost of Tk. 99.1 million (Huq 2014). The damage estimates demonstrate that this investment cost in flood mitigation will lead to a much higher reduction in damage of Tk. 1,002 million in 2050 when faced with a 100-year, return-period storm, with a cumulative damage reduction (2014–50) of Tk. 13,915 million. Table 6.9 Summary of Adaptation Costs, Old Dhaka Recommended measure Cost (Tk. million) Addressing current adaptation deficit Temporary pumps 54.6 Sludge cleaning (sediment removal from box-culverts) 10.2 Total estimated cost 64.8 Addressing climate change deficit Temporary pumps 34.0 Sludge cleaning (drainage pipe cleaning) 0.3 Total estimated cost 34.3 Total combined cost 99.1 Sources: DWASA officials and authors’ calculations. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 114 Assessment of Dhaka’s Modeled Zones Central Dhaka Results Hydrological Modeling Output The study generated maps simulating the inundation depth and duration of urban flooding for the historic 100-year, return-period rainfall event in 2004 and alternative scenarios for 2050 with and without climate change. Upon consulta- tion with local experts, the 0.1 m inundation criteria were adopted for Central Dhaka. For the 100-year rainfall event in 2050 without climate change, there was relatively little additional flooded area; at inundation depths of 0–0.1 m, 0.1–0.25 m, 0.25–0.75 m, 0.75–1.5 m, and >1.5 m, maximum flooded areas were 26.4 km2, 8.8 km2, 1.9 km2, 0.3 km2, and 1.8 km2, respectively. However the expected duration of flooding above 0.1 m depth—even without climate change—was significant, at 11.4 hours on average. With only planned improvements (PI), the maps generated showed that, even without climate change, the drainage system is unable to cope with a 100-year return-period storm (341 mm per 24 hours) (photo 6.1). Thus, additional invest- ments (AI) were incorporated into the model to satisfy the set criteria for accept- able flooding in Central Dhaka (i.e., 90 percent of each administrative area must have less than 0.1 m flooding, and no area can be inundated for more than 12 hours). Map 6.2 compares the simulated results for the extreme rainfall event in 2050 with PI and the AI to address the adaptation deficit in the current climate. The results highlight the significant reduction in the extent and duration of flooding from the AI (table 6.10). Photo 6.1 Original Rampura sluice gates (at right) and new bypass channel used when gates are closed (Central Dhaka system) Credit: © Asif Zaman / World Bank. Used with permission. Further permission required for reuse. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Assessment of Dhaka’s Modeled Zones 115 Map 6.2 Extreme Rainfall Scenarios in 2050 with 0.1 m Flood Criteria, Central Dhaka Source: IWM 2014. Note: PI = planned improvements; AI = additional investments. Table 6.10  2050 Rainfall Event in a Changing Climate, Central Dhaka Maximum flooded area (km2) Average duration Flood free above 0.1 m depth Extreme storm in 2050 (0–0.1 m) 0.1–0.25 m 0.25–0.75 m 0.75–1.5 m >1.5 m (hours to recede) Future SE + PI (341 mm/24 hour) 26.4 8.8 1.9 0.3 1.8 10 Future SE + PI + CC (396 mm/24 hour) 25.1 10.1 1.9 0.3 1.8 12 Future SE + PI + AI (341 mm/24 hour) 36.4 0.7 0.3 0.1 1.7 1 Future SE + PI + AI + CC (396 mm/24 hour) 35.7 1.1 0.5 0.2 1.8 2 Future SE + PI + AI + CC + AD (396 mm/24 hour) 36.1 0.9 0.4 0.1 1.8 1 Source: Zaman 2014. Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; AD = adaptation to climate change. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 116 Assessment of Dhaka’s Modeled Zones Photo 6.2 Construction of new pump station at Hatir Jheel, Rampura (Central Dhaka system) Credit: © Asif Zaman / World Bank. Used with permission. Further permission required for reuse. The part of Central Dhaka beyond Pragati Sharani is unprotected, and some new roads being constructed may work as embankments. Also, new pump stations are needed to drain out excess water from heavy rainfall (photo 6.2). The analysis identified the following specific measures to address the current adaptation deficit: • At Gulshan, Banani, and Hatir Jheel lakes—Reduce initial water levels from 5 m PWD to 4 m PWD (lowering of the lakes’ water levels represents early warning of large storm event). • At outfall of Panthapath Box-Culvert—Install a new pump station with 12.5 m3/s capacity and 80 percent efficiency. • Near Bashabo Water Pump Station of DWASA—Install new pump stations and sluice gate; pump capacity is 15 m3/s, with 80 percent efficiency, and gate size is 4 m × 3 m. When the effect of climate change was simulated (396 mm per 24 hours), it was found that the AI were insufficient to meet the acceptable flooding criteria in all administrative areas. Thus, further adaptation measures were required to reduce the impact of increased rainfall intensity due to climate change. Therefore, alternative measures were designed and simulated in the flood model. The process was repeated until the acceptable flooding criteria for Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Assessment of Dhaka’s Modeled Zones 117 Central Dhaka were met. Further adaptation measures required to cope with climate change were identified as follows: • Near Russel Square—Install a new sluice gate (3.4 m × 3.4 m) with automatic operation to divert excess water to Dhanmondi Lake at the time of peak r­ unoff and prevent backwater flow from Panthapath Box-Culvert. • Near Russel Square—Reactivate the existing sluice gate (2 m × 2 m) and auto- matic operation with the new gate (above) to divert excess water to Dhanmondi Lake at the time of peak runoff. Vulnerability to Urban Flooding Despite the more stringent criteria, Central Dhaka is one of the least vulnerable areas in Dhaka, with only two of its 44 wards—25 and 75—figuring among the top 20 vulnerable wards/regions. When the weightage assigned to flooding is increased, with the transition from equal weight to 50-percent flood weight, these two wards become even less vulnerable. Vulnerability also decreases with the transition from planned improvements (PI) to additional investments (AI) as flooding levels fall. From this, one can infer that flooding plays a role in the higher vulnerability of the two wards, but not as large as the lower-than-average indices of the Climate Disaster Resilience Index (CDRI) in the two wards. Table 6.11 provides a comprehensive listing of flood vulnerability indices for all of the con- stituent wards of Central Dhaka. Table 6.11  FVI Rankings of Central Dhaka Wards under Alternative Flooding Scenarios Future scenarios Equal weight a 50% flood weight a SE + PI SE + PI SE + PI SE + PI SE + PI + + AI + CC SE + PI SE + PI SE + PI + AI + CC Ward SE + PI + CC + AI AI + CC + AD SE + PI + CC + AI + AI + CC + AD 25 4 4 14 16 13 11 10 16 17 11 75 19 10 43 43 42 16 12 36 38 37 16 24 21 24 24 23 32 30 40 40 39 23 26 24 21 22 21 31 29 25 27 24 26 31 26 25 25 24 43 40 41 41 40 33 32 31 26 26 26 26 28 22 22 22 28 33 33 30 31 29 35 41 44 47 46 22 34 30 27 28 25 37 37 31 34 31 76 35 32 39 30 28 30 25 35 30 29 85 36 34 42 42 41 39 33 56 58 55 27 37 35 31 32 30 52 47 46 48 47 24 38 36 32 33 32 53 49 47 49 48 29 39 37 35 36 34 51 48 49 50 50 34 40 43 40 40 39 44 44 43 45 44 31 41 38 34 35 33 54 50 50 51 51 table continues next page Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 118 Assessment of Dhaka’s Modeled Zones Table 6.11  FVI Rankings of Central Dhaka Wards under Alternative Flooding Scenarios (continued) Future scenarios Equal weight a 50% flood weight a SE + PI SE + PI SE + PI SE + PI SE + PI + + AI + CC SE + PI SE + PI SE + PI + AI + CC Ward SE + PI + CC + AI AI + CC + AD SE + PI + CC + AI + AI + CC + AD 30 43 42 37 39 38 58 53 54 55 53 35 45 45 41 41 40 59 56 55 56 54 32 48 48 44 44 43 61 58 58 59 57 84 49 49 45 46 45 62 59 61 60 58 77 50 50 47 47 46 50 45 45 53 49 36 51 51 48 48 47 64 60 62 61 61 62 54 54 53 53 52 38 36 51 36 35 38 59 76 61 59 60 24 79 21 20 17 19 62 62 56 60 57 47 54 19 26 19 37 63 63 62 62 61 45 57 26 32 25 51 67 82 81 81 81 55 84 83 83 83 54 68 67 65 65 65 70 71 57 65 59 55 69 69 68 67 66 67 70 63 62 56 45 70 75 76 75 75 74 80 78 78 77 18 71 74 73 74 73 68 72 72 72 72 39 72 70 71 71 70 77 76 76 74 74 52 73 71 67 69 68 60 64 38 42 38 49 74 68 66 68 67 69 65 39 52 42 21 75 73 72 73 72 75 74 73 73 73 20 76 72 69 70 69 76 73 67 68 68 17 77 77 78 78 78 81 81 79 79 78 40 78 79 74 76 74 80 82 74 75 75 57 80 84 82 83 84 78 77 77 77 80 50 81 83 80 80 80 83 86 82 82 82 44 82 85 79 79 79 84 87 81 81 81 70 84 81 84 84 83 85 75 84 84 84 56 85 88 85 85 85 86 88 85 85 85 53 86 89 86 86 86 87 89 86 86 86 74 87 90 87 87 87 88 90 87 87 87 Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; AD = adaptation to climate change; FVI = Flood Vulnerability Index. a. Lower numbers indicate a higher vulnerability. Damage Estimates The total damage for Central Dhaka in 2050 was estimated for the five future scenarios, using alternative assumptions about how frequently a 100-year, return- period storm would occur (chapter 5). Figure 6.3 shows the total monetary value of the damage that would result from the actual 100-year event in 2050 under the five future scenarios, followed by expected damage estimates using Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Assessment of Dhaka’s Modeled Zones 119 Figure 6.3 Total Estimated Damage for Central Dhaka in 2050 5,000 4,000 3,000 Million Tk. 2,000 1,000 0 100-year event RP20 RP33 RP50 RP100 Assumptiona SE + PI SE + PI + Al SE + PI + Al + CC + AD SE + PI + CC SE + Al + PI + CC Source: Roy 2014. Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; AD = adaptation to climate change. RP = return period (years). a. 100-year event = actual damage from a 100-year event occurring in 2050 with certainty; RP20, 33, 50, and 100 = damage from a 100-year event occurring once every 20, 33, 50, or 100 years, respectively. alternative probabilities of the 100-year event’s recurrence (i.e., every 20, 33, 50, or 100 years). Figure 6.4 shows the expected cumulative damage over the 2014–50 period for the five future scenarios, with random assignment of various return-period storms under alternative assumptions of increased frequency of the 100-year event. As described in chapter 5, alternative probabilities of the recurrence of the 100-year event are used to compute the expected value of the total cumulative damage between 2014 and 2050 if the event were to occur every 20, 33, 50, or 100 years. The results show how the damage that would occur under the scenario with planned improvements (PI) alone is reduced when additional investments (AI) are made to address the current adaptation deficit and how climate change exac- erbates the extent of damage for each scenario. Comparison of the damage in a changing climate with and without AI reveals that increased damage due to cli- mate change is less if AI are made and the current adaptation deficit is addressed. With measures to address climate change adaptation, the damage in a changing climate decreases somewhat, but is still larger than in the AI scenario without climate change (photo 6.3). Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 120 Assessment of Dhaka’s Modeled Zones Figure 6.4 Total Cumulative Damage for Central Dhaka, 2014–50 80,000 60,000 Million Tk. 40,000 20,000 0 RP20 RP33 RP50 RP100 Assumptiona SE + PI SE + PI + Al SE + PI + Al + CC + AD SE + PI + CC SE + Al + PI + CC Source: Roy 2014. Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; AD = adaptation to climate change. RP = return period (years). a. 100-year event = actual damage from a 100-year event occurring in 2050 with certainty; RP20, 33, 50, and 100 = damage from a 100-year event occurring once every 20, 33, 50, or 100 years, respectively. Photo 6.3 Laying new stormwater pipe in Gulshan area of Central Dhaka system Credit: © Asif Zaman / World Bank. Used with permission. Further permission required for reuse. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Assessment of Dhaka’s Modeled Zones 121 Table 6.12 Increased Damage in Central Dhaka from Climate Change for Future Scenarios under Alternative Assumptions Increased damage occurring in 2050 (Tk. million) Return period based on probability Scenario 100 year RP20 RP33 RP50 RP100 SE + PI + CC 671 415 390 376 358 SE + PI + AI + CC 111 69 65 63 60 Increased cumulative damage 2014–50 (Tk. million) Using random numbers for each return period Scenario RP20 RP33 RP50 RP100 SE + PI + CC 10,884 10,611 9,797 9,439 SE + PI + AI + CC 1,799 1,754 1,621 1,564 Source: Roy 2014. Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; RP = return period (years). Table 6.12 captures the increased damage caused by climate change under the planned improvements (PI) and additional investments (AI) scenarios for the damage occurring in 2050 and the cumulative damage in 2014–50 using various alternative assumptions. Thus, in 2050, the increased damage from a 100-year event due to climate change would amount to Tk. 671 million if only PI were made and would be reduced to Tk. 111 million with AI to address the adaptation deficit in the current climate. Table 6.12 shows similar outcomes of increased damage caused by climate change under a weighted probability of various return-period storms for both PI and AI scenarios. It also shows similar cumulative effects of climate change in 2014–50, with random assignment of various return-period storms under alternative assumptions of increased frequency of a 100-year, return-period storm. In terms of sector-wise comparisons, the largest damage for Central Dhaka occurs in residential property, followed by residential buildings, health care, and industry. Cost of Adaptation Results of the hydrological modeling exercise show that, for Central Dhaka, the cost of adaptation beyond planned improvements would involve installing two permanent pump stations—a 15 m3/s capacity station near Bashabo and a 12 m3/s capacity station on Hatir Jheel behind Hotel Sonargaon; the site loca- tion would be the outfall of the Panthapath Box-Culvert (Zaman 2014). The station near Bashabo would be installed adjacent to DWASA’s water supply pump station to relieve inundation in large parts of Khilgaon and Bashabo. Pump-station cost components are mechanical, electrical, and civil works. For Bashabo, these are estimated at Tk. 450 million, 100 million, and 160 million, respectively. For Panthapath, they are estimated at Tk. 380 million, 80 million, Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 122 Assessment of Dhaka’s Modeled Zones Table 6.13 Estimated Cost of Meeting the Current Adaptation Deficit, Central Dhaka Pump Pump Mechanical Electrical Civil works Total cost capacity Head price/unit Unit works works (million (million Required item (m3/s) (m) (million Tk.) (no.) (million Tk.) (million Tk.) Tk.) Tk.) Bashabo Pump station, adjacent to DWASA’s water supply pump station 15 6 150 3 450 100 160 710 Land acquisition (15 decimals @ 3 million Tk./decimal)a 45 Sluice gate (4 m × 3 m) = 92.5 million Tk., plus 2 million Tk. (mechanical hoist) 94.5 Subtotal 849.5 Panthapath Pump station at outfall of box-culvert 12.5 6 190 2 380 80 140 600 Subtotal 600 Total cost 1,449.5 Sources: DWASA officials and authors’ calculations. a. If land is state-owned, no cost would be incurred. Table 6.14 Estimated Cost of Meeting the Climate Change Adaptation Deficit, Central Dhaka Required item Total cost (million Tk.) Russel Square Sluice gate, 3.4 m × 3.4 m 89.1 Mechanical hoist 2.0 Total cost 91.1 Sources: DWASA officials and authors’ calculations. and 140 ­ million. At current market prices, pump-station construction costs would total Tk. 710 million for Bashabo and Tk. 600 million for Panthapath (table 6.13). The Bashabo site would also entail costs for land acquisition and sluice- gate construction. The pump station would be located across the road from a DWASA water supply pump station. Using a portion of land at this existing station as a rest area for pump operators, office space, and storage would minimize the amount of land that would need to be acquired, which is esti- mated at 15 decimals. If the land is state-owned (i.e., khas land), no cost would be incurred. The recommended sluice gate (4 m × 3 m) and mechani- cal hoist to make it operable would cost an estimated Tk. 94.5 million (table 6.13). The cost of addressing the current adaptation deficit for Central Dhaka—the area with the most stringent acceptable flooding criteria—totals nearly Tk. 1.45 billion, the highest among the areas studied in the hydrological modeling exer- cise. However, once the current adaptation deficit is met, only Tk. 91.1 million more would be required to meet the climate change deficit; this would cover the Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Assessment of Dhaka’s Modeled Zones 123 Table 6.15 Summary of Adaptation Costs, Central Dhaka Recommended measure Cost (Tk. million) Addressing current adaptation deficit Pump station, Bashabo 849.5 Pump station, Panthapath 600.0 Total estimated cost 1,449.5 Addressing climate change deficit Total estimated cost, sluice gate at Russel Square 91.1 Total combined cost 1,540.6 Sources: DWASA officials and authors’ calculations. cost of installing a sluice gate near Russel Square to prevent backflow toward Asad Gate (table 6.14). Table 6.15 summarizes the total adaptation costs for the proposed recommendations (Huq 2014). The damage estimates demonstrate that ­ this investment of Tk. 1,541 million in flood mitigation will lead to a much higher reduction in damage of Tk. 3,888 million when faced with a ­ 100-year, return-period storm and a cumulative damage reduction (2014–50) of Tk. 53,387 million. Kallyanpur System Results Hydrological Modeling Output The study generated maps simulating the inundation depth and duration of urban flooding for the historic 100-year, return-period rainfall event in 2004 and alternative scenarios for 2050 with and without climate change. The 2004 base scenario led to prolonged flooding, which took an average of 13.5 hours to recede. For 2050, alternative scenarios were simulated using flood criteria of 0.1 m and 0.25 m. Comparison of 100-year, return-period storms for 2004 and 2050 without climate change reveal a greater extent of flooding in 2004 due to considerably less total pump capacity than in future scenarios. In 2050, there was a marked reduction in flooding duration, owing mainly to a doubling of pump capacity and the laying of new drainage pipes (figure 6.5). Results of the extreme rainfall event in 2050 were re-analyzed using the flood criteria for 0.25 m depth under the premises that inhabitants of Kallyanpur are accustomed to ankle-deep water since flooding is a recurrent phenomenon in their neighborhood. The maps generated showed that, for some areas, there is an adaptation deficit; that is, the planned improvements (PI) are insufficient to cope with rainfall extremes (341 mm per 24 hours) (photo 6.4). Therefore, additional investments (AI) were incorporated into the model. These included increasing the water-body storage volume in one ward (ward 9), laying new pipes in four wards (wards 9, 47, 48, and 58), replacing a pipe in one ward (ward 48), and increasing the working efficiency and capacities at temporary pump locations (photo 6.5). Map 6.3 highlights the impact of making these AI. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 124 Assessment of Dhaka’s Modeled Zones Figure 6.5  Depth and Duration of Urban Flood: Comparisons for Rainfall Extremes (341 mm/24 hours) over the Study Period, Kallyanpur System Model a. Flood depth b. Flood duration 30 15 25 Maximum flooded area (km2) 12 20 9 Hours 15 6 10 3 5 0 0 0–0.1 0.1–0.25 0.25–0.75 0.75–1.5 >1.5 Maximum flood depth (m) 2004 2050 Source: Zaman 2014. Photo 6.4 Resectioning and lining work at Katashur Khal near Mohammadpur (Kallyanpur system) Credit: © Asif Zaman / World Bank. Used with permission. Further permission required for reuse. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Assessment of Dhaka’s Modeled Zones 125 Photo 6.5 Temporary pumps located along the Western Embankment, Kallyanpur Credit: © Subhendu Roy / World Bank. Used with permission. Further permission required for reuse. Map 6.3 Extreme Rainfall Scenarios in 2050 with 0.25 m Flood Criteria, Kallyanpur System Model Source: IWM 2014. Note: PI = planned improvements; AI = additional investments. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 126 Assessment of Dhaka’s Modeled Zones The analysis identified the following specific measures to address the current adaptation deficit: • In ward 9—Deepen water-body storage volume of pond (excavating 262,000 m3 of soil and not reducing the present pond size). • In wards 9, 47, 48, and 58—Lay new pipes, of 362 m length, 0.9 m diameter (ward 9); 235 m length, 1.0 m diameter (linking to ward 9 khal); 814 m length, 3.0 m diameter (to Rayerbazar Temporary Pump Station); 565 m length, 2.0 m diameter (to Sikdar Temporary Pump Station); 111 m length, 1.0 m diam- eter (just south of Sikdar Temporary Pump Station); and 106 m length, 1.0 m diameter (northwest of Hazaribag Temporary Pump Station 2). • In ward 48—Replace drain pipe (swap direction of drain of 210 m length, 1.52 m diameter). • At Rayerbazar Temporary Pump Station—Increase working efficiency of temporary pump from 60 percent to 80 percent (i.e., effective capacity of ­ 0.336 m3/s). • At Sikdar Temporary Pump Station—Increase working efficiency of tempo- rary pump from 60 percent to 80 percent (i.e., effective capacity of 0.448 m3/s). • At Hazaribag Temporary Pump Station—Increase capacity of pump 1 from 0.35 m3/s to 2.5 m3/s and working efficiency from 60 percent to 80 percent. • At Hazaribag Temporary Pump Station—Increase capacity of pump 2 from 0.35 m3/s to 1.25 m3/s and working efficiency from 60 percent to 80 percent. When the effect of climate change was simulated (396 mm per 24 hours), it was found that these additional improvements were insufficient to meet the acceptable flooding criteria, thus making further interventions necessary (table 6.16). Table 6.16  2050 Rainfall Event in a Changing Climate, Kallyanpur System Model Maximum flooded area (km2) Average duration Flood free above 0.25 m depth Extreme storm in 2050 (0–0.25 m) 0.25–0.75 m 0.75–1.5 m >1.5 m (hours to recede) Future SE + P1 (341 mm/24 hour) 25.1 0.9 0.8 1.0 8 Future SE + PI + CC (396 mm/24 hour) 21.6 1.1 0.8 1.0 9 Future SE + PI + AI (341 mm/24 hour) 26.5 0.5 0.3 0.5 3 Future SE + PI + AI + CC (396 mm/24 hour) 26.3 0.6 0.3 0.6 4 Future SE + PI + AI + CC + AD (396 mm/24 hour) 26.5 0.5 0.3 0.5 3 Source: Zaman 2014. Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; AD = adaptation to climate change. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Assessment of Dhaka’s Modeled Zones 127 To cope with the increased rainfall intensity due to climate change, the alter- native measures were designed and simulated in the flood model. The process was repeated until the acceptable flooding criteria were met. The required adap- tation measures were identified as follows: • At Hazaribag Temporary Pump Station—Increase capacity of pump 1 from 2.5 m3/s to 5.0 m3/s. Vulnerability to Urban Flooding Kallyanpur is quite vulnerable, with 6 of its 15 wards—8–11, 48, and 58—figuring among Dhaka’s top 20 vulnerable wards/regions. Under planned improvements (PI), ward 9 has the highest vulnerability. Kallyanpur’s vulnerable wards generally have lower-than-average CDRI rankings, but three wards (9, 48, and 58) exhibit greater vulnerability due to their high Flood Vulnerability Index (FVI). Their vulnerability decreases from PI to additional investments (AI) scenarios as flood- ing levels fall and increases as greater weight is assigned to flood vulnerability under the 50 percent flood weight. Wards 8, 10, and 11 are vulnerable mainly because of their lower-than-average CDRI rankings; their vulnerability changes little or falls with the transition from equal weight to 50 percent flood weight and increases from PI to AI scenarios as flooding levels go down (table 6.17). Table 6.17  FVI Rankings of Kallyanpur Wards under Alternative Flooding Scenarios Future scenarios Equal weighta 50% flood weighta SE + PI SE + PI SE + PI SE + PI SE + PI + AI + CC SE + PI SE + PI SE + PI + + AI + CC Ward SE + PI + CC + AI + AI + CC + AD SE + PI + CC + AI AI + CC + AD 09 1 1 4 6 6 5 7 20 19 18 10 6 5 2 4 3 22 17 6 11 9 58 7 15 33 27 49 1 3 11 10 21 08 11 7 12 2 5 17 14 10 3 7 11 13 6 6 8 12 23 20 12 18 15 48 18 25 59 57 59 2 5 70 69 70 13 21 16 16 18 15 34 31 27 28 26 41 23 18 19 20 18 41 38 37 37 36 14 25 22 20 21 19 36 34 30 31 30 12 29 23 23 23 22 42 39 32 35 34 07 44 44 38 37 35 56 51 48 46 45 47 61 59 64 64 64 29 21 60 64 60 46 64 61 60 61 62 48 43 23 29 28 43 65 65 63 63 63 63 62 34 39 41 42 79 78 77 77 76 82 85 80 80 79 Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; AD = adaptation to climate change. a. Lower numbers indicate a higher vulnerability. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 128 Assessment of Dhaka’s Modeled Zones Damage Estimates The total damage for Kallyanpur in 2050 was estimated for the five future scenarios, using alternative assumptions about how frequently a 100-year, ­ ­ return-period storm would occur (chapter 5). Figure 6.6 shows the total mone- tary value of the damage that would result from the actual 100-year event in 2050 under the five future scenarios, followed by expected damage estimates using alternative probabilities of the 100-year event’s recurrence (i.e., every 20, 33, 50, or 100 years). Figure 6.7 shows the expected cumulative damage over the 2014–50 period for the five future scenarios, with random assignment of various return-period storms under alternative assumptions of increased frequency of the 100-year event. As described in chapter 5, alternative probabilities of the recurrence of the 100-year event are used to compute the expected value of the total cumu- lative damage between 2014 and 2050 if the event were to occur every 20, 33, 50, or 100 years. The results show how the damage that would occur under the scenario with planned improvements (PI) alone is reduced when additional investments (AI) are made to address the current adaptation deficit and how climate change exac- erbates the extent of damage for each scenario. Comparison of the damage in a Figure 6.6 Total Estimated Damage for Kallyanpur in 2050 2,500 2,000 1,500 Million Tk. 1,000 500 0 100-year event RP20 RP33 RP50 RP100 Assumptiona SE + PI SE + PI + Al SE + PI + Al + CC + AD SE + PI + CC SE + Al + PI + CC Source: Roy 2014. Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; AD = adaptation to climate change. RP = return period (years). a. 100-year event = actual damage from a 100-year event occurring in 2050 with certainty; RP20, 33, 50, and 100 = damage from a 100-year event occurring once every 20, 33, 50, or 100 years, respectively. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Assessment of Dhaka’s Modeled Zones 129 Figure 6.7 Total Cumulative Damage for Kallyanpur, 2014–50 40,000 30,000 Million Tk. 20,000 10,000 0 RP20 RP33 RP50 RP100 Assumptiona SE + PI SE + PI + Al SE + PI + Al + CC + AD SE + PI + CC SE + Al + PI + CC Source: Roy 2014. Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; AD = adaptation to climate change. RP = return period (years). a. 100-year event = actual damage from a 100-year event occurring in 2050 with certainty; RP20, 33, 50, and 100 = damage from a 100-year event occurring once every 20, 33, 50, or 100 years, respectively. changing climate with and without AI reveals that the increased damage due to climate change is less if AI are made and the current adaptation deficit is addressed. With measures to address climate change adaptation, the damage in a changing climate decreases somewhat, but is still larger than in the AI scenario without climate change. Table 6.18 captures the increased damage caused by climate change under the planned improvements (PI) and additional investments (AI) scenarios for the damage occurring in 2050 and the cumulative damage in 2014–50 using various alternative assumptions. In 2050, the increased damage from a 100-year event due to climate change would amount to Tk. 148 million if only PI were made and would be reduced to Tk. 113 million with AI. Table 6.18 shows simi- lar outcomes of increased damage caused by climate change under a weighted probability of various return-period storms for both PI and AI scenarios. It also shows similar cumulative effects of climate change in 2014–50, with random assignment of various return-period storms under alternative assumptions of increased frequency of a 100-year, return-period storm. In terms of sector-wise comparisons for Kallyanpur, the largest damage is to residential buildings, ­ followed by residential property (moveable goods inside the house, including furniture, electronic goods, clothing, and any other stored personal possessions) and industrial units. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 130 Assessment of Dhaka’s Modeled Zones Table 6.18 Increased Damage in Kallyanpur from Climate Change for Future Scenarios under Alternative Assumptions Increased damage occurring in 2050 (Tk. million) Return period based on probability Scenario 100 year RP20 RP33 RP50 RP100 SE + PI + CC 148 278 192 145 82 SE + PI + AI + CC 113 142 107 87 61 Increased cumulative damage, 2014–50 (Tk. million) Using random numbers for each return period Scenario RP20 RP33 RP50 RP100 SE + PI + CC 2,429 2,372 2,197 2,123 SE + PI + AI + CC 1,842 1,797 1,661 1,602 Source: Roy 2014. Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; RP = return period (years). Cost of Adaptation For Kallyanpur, the recommended structural measures against a 100-year ­ return-period rainfall event in 2050 beyond planned improvements focus mainly on increasing temporary-pump capacity and laying new drainage pipes (Zaman 2014). The recommended temporary pumps have 10 m head with two discharge capacities: 5 ft3/s (0.14 m3/s) and 25 ft3/s (0.70 m3/s). Pump operation requires both pipes and transformers. This study recommends 100 kVA and 200 kVA transformers, whose unit costs are the same as those for Old Dhaka (table 6.19). Addressing the current structural adaptation deficit will require additional temporary pumps at the following points: Rayerbazar, Sikdar, and Hazaribag. It will also require laying new drainage pipes and switching the direction of an existing pipe in ward 48 by 180 degrees. Installation expenses, in addition to pipe costs, include laying pipes, as well as digging roads, for which compensatory pay- ment must be made to the Dhaka City Corporations (DCCs). The per-meter cost of a pipe is a function of its diameter, and the pipe-laying cost increases with the pipe diameter to be laid. Also, the ditch in which the pipes are laid must be wide enough for laborers to work on the pipe. For road digging, the study assumes a high-end rate (Tk. 6,250 per m2) to avoid underestimating the needed payment. According to practitioners, the proposed pipes to temporary pump stations at Rayerbazar (3 m diameter and 814 m length) and Sikdar (2 m diameter and 565 m length) cannot be easily installed. For Rayerbazar, they suggest building a brick sewer. Assuming a cost of Tk. 0.25 million per meter, the total cost of the brick sewer is about Tk. 223.9 million. For the link to Sikdar Temporary Pump Station, they suggest that two 1.52 m diameter pipes be laid, one on top of the other. This study estimates the total cost of the pipes, pipe laying, and payment for road digging at about Tk. 55.4 million. The pipe whose direction is to be switched has Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Table 6.19 Estimated Cost of Meeting the Current Adaptation Deficit, Kallyanpur Additional temporary pump capacity (Tk. million) Temporary pump Added pumps (5 ft 3/s) and transformers Added pumps (25 ft 3/s) and transformers location, added Cost of pumps 100 kVA Cost of Cost of pumps 200 kVA Cost of discharge capacity and pipes transformers transformers and and pipes transformers transformers and Total (m3/s)a Pumps (no.) (0.8/unit) (no.) cables (1.2/unit) Pumps (no.) (2.6/unit) (no.) cables (2.0/unit) cost Rayerbazar, 0.12 1 0.8 1 1.2 0 0 0 0 2.0 Sikdar, 0.15 1 0.8 1 1.2 0 0 0 0 2.0 Hazaribag, pump 1, 2.27 2 1.6 0 0 3 7.8 3 6.0 15.4 Hazaribag, pump 2, 1.02 3 2.4 1 1.2 1 2.6 1 2.0 8.2 Subtotal 27.6 New pipe laying/switching pipe direction (Tk. million) Location/link to Pipe diameter Pipe installation Road-cutting Total temporary pump station (m) Pipe length (m) Pipe price/m cost/m cost/m cost Ward 9 0.9 362 5,126 14,400 12,000 11.4 Ward 9 khal 1 235 7,885 20,400 12,500 9.6 Rayerbazar (brick sewer) 3 814 250,000 0 25,000 223.9 Sikdar 2 565 30,600 50,000 17,500 55.4 South of Sikdar 1 111 7,885 20,400 12,500 4.5 Northwest of Hazaribag, 2 1 106 7,885 20,400 12,500 4.3 Switching pipe direction in ward 48 1.52 210 0 26,400 15,100 8.7 Subtotal 317.8 Total cost: 345.4 Sources: DWASA officials and authors’ calculations. a. Head = 10 m. 131 132 Assessment of Dhaka’s Modeled Zones Table 6.20 Estimated Cost of Meeting the Climate Change Adaptation Deficit, Kallyanpur Additional temporary pump capacity (Tk. million) Added pumps (5 ft3/s) and transformers Added pumps (25 ft3/s) and transformers Temporary pump Cost of Cost of Cost of Cost of location, added pumps 100 kVA transformers pumps 200 kVA transformers discharge Pumps and pipes transformers and cables Pumps and pipes transformers and cables Total capacity (m3/s)a (no.) (0.8/unit) (no.) (1.2/unit) (no.) (2.6/unit) (no.) (2.0/unit) cost Hazaribag, pump 1, 2.5 3 2.4 0 0 3 7.8 3 6 16.2 Subtotal 16.2 Total cost: 16.2 Sources: DWASA officials and authors’ calculations. a. Head = 10 m. Table 6.21 Summary of Adaptation Costs, Kallyanpur Recommended measure Cost (Tk. million) Addressing current adaptation deficit Temporary pumps 27.6 Laying new pipes/switching pipe direction 317.8 Total estimated cost 345.4 Addressing climate change deficit Total estimated cost, temporary pumps 16.2 Total combined cost 361.6 Sources: DWASA officials and authors’ calculations. a 1.5 m diameter and is 210 m in length; the cost of changing its direction is estimated at Tk. 8.7 million. Since an older pipe can be reused for this purpose, no purchase is necessary. Four additional pipes, with lengths in a range of 106– 363 m, are required; three are of 1 m diameter and the other is of 0.9 m diameter. Table 6.19 breaks down the cost estimates for meeting the current adaptation deficit. As shown, the costs for added temporary-pump capacity and pipes are Tk. 27.6 million and Tk. 317.8 million, respectively, for a total estimated adaptation cost of Tk. 345.4 million. Adapting to climate change would mean increasing Hazaribag’s temporary pump capacity. Table 6.20 breaks down the total additional cost for addressing climate change adaptation, estimated at Tk. 16.2 million. Finally, table 6.21 provides a summary comparison of the total costs to meet the current adaptation deficit and the added cost of climate change adaptation. As shown, addressing the current deficit would cost Tk. 345.4 million, while meeting the climate change deficit would require another Tk. 16.2 million, for a total combined adaptation cost of Tk. 361.6 million (Huq 2014). The adaptation cost estimate for the area’s drainage system excludes the hydrological modeling exercise’s recommended deepening of water bodies, which would provide a one- time income source rather than an added expense. Chapter 8 discusses Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Assessment of Dhaka’s Modeled Zones 133 this proposed activity in detail. The damage estimates demonstrate that this investment of about Tk. 362 million in flood mitigation will lead to a much higher reduction in damage of Tk. 1,410 million in 2050 when faced with a ­ 100-year, return-period storm and a cumulative damage reduction (2014–50) of Tk. 19,562 million. Goranchatbari System Results Hydrological Modeling Output Simulations of inundation depths and flood duration were run for the historic 100-year, return-period rainfall event in 2004 and alternative scenarios for 2050 with and without climate change. The average duration of the actual 100-year flood in 2004 was 15 hours. For 2050, alternative scenarios were simulated using flood criteria of 0.25 m (i.e., 90 percent of the administrative areas with a maxi- mum inundation depth less than 0.25 m and duration less than 12 hours). Comparison of the 100-year rainfall event in 2050 with the actual 2004 storm showed a marked improvement in the flooding situation (map 6.4), with Map 6.4 Changes in Extent of Flooding for Extreme Rainfall Event over Study Period, Goranchatbari System Model Source: IWM 2014. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 134 Assessment of Dhaka’s Modeled Zones Table 6.22 Comparison of Flooding Results, Goranchatbari System Model Extreme storm Maximum flooded area (km2) Average duration above 0.25 m (341 mm/24 hour) Flood free (0–0.25 m) 0.25–0.75 m 0.75–1.5 m >1.5 m depth (hours to recede) 2004 (actual storm) 45.3 3.0 4.5 11.0 15 2050 (SE + PI) 56.1 2.5 1.7 3.5 6 Source: Zaman 2014. significantly less flooding depth and average duration of flooding (table 6.22). The improved performance is made possible by the laying of new drainage pipes, doubling the capacity of the Goranchatbari Pump Station, and better drainage to the pump’s retention pond. The raising of land levels with the ongoing process of urbanization in the area also contributes to the reduced extent of flooding. However, even with the planned improvements (PI), the analysis showed that some areas were unable to cope with extreme rainfall events (i.e., 341 mm per 24 hours). The extent of shallow submergence remained significantly higher in 2050 than in 2004. Therefore, additional investments (AI) were incorporated into the model to address the existing adaptation deficit. The additional investments identified to meet the current adaptation deficit were as follows. • From the airport to Baunia Khal and Uttara Model Town Lake—Lay two new pipes, with a total length of 5,325 m and a 1.524 m diameter. • From Pallabi—Redirect flow from catchment 403 (ward 15) to Baunia Khal. • At Mirpur Zoo and Botanical Garden—Deepen ponds by excavating about 702,400 m3. Simulating the likely impacts of these additional investments in the flood model showed they were insufficient to meet the acceptable flooding criteria under the A1FI climate change scenario (table 6.23). Thus, to cope with the impact of increased rainfall intensity due to climate change, supplementary adap- tation measures were required. These were identified as follows: • From Pallabi to Diabari Khal—Lay a new pipe, with a total length of 518 m and a pipe diameter of 2 m. • From Catchment 272_1 (ward 4)—Divert flow to Diabari Khal. Vulnerability to Urban Flooding Goranchatbari is somewhat vulnerable to urban flooding, with three of its eight areas (i.e., seven wards and Uttara, which is outside the DCC wards) included among Dhaka’s top 20 vulnerable wards/regions. All three vulnerable wards— wards 2, 4, and 15—have lower-than-average rankings for all four dimensions in the CDRI index (physical, social, economic, and institutional). These wards also have higher-than-average flood vulnerability; when combined with their lower CDRI indices, they become quite vulnerable and thus place among the Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Assessment of Dhaka’s Modeled Zones 135 Table 6.23  2050 Rainfall Event in a Changing Climate, Goranchatbari System Model Maximum flooded area (km2) Average duration Flood free above 0.25 m depth Extreme storm in 2050 (0–0.25 m) 0.25–0.75 m 0.75–1.5 m >1.5 m (hours to recede) Future SE + PI (341 mm/24 hour) 56.0 2.5 1.7 3.5 6 Future SE + PI + CC (396 mm/24 hour) 55.1 3.2 1.7 4.1 7 Future SE + PI + AI (341 mm/24 hour) 57.4 1.6 1.2 3.5 4 Future SE + PI + AI + CC (396 mm/24 hour) 56.5 2.0 1.3 3.9 5 Future SE + PI + AI + CC + AD (396 mm/24 hour) 56.7 1.9 1.2 3.9 4 Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; AD = adaptation to climate change. Table 6.24  FVI Rankings of Goranchatbari Wards/Region under Alternative Flooding Scenarios Future scenarios Equal weight a Ward/region SE + PI SE + PI + CC SE + PI + AI SE + PI + AI + CC SE + PI + AI + CC + AD 15 14 11 15 14 16 02 16 13 17 11 17 04 17 19 8 10 11 05 22 20 13 15 14 06 28 28 18 17 20 03 42 40 36 38 36 01 60 60 58 58 58 Uttara 94 94 94 94 94 50% flood weight a 15 18 16 15 14 12 02 21 18 18 12 14 04 20 19 7 9 10 05 25 24 13 16 13 06 28 27 17 15 16 03 57 52 52 54 52 01 71 69 66 67 66 Uttara 94 94 94 94 94 Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; AD = adaptation to climate change; FVI = Flood Vulnerability Index. a. Lower numbers indicate a higher vulnerability. top 20 vulnerable wards/regions under planned improvements with equal weight (table 6.24). Vulnerability to urban flooding and the four CDRI dimensions play equally important roles in increasing vulnerability in Goranchatbari. The area’s vulnera- bility rankings generally are little changed as greater weight is assigned to flood- ing vulnerability under the 50 percent flood weight or AI scenarios, where the effect of flooding decreases (table 6.24). Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 136 Assessment of Dhaka’s Modeled Zones Damage Estimates The total damage for Goranchatbari in 2050 was estimated for the five future scenarios, using alternative assumptions about how frequently a 100-year, return- period storm would occur (chapter 5). Figure 6.8 shows the total monetary value of the damage that would result from the actual 100-year event in 2050 under the five future scenarios, followed by expected damage estimates using alternative probabilities of the 100-year event’s recurrence (i.e., every 20, 33, 50, or 100 years). Figure 6.9 shows the expected cumulative damage over the 2014–50 period for the five future scenarios, with random assignment of various return-period storms under alternative assumptions of increased frequency of the 100-year event. As described in chapter 5, alternative probabilities of the recurrence of the 100-year event are used to compute the expected value of the total cumulative damage between 2014 and 2050 if the event were to occur every 20, 33, 50, or 100 years. The results reveal how the damage that would occur under the scenario with planned improvements (PI) alone decreases with additional investments (AI)— but less so than for the other modeled zones—and how climate change exacer- bates the extent of damage under each scenario. In Goranchatbari, unlike other Figure 6.8 Total Estimated Damage for Goranchatbari in 2050 1,000 800 600 Million Tk. 400 200 0 100-year event RP20 RP33 RP50 RP100 Assumptiona SE + PI SE + PI + Al SE + PI + Al + CC + AD SE + PI + CC SE + Al + PI + CC Source: Roy 2014. Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; AD = adaptation to climate change. RP = return period (years). a. 100-year event = actual damage from a 100-year event occurring in 2050 with certainty; RP20, 33, 50, and 100 = damage from a 100-year event occurring once every 20, 33, 50, or 100 years, respectively. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Assessment of Dhaka’s Modeled Zones 137 Figure 6.9 Total Cumulative Damage for Goranchatbari, 2014–50 16,000 12,000 Million Tk. 8,000 4,000 0 RP20 RP33 RP50 RP100 Assumptiona SE + PI SE + PI + Al SE + PI + Al + CC + AD SE + PI + CC SE + Al + PI + CC Source: Roy 2014. Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; AD = adaptation to climate change. RP = return period (years). a. 100-year event = actual damage from a 100-year event occurring in 2050 with certainty; RP20, 33, 50, and 100 = damage from a 100-year event occurring once every 20, 33, 50, or 100 years, respectively. areas of Dhaka, the increased damage caused by climate change is higher with AI, compared to the scenario with PI. With climate change adaptation, the dam- age decreases somewhat, but is still larger than that under the scenario with AI without climate change. Table 6.25 captures the increased damage caused by climate change under the planned improvements (PI) and additional investments (AI) scenarios for the damage occurring in 2050 and the cumulative damage in 2014–50 using various alternative assumptions. Thus, in 2050, the increased damage from a 100-year event due to climate change would amount to Tk. 224 million if only planned improvements were made and would rise to Tk. 339 million with AI. Table 6.25 shows similar outcomes of increased damage caused by climate change under a weighted probability of various return-period storms for both PI and AI s­ cenarios. It also shows similar cumulative effects of climate change in 2014–50, with ran- dom assignment of various return-period storms under alternative assumptions of increased frequency of a 100-year, return-period storm. In terms of sector-wise comparisons for Goranchatbari, the largest damage is to residential buildings, followed by industry and health care. Cost of Adaptation Meeting the current adaptation deficit to safeguard Goranchatbari against a 100- year, return-period rainfall event in 2050 beyond planned improvements involves Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 138 Assessment of Dhaka’s Modeled Zones Table 6.25 Increased Damage in Goranchatbari from Climate Change for Future Scenarios under Alternative Assumptions Increased damage occurring in 2050 (Tk. million) Return period based on probability Scenario 100 year RP20 RP33 RP50 RP100 SE + PI + CC 224 213 175 154 126 SE + PI + AI + CC 339 266 231 212 186 Increased cumulative damage, 2014–50 (Tk. million) Using random numbers for each return period Scenario RP20 RP33 RP50 RP100 SE + PI + CC 3,687 3,603 3,339 3,231 SE + PI + AI + CC 5,527 5,393 4,989 4,817 Source: Roy 2014. Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; RP = return period (years). Table 6.26 Estimated Cost of Meeting the Adaptation Deficit, Goranchatbari Laying new pipe Pipe Pipe Pipe Pipe installation Road-cutting Total (location/link) diameter (m) length (m) price/m cost/m cost/m cost Meeting current adaptation deficit (Tk. million) From airport to Baunia Khal and Uttara Model Town 1.52 5,325 15,295 26,400 15,100 302.4 Meeting climate change adaptation deficit (Tk. million) From Pallabi to Diabari Khal 2.00 518 30,600 50,000 17,500 50.8 Sources: DWASA officials and authors’ calculations. only laying of a new drainage pipe (Zaman 2014). The total cost of laying the new pipe, which would run from the airport to Baunia Khal and Uttara Model Town, is estimated at Tk. 302.4 million (table 6.26). Addressing the climate change adaptation deficit would require only a 2 m diameter pipe from Pallabi to Diabari Khal. Because of space constraints, the modeling study recommends laying two 1.52 m diameter pipes, one atop the other, which would have the equivalent discharge capacity of a 2 m diameter pipe (Zaman 2014). The cost of laying this pipe is estimated at Tk. 50.8 million (table 6.26). The total combined adaptation cost for Goranchatbari is Tk. 353.2 million (table 6.27) (Huq 2014). The estimate excludes the hydrological modeling exer- cise’s recommended deepening of ponds at the Mirpur Zoo and Botanical Garden, which should provide a one-time source of income rather than an added expense. The damage estimates demonstrate that this investment of about Tk. 353 million in flood mitigation will lead to a reduction in damage of Tk. 359 million in 2050 when faced with a 100-year, return-period storm and a cumula- tive damage reduction (2014–50) of Tk. 5,062 million. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Assessment of Dhaka’s Modeled Zones 139 Table 6.27 Summary of Adaptation Costs, Goranchatbari Recommended measure Cost (Tk. million) Addressing current adaptation deficit Total estimated cost, laying new drainage pipe 302.4 Addressing climate change deficit Total estimated cost, laying new drainage pipe 50.8 Total combined cost 353.2 Sources: DWASA officials and authors’ calculations. Eastern Dhaka Results Hydrological Modeling Output Simulations of inundation depths and duration of river flood were run for the historic 100-year, return-period rainfall event in 2004 and 2050 alternative sce- narios with and without climate change. For 2050, alternative scenarios were simulated using flood criteria of 0.25 m (i.e., 90 percent of the administrative areas with a maximum flood depth less than 0.25 m and duration less than 12 hours). Table 6.28 shows a significant decrease in flooded areas for all inundation depths in 2050, including more than a threefold increase in flood-free area; however, the time required for flooding to recede decreased only slightly compared to the 2004 event. Analysis of the flood maps generated for the extreme rainfall event in 2050 using the 0.25 flood criteria showed that the planned improvements (PI)—­ protection from river flooding, increased conveyance capacity of khals, introduc- tion of drainage pumps, and raising of land levels—if completed as planned and on time, could cope with the 100-year storm (341 mm per 24 hours). Thus, no additional investments (AI) would be required for the Eastern Dhaka area. Furthermore, the PI could cope with effect of climate change (396 mm per 24 hours) (map 6.5). Thus, no additional adaptation measures would be required for Eastern Dhaka (table 6.29). These results highlight the effectiveness of the planned land use and drainage improvements and their value in reducing the vulnerability of this area to the impacts of climate change. Table 6.28 Comparison of Flooding Results, Eastern Dhaka Model Maximum flooded area (km2) Average duration above Flood free 0.25 m depth (hours to Scenario (0–0.25 m) 0.1–0.25 m 0.25–0.75 m 0.75–1.5 m >1.5 m recede) 2004 (actual storm, 341 mm/24-hour) 34.40 2.87 7.89 10.81 64.73 >240 2050 100-year event (SE + PI, 341 mm​ /24-hour) 104.8 1.1 2.2 4.0 8.6 234 Source: Zaman 2014. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 140 Assessment of Dhaka’s Modeled Zones Map 6.5  Flood-Depth Comparisons of Extreme Rainfall Scenarios in 2050, Eastern Dhaka Model Source: IWM 2014. Table 6.29  2050 Rainfall Event in a Changing Climate, Eastern Dhaka Model Maximum flooded area (km2) Average duration Flood free above 0.25 m depth Extreme storm in 2050 (0–0.25 m) 0.25–0.75 m 0.75–1.5 m >1.5 m (hours to recede) Future SE + PI (341 mm/24 hour) 106.0 2.24 3.97 8.58 >240 Future SE + PI + CC (396 mm/24 hour) 104.6 2.73 2.91 10.59 >240 Source: Zaman 2014. Note: SE = socioeconomic changes; PI = planned improvements; CC = climate change. Additional scenarios were simulated for 2050, using various assumptions regarding planned improvements (PI), including changes in the drainage system, land levels, and land use. The first simulation assumed that only socioeconomic changes (SE) would occur by 2050, with no PI to the drainage system (i.e., no river embankment, regulators, or pumps). The results showed that, for the 30-year event with A1FI climate change, more than 80 percent of the area would remain flood free. This finding is attributed to the raising of land levels in Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Assessment of Dhaka’s Modeled Zones 141 Figure 6.10  Flood-Depth Simulations for 30-Year Rainfall Event in Eastern Dhaka, Showing Effects of Planned Improvements 100 80 Percent of area 60 40 20 0 Flood free 0–0.1 0.1–0.25 0.25–0.75 0.75–1.5 >1.5 Flood depth (m) 2004 base year 2050 lands levels and 2004 lands levels and land use with PI and CC land use with PI Source: Zaman 2014. built-up areas that is likely to occur and prevention of encroachment of khals and ponding areas. A second simulation asked what would have happened in 2004 if planned improvements (PI) had been made to the drainage system without changing 2004 land levels and land use. Results for the 30-year event with A1FI climate change showed that, in this case, about 59 percent of the area would have remained flood-free (compared to 24 percent for 2004 base case). By raising land levels, along with making the drainage PI, the percentage of flood-free area would reach 90 percent by 2050 (figure 6.10). These findings highlight the value of raising land levels and planned development in reducing the extent of flooding in Eastern Dhaka. Vulnerability to River Flood For Eastern Dhaka, the flood vulnerability estimate covers two regions outside the DCC. Since larger proportions of the DCC wards that lie in Eastern Dhaka are found in the adjacent areas of Central Dhaka and DND, the relevant wards are covered in the flood vulnerability assessments of those areas and thus are excluded here. Since both of Eastern Dhaka’s outside regions figure among Dhaka’s top 20 vulnerable wards/regions, the Eastern Dhaka area figures high on the FVI. The two outside regions rank low in terms of the CDRI’s institutional and social dimensions but exhibit high flood vulnerability; when combined with their other lower CDRI indices, these areas become quite vulnerable and thus Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 142 Assessment of Dhaka’s Modeled Zones Table 6.30  FVI Rankings of Eastern Dhaka Regions under Alternative Flooding Scenarios Future scenarios Equal weight a SE + PI + SE + PI + AI + Region SE + PI SE + PI + CC SE + PI + AI AI + CC CC + AD Badda, outside DNCC 12 8 n.a. n.a. n.a. Khilgaon, outside DNCC 10 14 n.a. n.a. n.a. 50% flood weight a Badda, outside DNCC 8 8 n.a. n.a. n.a. Khilgaon, outside DNCC 3 2 n.a. n.a. n.a. Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; AD = adaptation to climate change; n.a. = not applicable. a. Lower numbers indicate a higher vulnerability. place among Dhaka’s top 20 vulnerable wards/regions under planned improve- ments (PI) for equal weight (table 6.30). Unlike other areas, only two scenarios were created for Eastern Dhaka. Flood vulnerability and the four CDRI dimensions play equally important roles in increasing Eastern Dhaka’s vulnerability. The area’s vulnerability rankings gener- ally increase slightly as greater weight is assigned to flood vulnerability for 50 percent flood weight. Damage Estimates The total damage for Eastern Dhaka in 2050 was estimated for two future scenarios—planned improvements with and without climate change—using ­ alternative assumptions about how frequently a 100-year, return-period storm would occur (chapter 5). Figure 6.11 shows the total monetary value of the dam- age that would result from the actual 100-year event in 2050 for these two future scenarios, followed by expected damage estimates using alternative prob- abilities of the 100-year event’s recurrence (i.e., every 20, 33, 50, or 100 years). Figure 6.12 shows the expected cumulative damage over the 2014–50 period for the two future scenarios, with random assignment of various return-period storms under alternative assumptions of increased frequency of the 100-year event. As described in chapter 5, alternative probabilities of the recurrence of the 100- year event are used to compute the expected value of the total cumulative damage between 2014 and 2050 if the event were to occur every 20, 33, 50, or 100 years. The findings show how the damage that would occur under the scenario with planned improvements increases with climate change. Table 6.31 captures the increased damage caused by climate change under the planned improvements (PI) scenario for the damage occurring in 2050 and the cumulative damage in 2014–50 using various alternative assumptions. Thus, in 2050, the increased damage from a 100-year event due to climate change would amount to Tk. 34 million if PI were made. Table 6.31 shows similar outcomes of increased damage caused by climate change under a weighted probability of Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Assessment of Dhaka’s Modeled Zones 143 Figure 6.11 Total Estimated Damage for Eastern Dhaka in 2050 140 120 100 Million Tk. 80 60 40 20 0 100-year event RP20 RP33 RP50 RP100 Assumptiona SE + PI SE + PI + CC Source: Roy 2014. Note: SE = socioeconomic changes; PI = planned improvements; CC = climate change. RP = return period (years). a. 100-year event = actual damage from a 100-year event occurring in 2050 with certainty; RP20, 33, 50, and 100 = damage from a 100-year event occurring once every 20, 33, 50, or 100 years, respectively. Figure 6.12 Total Cumulative Damage for Eastern Dhaka, 2014–50 2,500 2,000 1,500 Million Tk. 1,000 500 0 RP20 RP33 RP50 RP100 Assumptiona SE + PI SE + PI + CC Source: Roy 2014. Note: SE = socioeconomic changes; PI = planned improvements; CC = climate change. RP = return period (years). a. 100-year event = actual damage from a 100-year event occurring in 2050 with certainty; RP20, 33, 50, and 100 = damage from a 100-year event occurring once every 20, 33, 50, or 100 years, respectively. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 144 Assessment of Dhaka’s Modeled Zones Table 6.31 Increased Damage in Eastern Dhaka from Climate Change for Future Scenarios under Alternative Assumptions Increased damage occurring in 2050 (Tk. million) Return period based on probability Scenario 100 year RP20 RP33 RP50 RP100 SE + PI + CC 34 22 21 20 19 Increased cumulative damage, 2014–50 (Tk. million) Using random numbers for each return period Scenario RP20 RP33 RP50 RP100 SE + PI + CC 565 552 511 494 Source: Roy 2014. Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; RP = return period (years). v ­ arious return-period storms for the PI scenario. It also shows similar cumulative effects of climate change in 2014–50, with random assignment of various return-period storms under alternative assumptions of increased frequency of a ­ 100-year, return-period storm. In terms of sector-wise comparisons for Eastern Dhaka, the largest damage is found in industry, followed by other surface transport and residential buildings. ­ DND Results Hydrological Modeling Output The study generated flood maps showing inundation depths and duration of urban flooding for the historic 100-year, return-period rainfall event in 2004 and alterna- tive scenarios for 2050 with and without climate change using the 0.25 m flood criteria. The actual 100-year, return-period storm in 2004 inundated more than 70 percent of the DND area. Only 12.25 km2 remained flood free, while 11.42 km2 experienced inundation above 1.5 m depth. Compared to the 2004 storm, the simulated 2050 100-year event showed a decrease in deeper pockets of waterlog- ging (>1.5 m depth), even for the A1FI climate change scenario (table 6.32). The improved flooding situation is attributable to planned drainage improvements (i.e., increased conveyance capacity of khals, installation of new pump stations, and increased capacity at Shimrail Pump Station), as well as greater land elevation to accommodate future population growth in built-up areas. For the 100-year event in 2050 without climate change (341 mm per 24 hours), the flood-free area is increased to 48.64 km2, meaning that about 84 percent of the modeled area remains flood free (table 6.32). When the effect of A1FI climate change (396 mm per 24 hours) is simulated, the DND’s flood- free area is reduced to 66 percent. About three-fifths of the area is inundated at 0.25 m depth, while another 15 percent is under more than 1.5 m flood depth—­​ 9 percent more than in the simulation without climate change (map 6.6). Because these findings showed that planned improvements would be insufficient to cope with extreme rainfall events, even without climate change, ­ Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Assessment of Dhaka’s Modeled Zones 145 Table 6.32  Flood-Depth Comparisons, DND Model Maximum flooded area (km2) Scenario simulated Flood free (0–0.1 m) 0.1–0.25 m 0.25–0.75 m 0.75–1.5 m >1.5 m 2004 100-year event (actual storm, 341 mm/24-hour) 12.25 4.59 11.33 19.55 11.42 2050 100-year event (SE + PI, 341 mm/24-hour) 48.64 1.59 2.60 2.83 3.52 2050 100-year event (SE + PI + CC, 396 mm/24-hour) 1.20 37.70 5.68 5.61 8.99 Source: Zaman 2014. Note: DND = Dhaka-Narayanganj-Demra. Map 6.6  Flood-Depth Comparisons of Extreme Rainfall Events in 2050, DND Model Source: IWM 2014. additional measures were incorporated into the model to meet acceptable flood criteria for the DND study area. The following additional investments were identified to address the current adaptation deficit: • At Shimrail Pump Station—Increase planned pump capacity by 10 percent to 42.25 m3/s. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 146 Assessment of Dhaka’s Modeled Zones Table 6.33  2050 Rainfall Event in a Changing Climate, DND Model Area Maximum flooded area (km2) Average duration above 0.25 m Extreme storm in 2050 Flood free (0–0.25 m) 0.25–0.75 m 0.75–1.5 m >1.5 m depth (hours to recede)a Future SE + PI (341 mm/24 hour) 48.6 1.6 2.6 2.8 35 Future SE + PI + CC (396 mm/24 hour) 1.2 37.7 5.7 5.6 77 Future SE + PI + AI (341 mm/24 hour) 51.5 1.2 2.0 2.1 20 Future SE + PI + AI + CC (396 mm/24 hour) 43.2 2.4 3.8 4.0 64 Future SE + PI + AI + CC + AD (396 mm/24 hour) 51.3 1.3 2.1 2.2 34 Source: Zaman 2014. Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; AD = adaptation to climate change; DND = Dhaka-Narayanganj-Demra. a. Analysis of depth duration could not be calculated directly since drainage-system simulations used the 1-dimensional hydrodynamic model. Thus, for each administrative area, flooding duration was estimated by analyzing the change in water levels at appropriate points in the drainage system. For flood depths above 1.5 m, duration calculations used GIS analysis and generated area elevation curves for each administrative area. Lowest land level was estimated by comparing the total water body area (proposed by the DAP) and area elevation curve. Details are provided in Zaman (2014). • At Adamjinagar Station—Increase planned pump capacity by 20 percent to 36.25 m3/s. • At Pagla Station—Increase planned pump capacity by 20 percent to 3.84 m3/s. Even with these additional investments, the findings showed that, with cli- mate change, about 23 percent of the DND study area would be inundated, and some areas would experience more than 0.25 m inundation over 12 hours (table 6.33). Thus, further adaptation measures were needed to meet acceptable flooding criteria under the A1FI climate change scenario in 2050. The specific adaptation measures required to cope with climate change were identified as follows: • At Shimrail Pump Station—Increase pump capacity by another 30 percent to 54.92 m3/s. • At Adamjinagar Pump Station—Increase pump capacity by another 30 per- cent to 47.13 m3/s. • At Fatulla Pump Station—Increase planned pump capacity by 25 percent to 4.06 m3/s. With the implementation of these adaptation measures, 90 percent of the DND area will be flood-free. Additional 2050 Scenarios Owing to political, social, and other pressures, it may be difficult to maintain the recommended design of khal cross-sections in the future (BWDB 2010). Thus, additional simulations including bank encroachment of khals were run for Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Assessment of Dhaka’s Modeled Zones 147 Table 6.34  2050 Flood-Depth Comparisons in a Changing Climate, Highlighting the Impact of Khal Encroachment Maximum flooded area (km2) Flood free Scenario (0–0.1) 0.1–0.25 m 0.25–0.75 m 0.75–1.5 m >1.5 m 211 mm/6 hour rainfall event Without climate change 57.61 0.27 0.34 0.24 0.71 Without climate change + encroachment 57.05 0.42 0.53 0.39 0.79 With 16% rainfall increase due to climate change 56.63 0.50 0.72 0.48 0.84 With 16% rainfall increase due to climate change + encroachment 54.57 0.85 1.33 1.15 1.28 341 mm/24 hour rainfall event Without climate change 48.64 1.59 2.60 2.83 3.52 Without climate change + encroachment 1.51 0.21 0.48 43.08 13.90 With 16% rainfall increase due to climate change 1.20 37.70 5.68 5.61 8.99 With 16% rainfall increase due to climate change + encroachment 0.12 0.08 0.16 0.21 58.60 Source: Zaman 2014. various 2050 scenarios (30- and 100-year rainfall events with and without cli- mate change) (table 6.34). Results of simulations for the 211 mm per six-hour rainfall event (30-year, return-period design storm) show little change in the extent of inundation due to the effect of climate change, even with bank encroachment (table 6.34). However, findings for the 341 mm per 24-hour event show that no area would remain flood-free. Without climate change, about 72 percent of the area would experi- ence inundation at 0.75–1.5 m depth, while about 23 percent would be under 1.5 m water depth. With A1FI climate change, the situation would worsen markedly, with nearly 99 percent of the DND study area inundated at a depth above 1.5 m. Interestingly, the findings show that bank encroachment has a significant impact on the 2050 scenarios for the 341 mm per 24-hour event. Without cli- mate change or encroachment, about 80 percent of the DND study area would remain flood free. The addition of encroachment would have a significant impact, resulting in about 70 percent of the area being inundated at a 0.75–1.5 m depth. For the A1FI climate change scenario, the impact is even greater. Without encroachment, about 60 percent of the area would be inundated at 0.25 m water depth. With encroachment, however, nearly 100 percent of the DND study area would experience waterlogging above 1.5 m depth. Vulnerability to River Flood The DND area includes five wards inside the DCC and one outside region to the south, adjacent to the DSCC wards. Wards 86–88, 90 and the outside region are Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 148 Assessment of Dhaka’s Modeled Zones among Dhaka’s top 20 vulnerable wards/regions, making DND the most vulner- able of the six areas analyzed for flood vulnerability. All five wards—86, 87, 88, 89, and 90—rank low in terms of the CDRI’s economic and physical dimensions, while the outside region ranks low for the institutional and physical dimensions. In terms of the flood dimension, four of the five wards—86, 87, 88, and 90—and the outside region show high flood vulnerability as well; when combined with their other lower CDRI indices, they become quite vulnerable, placing among the top 20 vulnerable wards/regions under planned improvements (PI) scenarios for equal weights (table 6.35). High flood vulnerability is a significant factor in increasing vulnerability in the DND area; however, the four CDRI dimensions play an even larger role. This impact is evidenced by the general increase in vulnerability rankings for the DND wards/region as flooding levels go down with the transition from planned improvements to additional investments scenarios under equal weights and 50 percent flood weight. Damage Estimates The total damage for the DND area in 2050 was estimated for the five future scenarios, using alternative assumptions about how frequently a 100-year, return- period storm would occur (chapter 5). Figure 6.13 shows the total monetary value of the damage that would result from the actual 100-year event in 2050 under the five future scenarios, followed by expected damage estimates using Table 6.35  FVI Rankings of DND Wards/Region under Alternative Flooding Scenarios Equal weight a Future scenarios SE + PI + SE + PI + AI + Ward/region SE + PI SE + PI + CC SE + PI + AI AI + CC CC + AD Outside DCC, south area 15 12 7 5 10 86 5 3 1 1 2 87 2 29 3 3 1 88 8 39 10 7 4 89 52 52 49 50 48 90 9 41 11 9 8 50% flood weight a Outside DCC, south area 10 9 3 2 4 86 7 4 1 1 1 87 9 26 4 4 2 88 14 46 8 7 6 89 65 61 64 63 62 90 15 55 9 8 8 Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; AD = adaptation to climate change; DND = Dhaka-Narayanganj-Demra; FVI = Flood Vulnerability Index. a. Lower numbers indicate a higher vulnerability. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Assessment of Dhaka’s Modeled Zones 149 Figure 6.13 Total Estimated Damage for DND in 2050 600 500 400 Million Tk. 300 200 100 0 100-year event RP20 RP33 RP50 RP100 Assumptiona SE + PI SE + PI + Al SE + PI + Al + CC +AD SE + PI + CC SE + Al + PI + CC Source: Roy 2014. Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; AD = adaptation to climate change. RP = return period (years). a. 100-year event = actual damage from a 100-year event occurring in 2050 with certainty; RP20, 33, 50, and 100 = damage from a 100-year event occurring once every 20, 33, 50, or 100 years, respectively. alternative probabilities of the 100-year event’s recurrence (i.e., every 20, 33, 50, or 100 years). Figure 6.14 shows the expected cumulative damage over the 2014–50 period for the five future scenarios, with random assignment of various return-period storms under alternative assumptions of increased frequency of the 100-year event. As described in chapter 5, alternative probabilities of the recurrence of the 100-year event are used to compute the expected value of the total cumulative damage between 2014 and 2050 if the event were to occur every 20, 33, 50, or 100 years. The results show how the damage that would occur under the scenario with planned improvements (PI) alone is reduced when additional investments (AI) are made to address the current adaptation deficit and how climate change exac- erbates the extent of damage for each scenario. Comparison of the damage in a changing climate with and without AI reveals that increase in the damage due to climate change is less if AI are made and the current adaptation deficit is addressed. With measures to address climate change adaptation, the damage in a changing climate decreases somewhat, but is still larger than in the AI scenario without climate change. Table 6.36 captures the increased damage caused by climate change under the planned improvements (PI) and additional investments (AI) scenarios for the Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 150 Assessment of Dhaka’s Modeled Zones Figure 6.14 Total Cumulative Damage for DND, 2014–50 8,000 6,000 Million Tk. 4,000 2,000 0 RP20 RP33 RP50 RP100 Assumptiona SE + PI SE + PI + Al SE + PI + Al + CC +AD SE + PI + CC SE + Al + PI + CC Source: Roy 2014. Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; AD = adaptation to climate change. RP = return period (years). a. 100-year event = actual damage from a 100-year event occurring in 2050 with certainty; RP20, 33, 50, and 100 = damage from a 100-year event occurring once every 20, 33, 50, or 100 years, respectively. Table 6.36 Increased Damage in DND from Climate Change for Future Scenarios under Alternative Assumptions Increased damage occurring in 2050 (Tk. million) Return period based on probability Scenario 100 year RP20 RP33 RP50 RP100 SE + PI + CC 348 224 205 195 181 SE + PI + AI + CC 228 149 137 130 121 Increased cumulative damage, 2014–50 (Tk. million) Using random numbers for each return period Scenario RP20 RP33 RP50 RP100 SE + PI + CC 5,574 5,427 5,003 4,811 SE + PI + AI + CC 3,687 3,593 3,317 3,195 Source: Roy 2014. Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; RP = return period (years); DND = Dhaka-Narayanganj-Demra. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Assessment of Dhaka’s Modeled Zones 151 damage occurring in 2050 and the cumulative damage in 2014–50 using various alternative assumptions. Thus, in 2050, the increased damage from a 100-year event due to climate change would amount to Tk. 348 million if only PI were made and would decrease to Tk. 228 million with AI. Table 6.36 shows similar outcomes of increased damage caused by climate change under a weighted prob- ability of various return-period storms for both PI and AI scenarios. It also shows similar cumulative effects of climate change in 2014–50, with random assign- ment of various return-period storms under alternative assumptions of increased frequency of a 100-year, return-period storm. In terms of sector-wise compari- sons for the DND area, the largest damage is to residential property, followed by industry and residential buildings. Cost of Adaptation Results of the hydrological modeling exercise show that, for the DND drainage system, the cost of meeting the current adaptation deficit beyond planned improvements (PI) would require increasing the capacity of permanent pump stations at Shimrail and Adamjinagar by 3.85 m3/s and 6.05 m3/s, respectively. It would also require increasing the temporary pump capacity at Pagla by 0.65 m3/s (Zaman 2014). Among the PI, only the Shimrail Pump Station has been installed to date. This study’s cost estimates assume that all PI will be made. The costs of these additional investments (AI) are estimated at Tk. 540.0 million and Tk. 4.6 million, for a total cost of Tk. 544.6 million (table 6.37). Table 6.37 Estimated Cost of Meeting the Current Adaptation Deficit, DND Pump Electrical Civil Total Pump price/unit Mechanical works works cost capacity Head (million works (million (million (million Increased pump capacity (m3/s) (m) Tk.) Unit (no.) (million Tk.) Tk.) Tk.) Tk.) Shimrail Permanent pump station 3.85 6 80 1 80 65 70 215 Adamjinagar Permanent pump station 6.05 6 150 1 150 75 100 325 Subtotal 540 Pagla 25 ft3/s Unit cost pumps of pump Total Pump (pumps and pipes 200 kVA Unit cost of 200 kVA cost capacity Head and pipes) (million transformers transformers and (million Increased pump capacity (m3/s) (m) (no.) Tk.) needed (no.) cables (million Tk.) Tk.) Temporary pump station 0.65 10 1 2.6 1 2.0 4.6 Subtotal 4.6 Total cost 544.6 Sources: DWASA officials and authors’ calculations. Note: DND = Dhaka-Narayanganj-Demra. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 152 Assessment of Dhaka’s Modeled Zones Adapting to climate change would require further enhancement of the per- manent pump capacity at Shimrail and Adamjinagar by 12.7 m3/s and 10.9 m3/s, respectively. In addition, the temporary pump capacity at Fatulla would need to be increased by 0.82 m3/s. The cost of these adaptation measures to address cli- mate change would total nearly Tk. 1.11 billion (table 6.38). Finally, table 6.39 provides a summary comparison of the costs required to meet the current adaptation deficit and the added expense of climate change adaptation. As shown, Tk. 544.6 million would be required to close the current gap, while addressing the climate change deficit would require an additional Tk. 1.1 billion, for a total combined cost of Tk. 1.65 billion. It should be noted that the figure for DND’s climate change adaptation is far higher than those Table 6.38 Estimated Cost of Meeting the Climate Change Adaptation Deficit, DND Pump Electrical Total Pump price/unit Mechanical works cost capacity Head (million Unit works (million Civil works (million Increased pump capacity (m3/s) (m) Tk.) (no.) (million Tk.) Tk.) (million Tk.) Tk.) Shimrail Permanent pump station 12.7 6 190 2 380 80 140 600 Adamjinagar Permanent pump station 10.9 6 150 2 300 75 125 500 Subtotal 1,100 Fatulla 5 ft3/s pumps 25 ft3 /s pumps 200 kVA Total Pump (pumps and (pumps and pipes), transformers and cost capacity Head pipes), 0.8 per 2.6 per unit cables, 2.0 per (million Increased pump capacity (m3/s) (m) unit (million Tk.) (million Tk.) unit (million Tk.) Tk.) Temporary pump station 0.82 10 0.8 2.6 2.0 5.4 Subtotal 5.4 Total cost 1,105.4 Sources: DWASA officials and authors’ calculations. Note: DND = Dhaka-Narayanganj-Demra. Table 6.39 Summary of Adaptation Costs, DND Recommended measure Cost (Tk. million) Addressing current adaptation deficit Total estimated cost, increasing pump capacity 544.6 Addressing climate change deficit Total estimated cost, increasing pump capacity 1,105.4 Total combined cost 1,650.0 Sources: BWDB, DWASA officials, and authors’ calculations. Note: DND = Dhaka-Narayanganj-Demra. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Assessment of Dhaka’s Modeled Zones 153 for the other drainage systems in the detailed study area. In fact, DND accounts for more than 85 percent of the total Tk. 1.3 billion required to address the climate change–induced adaptation deficit (Huq 2014). The damage estimates demonstrate that this investment of Tk. 1,650 million in flood mitigation will lead to a reduction in damage of Tk. 352 million in 2050 when faced with a 100-year, return-period storm and cumulative damage reduction (2014–50) of Tk. 4,868 million. Narayanganj Results Hydrological Modeling Output River flood maps were generated simulating the inundation depth and duration of 2004 rainfall events (10-, 30-, and 100-year return-periods) and alternative 2050 scenarios with and without climate change. In 2004, the flooding extent and duration for the 30-year rainfall event is higher than for the 10-year event; however, the difference in flooding is minimal. By comparison, the actual 100- year, return-period storm exhibited more extensive flooding, with greater depth and prolonged duration. The average duration for the actual 100-year event to recede was 12 hours, compared to 3 hours and 2 hours, respectively, for the 30-year and 10-year events (table 6.40). For 2050, the study modeled 211 mm per six-hour event (equivalent to the 30-year return-period, 24-hour event in 2004) with and without climate change, as well as 341 mm per 24-hour event (equal to the 100-year event in 2004) without climate change. The maps generated for the 211 mm per six-hour event showed a minimal increase in flood depth and duration due to A1FI climate change. As expected, the 341 mm per 24-hour event showed less flood-free area compared to the smaller events, but only marginally (table 6.41). Compared to the actual 2004 storm, the 341 mm per 24-hour event in 2050 exhibited a reduction in the extent and average duration of flooding, attributed mainly to the planned increase in land elevation and planned improvements to the drainage system by 2050. Table 6.40 Comparison of 2004 Flooding Results, Narayanganj Model Maximum flooded area (km2) Average duration Flood free above 0.25 m depth Scenario (0–0.1 m) 0.1–0.25 m 0.25–0.75 m 0.75–1.5 m >1.5 m (hours to recede) 10-year event 30.64 3.54 2.11 0.65 2.29 2 30-year event 29.53 4.11 2.57 0.63 2.39 3 100-year event (actual 2004 storm) 28.62 2.86 3.89 0.96 2.91 12 Source: Zaman 2014. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 154 Assessment of Dhaka’s Modeled Zones Table 6.41 Comparison of 2050 Flooding Results, Narayanganj Model Maximum flooded area (km2) Average duration Flood free above 0.25 m depth Scenario (0–0.1 m) 0.1–0.25 m 0.25–0.75 m 0.75–1.5 m >1.5 m (hours to recede) 211 mm/6 hour (without climate change) 27.84 8.39 1.15 0.15 1.71 3 245 mm/6 hour (with A1FI climate change) 25.94 9.96 1.45 0.16 1.73 4 341 mm/24 hour (without climate change) 26.41 8.81 1.90 0.33 1.79 11 Source: Zaman 2014. Table 6.42  2050 Rainfall Events in a Changing Climate, Narayanganj Model Area Maximum flooded area (km2) Average duration Flood free above 0.25 m depth Extreme storm in 2050 (0–0.25 m) 0.25–0.75 m 0.75–1.5 m >1.5 m (hours to recede) One-dimensional model area Future SE + PI (341 mm/24 hour) 3.4 0.7 2.4 3.4 >240 Future SE + PI + CC (396 mm/24 hour) 3.3 0.7 2.4 3.4 >240 Future SE + PI + AI (341 mm/24 hour) 3.4 0.7 2.4 3.4 >240 Future SE + PI + AI + CC (396 mm/24 hour) 3.3 0.7 2.4 3.4 >240 Two-dimensional model area Future SE + PI (341 mm/24 hour) 12.2 0.1 0.1 0.1 2 Future SE + PI + CC (396 mm/24 hour) 11.2 0.2 0.2 0.9 3 Future SE + PI + AI (341 mm/24 hour) 12.3 0.1 0.1 0.1 2 Future SE + PI + AI + CC (396 mm/24 hour) 12.3 0.1 0.1 0.1 2 Source: Zaman 2014. Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; AD = adaptation to climate change. Results of the 341 mm per 24-hour rainfall event were analyzed with regard to meeting the acceptable 0.25 m flood criteria. Table 6.42 summarizes the simulation results for the extreme rainfall events. The findings showed that, for some areas, the planned drainage system was marginally unable to cope with extreme rainfall events (341 mm per 24 hours) (map 6.7). Therefore, additional drainage improvements were identified and incorporated into the model. Because some water bodies were not explicitly represented in the Digital Elevation Model (DEM) for 2050, the DEM was adjusted by lowering the land level by approximately 2 m in locations where water bodies were identified in the planned land use of the Detailed Area Plan (DAP). Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Assessment of Dhaka’s Modeled Zones 155 Map 6.7 Extreme Rainfall Event (341 mm/24 hours) in 2050 with 0.25 m Flood Criteria, Narayanganj Model Source: IWM 2014. Note: PI = planned improvements; AI = additional investments. The additional drainage improvements (i.e., refinements in the drainage model) to meet the adaptation deficit were identified as follows: • In wards 8, 12–18, and 25–26—Create local area ponds, as identified in the DAP, to store stormwater runoff during extreme events. When these additional drainage improvements were incorporated into the model, simulation of the A1FI climate change effect (396 mm per 24 hours) showed that the added measures were sufficient to meet acceptable flood criteria (map 6.8). Therefore, no measures for climate change adaptation were required for the Narayanganj area. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 156 Assessment of Dhaka’s Modeled Zones Map 6.8  Flood Depth of A1FI Rainfall Event in 2050, with Additional Drainage Improvements, Narayanganj Model Source: IWM 2014. Note: PI = planned improvements; AI = additional investments; CC = climate change. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Assessment of Dhaka’s Modeled Zones 157 References BWDB (Bangladesh Water Development Board). 2010. Feasibility Study on Drainage Improvement of Dhaka-Narayanganj-Demra (DND) Project. Dhaka: Bangladesh Water Development Board. Huq, Mainul. 2014. Cost of Adaptation to a Changing Climate. Background report pre- pared for this study, World Bank, Washington, DC. Roy, Subhendu. 2014. Urban Flooding of Greater Dhaka Area in a Changing Climate: Vulnerability, Adaptation and Potential Costs. Background report prepared for this study, World Bank, Washington, DC. Zaman, Asif M. 2014. Urban Flooding of Greater Dhaka Area in a Changing Climate: Vulnerability, Adaptation and Potential Costs. Background report prepared for this study, Dhaka, Bangladesh Climate Change Resilience Fund and Institute of Water Modelling. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Chapter 7 Assessment of Dhaka City’s Flood Resilience Introduction While the previous chapter provided the study findings by modeled zone, this chapter presents overall results for Dhaka city. First, findings from the hydrologi- cal modeling exercise are summarized for the detailed study area, h ­ ighlighting the varying effects of climate change by region. Results of the flood vulnerability assessment follow, including a ward-level analysis. Subsequent sections present total (2050) and cumulative (2014–50) estimated damage, including sector-wise estimates, using alternative-probability cases for extreme rainfall events; and the cost of adaptation and estimated benefits from these investments. Finally, the chapter considers what added expenditures might be needed to achieve a flood- free Dhaka. Summary of Modeling Findings Since the year 2000, Dhaka city has experienced three extreme rainfall events that far exceeded previous records. The underlying assumption is that such events will become more frequent in the future due to climate change processes. The simulation results (chapter 6) show that, with planned improvements, the detailed study area can cope with rainfall events of 200–250 mm per day—that is, up to a 50-year, return-period event, based on data up to 1990 (JICA 2000), with a peak intensity of about 100 mm per hour under the 2050 A1FI scenario. But for larger events exceeding 300 mm per day, which are becoming more fre- quent, additional structural measures will be required; these include installing new permanent and temporary pump stations and increasing the capacity of existing ones, laying new drainage pipes, deepening existing water bodies, and installing automatic sluice gates to prevent backflow in box-culverts. Implementing the recommended additional investments to address the cur- rent adaptation deficit would, except for rural parts of Narayanganj and DND, meet the study’s acceptable flooding criteria for the extreme event in 2050 Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7   159   160 Assessment of Dhaka City’s Flood Resilience Figure 7.1 Impact of Climate Change on Flooded Area in 2050 Narayanganj DND Eastern Dhaka Goranchatbari Kallyanpur Central Dhaka Old Dhaka 0 2 4 6 8 10 12 Percent change in flooded area (>0.25 m depth) due to A1FI climate change Source: Zaman 2014. Note: A1FI = Intergovernmental Panel on Climate Change scenario with future based on fossil fuel–intensive development; DND = Dhaka-Narayanganj-Demra. without climate change (341 mm per 24 hours). The effects of climate change (A1FI) by 2050 on the improved drainage system would have varying effects by region. With a 16 percent increase in rainfall—396 mm per day in 2050 versus 341 mm per day in the 2004 base year—the DND area would experience a 12 percent increase in flooded area, while the other six areas would remain within a 3 percent increase (figure 7.1). What Is Dhaka’s Flood Vulnerability? Results of the vulnerability assessment show that, when equal weight is assigned to the five dimensions—20 percent weight given to the physical, social, economic, and institutional CDRI dimensions used and the flood dimension (chapter 4)—the relative vulnerability among the study areas remains unchanged across all five alternative future scenarios. But increasing the weight assigned to flooding to 50 percent causes minor changes in the study areas’ relative rankings, especially for the SE + PI + CC scenario, where flooding is higher; Eastern Dhaka and Old Dhaka show some increased vulnerability, implying that flooding may have a larger impact in some pockets of these areas (table 7.1). Ward-Level Analysis Table 7.2 shows the Flood Vulnerability Index (FVI) rankings of Dhaka’s 20 most vulnerable wards/regions, which figured repeatedly under the five future Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Assessment of Dhaka City’s Flood Resilience 161 Table 7.1  FVI Rankings of Dhaka Study Areas Future scenarios Equal weightb SE + PI + SE + PI + AI + Study areaa SE + PI SE + PI + CC SE + PI + AI AI + CC CC + AD DND 1 1 1 1 1 Eastern Dhaka 2 2 2 2 2 Kallyanpur 3 3 3 3 3 Goranchatbari 4 4 4 4 4 Central Dhaka 5 5 5 5 5 Old Dhaka 6 6 6 6 6 50% flood weightb DND 1 2 1 1 1 Eastern Dhaka 2 1 2 2 2 Kallyanpur 3 3 4 4 4 Goranchatbari 4 5 3 3 3 Central Dhaka 6 6 5 5 5 Old Dhaka 5 4 6 6 6 Source: Roy 2014. Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; AD = adaptation to climate change; DND = Dhaka-Narayanganj-Demra; FVI = Flood Vulnerability Index. a. The flood vulnerability of each study area is separately examined in chapter 6, using the FVI to determine the relative vulnerability among wards and regions in each area; Narayanganj is excluded from the vulnerability analysis due to lack of detailed data. b. FVI rankings progress from 1 (highest vulnerability) to 6 (lowest vulnerability). scenarios for equal-weighted and 50 percent flood-weighted dimensions). Seven wards—02, 04, 08, 09, 15, 25, and 86—and two regions—Khilgaon (outside DNCC) and Demra (outside DSCC)—were among the top 20 under all 10 scenarios and thus appear the most vulnerable. Wards 11 and 87, which were among the top 20 in 9 out of the 10 scenarios, appear equally vulnerable. These top 20 vulnerable wards/regions were further examined to determine which of the CDRI (physical, social, economic, and institutional) and flood dimensions accounted for their added vulnerability and, within these dimensions, which particular parameters were driving the results. The findings showed that, in each ward/region, dimensions that scored below the first quartile (bottom 25 percent) for all wards/regions were the main cause of greater vulnerability. For all such dimensions, the indicators found to be “poor” were the main contributors to the dimension falling below the first quartile. All of the top 20 vulnerable wards/regions faced vulnerability from flooding extent and duration, and this vulnerability was heightened further by the other non-flood factors (table 7.3). The dimension of flood vulnerability in the overall vulnerability is based on the likely infrastructure in each area with a 100-year, return period storm, assum- ing completion of investments currently being implemented. The identified wards with high flood vulnerability can help local planners to focus on areas that need additional structural investments to minimize the higher vulnerability Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 162 Table 7.2  FVI Rankings of Dhaka Wards and Regions Future scenarios a 50% flood weighta Equal weight SE + PI+ SE + PI + SE + PI + SE + PI+ AI + Area Ward or region SE + PI SE + PI+ CC SE + PI + AI AI + CC AI + CC+ AD SE + PI SE + PI + CC SE + PI + AI AI + CC CC+ AD Kallyanpur 09 1 1 4 6 6 5 13 20 19 6 DND 87 2 >20 3 3 1 9 >20 3 4 1 Old Dhaka 82 3 2 >20 19 >20 4 >20 14 13 >20 DND 86 4 3 1 1 2 7 >20 1 1 2 Central Dhaka 25 5 4 14 16 13 11 >20 16 17 13 Kallyanpur 10 6 5 2 4 3 >20 14 6 11 3 Kallyanpur 58 7 15 >20 >20 >20 1 >20 11 10 >20 DND 88 8 >20 10 7 4 14 >20 8 7 4 DND 90 9 >20 11 9 7 15 >20 9 8 7 Eastern Dhaka Khilgaon, outside DNCC 10 14 5 13 8 3 2 2 5 8 Kallyanpur 08 11 7 12 2 5 17 12 10 3 5 Eastern Dhaka Badda, outside DNCC 12 8 8 12 9 8 1 5 6 9 Kallyanpur 11 13 6 6 8 12 >20 15 12 18 12 Goranchatbari 15 14 10 15 14 16 18 19 15 14 16 DND Demra, outside DSCC 15 12 7 5 10 10 3 4 2 10 Goranchatbari 02 16 13 17 11 17 >20 6 18 12 17 Goranchatbari 04 17 19 9 10 11 20 8 7 9 11 Kallyanpur 48 18 >20 >20 >20 >20 2 >20 >20 >20 >20 Central Dhaka 75 19 11 >20 >20 >20 16 >20 >20 >20 >20 Old Dhaka 83 20 17 >20 >20 >20 19 >20 >20 20 >20 Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; AD = adaptation to climate change; DND = Dhaka-Narayanganj-Demra; FVI = Flood Vulnerability Index. a. Wards are listed in descending order of vulnerability from 1 (highest) to 20 (lowest). Assessment of Dhaka City’s Flood Resilience 163 Table 7.3 Main Drivers of Greater Vulnerability for Dhaka’s Top 20 Vulnerable Wards/Regions Dimension Parameters: wards/regions Physical Electricity: 04, 08, 15 Water: 02, 04, 08, 15, 86, 87, 88, 90 Sanitation and solid-waste disposal: 04, 08 Road accessibility: 04, 08 Housing and land use: 04, 08, 86, 87, 88, 90 Social Population: 09, 11, 25 Health: 09, 10, 11 Education and awareness: 09, 10, 11, 25, Khilgaon Social capital: 09, 10, 11, 25 Community preparedness: Khilgaon Economic Income: 75, 82, 83, 86, 87, 88, 90 Employment: 09, 10, 11, 75, 82, 83, 86, 87, 88, 90 Household assets: 75, 82, 83, 86, 87, 88, 90 Finance and savings: 09, 10, 11, 75, 82, 83, 86, 87, 88, 90 Budget and subsidy: 09, 10, 11, 75, 82, 83, 86, 87, 88, 90 Institutional Mainstreaming of disaster risk reduction (DRR) and climate change adaptation (CCA): 09, 10, 11, Khilgaon, Badda, Demra Effectiveness of city’s crisis management framework: 09, 10, 11, Khilgaon, Badda, Demra Knowledge dissemination and management: 09, 10, 11, Khilgaon, Badda, Demra Institutional collaboration with other organizations and stakeholders during a disaster: Khilgaon, Demra Good governance: 09, 10, 11, Khilgaon, Badda, Demra Flooding Extent: 02, 04, 08, 09, 10, 11, 15, 25, 48, 58, 75, 82, 83, 86, 87, 88, 90, Khilgaon, Badda, Demra Duration: 02, 04, 08, 09, 10, 11, 15, 25, 48, 58, 75, 82, 83, 86, 87, 88, 90, Khilgaon, Badda, Demra Source: Adapted from Jahan 2014. Note: Khilgaon and Badda are regions outside the DNCC; Demra is a region outside the DSCC. caused by flooding. The variation in non-flood drivers of greater flood vulnerabil- ity across the top 20 vulnerable wards/regions can help planners to identify location-specific interventions that may be needed to minimize vulnerability. For example, in ward 9, Dhaka’s most vulnerable ward, more attention may be required to improve most parameters in the social, economic, and institutional dimensions, but not in the physical dimension (table 7.3). Summing Up Limited financial and other resource constraints make it necessary to tackle Dhaka’s river and urban flooding problems in a timely and planned manner. This requires prioritizing the city’s most vulnerable wards/regions and finding cost- effective solutions. The ward-level FVI results presented above can provide local policymakers a useful roadmap for future planning to minimize the effects of climate change. As noted, some wards have increased vulnerability owing to non- flood factors, which may exacerbate the impact caused by even low levels of flooding. In these wards, adaptation may need to focus more on reducing the non-flood drivers of vulnerability. Conversely, in more flood-prone wards, reducing the direct impact of flooding may take a higher priority. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 164 Assessment of Dhaka City’s Flood Resilience Damage Estimates Flooding in an urban area like Dhaka can lead to widespread damage resulting from submersion of buildings and property, as well as livelihood disruption among the flood-affected population. The damage is further aggravated as flooding dura- tion persists. To estimate the overall potential economic damage from urban flooding in the study area for alternative future scenarios, damage was estimated separately for the following major affected sectors of the economy: residential (buildings and property), commerce and industry, health care, roads and railway, and other surface transport (chapter 5). In each case, efforts were made to esti- mate the damage to physical capital and earnings, to the extent possible. Sector-Wise Estimates This section provides a detailed, sector-wise analysis of damage from the occur- rence of a 100-year, return-period storm in 2050 under the five alternative flood scenarios. Sector-wise cumulative damage between 2014 and 2050 is similarly examined, using random assignment of various return-period storms. Residential Buildings With the occurrence of a 100-year, return-period storm in 2050, 9.7 percent of Dhaka’s households—0.50 million out of 4.61 million households1—will face flooding and damage to their residential buildings to varying degrees if only planned improvements are made. When the added precipitation from climate change is taken into account, the number of affected households climbs to 0.61 million or 11.8 percent. Without climate change, 0.22 million households or 4.7 percent of all households are affected when additional investments to address the existing adaptation deficit are implemented along with planned improvements. However, when the extra flooding resulting from climate change is considered, the figure rises to 0.26 million households or 5.7 percent. Finally, with climate change adaptation, the number of households likely to be affected falls to 0.25 million or 5.6 percent of all households. Under a 100-year flood in 2050, the financial loss to residential buildings with planned improvements alone amounts to Tk. 1,458 million and increases by more than 70 percent (i.e., to Tk. 2,535 million) if the effect of climate change is included. Without climate change, losses total Tk. 230 million if additional investments for the current adaptation deficit are made along with planned improvements. With the added effect of climate change, losses increase some- what to Tk. 365 million, but fall to Tk. 266 million with the implementation of adaptation measures. The cumulative damage to residential buildings from 2014 to 2050 was esti- mated under varying return-period floods for each year picked at random for future years up to 2050 under the five future scenarios. The respective damage estimates amounted to Tk. 21,955 million (SE + PI), Tk. 38,177 million (SE + PI + CC), Tk. 3,460 million (SE + PI + AI), Tk. 5,492 million (SE + PI + AI + CC), and Tk. 3,998 million (SE + PI + AI + CC + AD). Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Assessment of Dhaka City’s Flood Resilience 165 The additional investments to meet the existing deficit in drainage infrastruc- ture appear to have a large beneficial effect on reducing likely damage in the housing sector. Residential Property The property damage that households face from a 100-year flood in 2050 with planned improvements only amounts to Tk. 1,672 million; this figure increases to Tk. 3,002 million with the added effect of climate change. With additional invest- ments to address the adaptation deficit, along with planned improvements without climate change, losses amount to Tk. 346 million. With the effect of climate change, they increase to Tk. 491 million, but fall to Tk. 404 million with adaptation. The cumulative property damage for households from 2014 to 2050 was estimated under varying return-period floods picked at random for future years up to 2050 under the five future scenarios. The respective damage estimates amounted to Tk. 27,879 million (SE + PI), Tk. 50,041 million (SE + PI + CC), Tk. 5,763 million (SE + PI + AI), Tk. 8,193 million (SE + PI + AI + CC), and Tk. 6,734 million (SE + PI + AI + CC + AD). Once again, the additional investments in drainage infrastructure appear to have a large beneficial effect on reducing property damage among households. Commerce and Industry With the occurrence of a 100-year flood in 2050, the loss to business from damage to assets and property totals Tk. 372 million if only planned improvements are made, and climate change causes that figure to rise to Tk. 442 million. Without climate change, the loss amounts to Tk. 224 million if additional investments are made along with planned improvements. While the added effect of climate change causes the figure to climb to Tk. 263 million, implementing ­ adaptation measures causes it to fall to Tk. 262 million. In addition, the cumulative loss to commercial establishments from 2014 to 2050 was estimated under varying return-period floods picked at random for future years up to 2050 under the five future scenarios. These respective estimates amounted to Tk. 6,150 million (SE + PI), Tk. 7,305 million (SE + PI + CC), Tk. 3,707 million (SE + PI + AI), Tk. 4,352 million (SE + PI + AI + CC), and Tk. 4,332 million (SE + PI + AI + CC + AD). The damage to assets and finished products faced by industry in 2050 from a 100-year, return-period flood with planned improvements amounts to Tk. 967 million. With the effect of climate change, the figure increases to Tk. 1, 146 million. If additional investments for the current adaptation deficit are made along with planned improvements without climate change, the damage totals Tk. 591 million. The added effect of climate change causes the figure to rise to Tk. 694 million. However, with adaptation, the figure falls to Tk. 691 million. The cumulative loss to industry from 2014 to 2050 was estimated under varying return-period floods picked at random for future years up to 2050 under the five future scenarios. These respective estimates amounted to Tk. 15,980 million (SE + PI), Tk. 18,929 million (SE + PI + CC), Tk. 9,768 million (SE + PI + AI), Tk. 11,456 million (SE + PI + AI + CC), and Tk. 11,406 million (SE + PI + AI + CC + AD). Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 166 Assessment of Dhaka City’s Flood Resilience Health Care In Dhaka, diarrhea and dengue fever are the two main diseases with increased incidence resulting from flooding, which cause increased mortality and morbidity in the affected area. The total quantifiable health damage in 2050 from a 100 year, return-period flood with planned improvements only amounts to Tk. 906 million, and this figure rises to Tk. 1,127 million with the effect of climate change. With additional investments for the existing adaptation deficit, along with planned improvements without climate change, it totals Tk. 414 million. With the added effect of climate change, it increases to Tk. 495 million. However, with adapta- tion, it falls to Tk. 491 million. The cumulative health damage from 2014 to 2050 was estimated under varying return-period floods picked at random for future years up to 2050 under the five future scenarios. These respective estimates amounted to Tk. 14,961 million (SE + PI), Tk. 18,620 million (SE + PI + CC), Tk. 6,832 million (SE + PI + AI), Tk. 8,183 million (SE + PI + AI + CC), and Tk. 8,113 million (SE + PI + AI + CC + AD). Roads and Railway The total economic loss to road infrastructure in 2050 from a 100-year, return- period flood with planned improvements amounts to Tk. 516 million, and it increases to Tk. 601 million with climate change effects. With additional investments to address the adaptation deficit in the current climate, along with planned improvements without climate change, it amounts to Tk. 414 million. The effect of climate change causes it to increase to Tk. 378 million. However, with adaptation to climate change, it falls to Tk. 372 million. The cumulative road damage from 2014 to 2050 was estimated under varying return-period floods picked at random for future years up to 2050 under the five future sce- narios. These respective estimates amounted to Tk. 8,524 million (SE + PI), Tk. 9,930 million (SE + PI + CC), Tk. 5,268 million (SE + PI + AI), Tk. 6,252 million (SE + PI + AI + CC), and Tk. 6,139 million (SE + PI + AI + CC + AD). The total economic loss to rail infrastructure in 2050 from a 100-year flood event with planned improvements alone totals Tk. 13 million, and rises to Tk. 18 million when the effect of climate change is taken into account. With addi- tional investments to address the existing adaptation deficit, along with planned improvements without climate change, the loss falls to Tk. 8 million. With the added effect of climate change, it rises to Tk. 11 million. With adaptation to climate change, however, it again decreases to Tk. 9 million. The cumulative damage to railways from 2014 to 2050 was estimated under varying return- period floods picked at random for future years up to 2050 under the five future scenarios. These respective estimates amounted to Tk. 212 million (SE + PI), Tk. 301 million (SE + PI + CC), Tk. 137 million (SE + PI + AI), Tk. 179 million (SE + PI + AI + CC), and Tk. 157 million (SE + PI + AI + CC + AD). Other Surface Transport The total economic loss to other forms of surface transport in 2050 from a 100-year flood with planned improvements amounts to Tk. 601 million, and Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Assessment of Dhaka City’s Flood Resilience 167 increases to Tk. 690 million with climate change effects. With additional invest- ments for the adaptation deficit, along with planned improvements without ­ climate change, it totals Tk. 426 million, but climate change causes it to climb to Tk. 495 million. With adaptation to climate change, however, it falls to Tk. 494 million. The cumulative damage in the transport sector from 2014 to 2050 was also estimated under varying return-period floods picked at random for future years up to 2050 under the five future scenarios. These respective esti- mates amounted to Tk. 9,934 million (SE + PI), Tk. 11,396 million (SE + PI + CC), Tk. 7,029 million (SE + PI + AI), Tk. 8,173 million (SE + PI + AI + CC), and Tk. 8,152 million (SE + PI + AI +CC+ AD). Indirect Damage from Income Loss Loss of income for Dhaka residents and migratory labor commuting to Dhaka in 2050 from a 100-year flood with planned improvements amounts to Tk. 130 million, and increases to Tk. 214 million with the effect of climate change. With additional investments to address the current adaptation deficit, along with planned improvements without climate change, it totals Tk. 49 million, but rises to Tk. 61 million with the added effect of climate change. With climate change adaptation, however, it falls to Tk. 56 million. The cumulative income loss for Dhaka residents and migratory labor from 2014 to 2050 was also esti- mated under varying return-period floods picked at random for future years up to 2050 under the five future scenarios. These respective estimates amounted to Tk. 2,069 million (SE + PI), Tk. 3,410 million (SE + PI + CC), Tk. 787 million (SE + PI + AI), Tk. 977 million (SE + PI + AI + CC), and Tk. 839 million (SE + PI + AI + CC + AD). Summary of Sector-Wise Results Table 7.4 demonstrates how the total sector-wise damage changes under the five alternative scenarios and various assumptions about the frequency of a 100-year, return-period storm. The highest damage of Tk. 9,938 million in 2050 occurs under planned improvements with climate change when a 100-year, return- period storm happens with certainty. The damage drops rapidly with additional investments (without climate change) to Tk. 2,580 million and rises somewhat when the effect of climate change is factored in. With adaptation, the effect of climate change is greatly mitigated. A similar trend continues under each of the various assumptions about the frequency of a 100-year, return period storm. Table 7.4 shows that the damage decreases as the frequency of a 100-year, return period storm goes down from once every 20 years (RP 20), to once every 33 (RP 33), 50 (RP 50), and 100 years (RP 100). Aggregate Damage for Dhaka The total combined damage for Dhaka in 2050 was estimated for the five future scenarios, using alternative assumptions about how frequently a 100-year, return- period storm would occur. Figure 7.2 shows the total monetary value of the Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 168 Assessment of Dhaka City’s Flood Resilience Table 7.4 Summary of Sector-Wise Damage in Dhaka under Alternative Scenarios Damage occurring in 2050 (Tk. million) Return period based on probability Scenario 100 year RP20 RP33 RP50 RP100 SE + PI 7,904 4,840 4,544 4,380 4,161 SE + PI + CC 9,938 6,097 5,725 5,520 5,245 SE + PI + AI 2,580 1,639 1,542 1,488 1,416 SE + PI + AI +CC 3,469 2,196 2,067 1,995 1,898 SE + PI + AI + CC + AD 2,818 1,788 1,683 1,623 1,546 Cumulative damage, 2014–50 (Tk. million) Using random numbers for each return period Scenario RP20 RP33 RP50 RP100 SE + PI 127,183 123,883 114,273 109,971 SE + PI + CC 160,069 155,936 143,868 138,479 SE + PI + AI 42,135 41,123 38,048 36,743 SE + PI + AI + CC 56,575 55,208 51,070 49,305 SE + PI + AI + CC + AD 46,004 44,898 41,541 40,113 Source: Adapted from Roy 2014. Note: Sectors covered include residential buildings, residential property, commerce, industry, health care, roads, railway, other surface transport, and income loss. SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; AD = adaptation to climate change; RP = return period (years). Figure 7.2 Total Estimated Damage for Dhaka in 2050 10,000 8,000 6,000 Million Tk. 4,000 2,000 0 100-year event RP20 RP33 RP50 RP100 Assumption a SE + PI SE + PI + AI SE + PI + AI +CC + AD SE + PI + CC SE + PI + AI +CC Source: Roy 2014. Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; AD = adaptation to climate change; RP = return period (years). a. 100-year event = actual damage from a 100-year event occurring in 2050 with certainty; RP20, 33, 50, and 100 = damage from a 100-year event occurring once every 20, 33, 50, or 100 years, respectively. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Assessment of Dhaka City’s Flood Resilience 169 damage that would result from the actual 100-year event in 2050 under the five future scenarios, followed by expected damage estimates using alternative prob- abilities of the 100-year event’s recurrence (i.e., every 20, 33, 50, or 100 years), as explained in chapter 5. Table 7.5 demonstrates that, with a 100-year, return-period storm, the total additional damage from climate change for Dhaka in 2050 would amount to Tk. 2,033 million with only planned improvements. With additional investments, that figure would fall to Tk. 888 million, reducing the extent of climate change damage by Tk. 1,145 million. Table 7.5 also provides damage estimates assuming a probability distribution of return-period storms under alternative assumptions of increased frequency of 100-year storms. Figure 7.3 shows the expected cumulative damage over the 2014–50 period for the five future scenarios, with random assignment of various return-period storms under alternative assumptions of increased frequency of the 100-year event. As described in chapter 5, alternative probabilities of the recurrence of the 100-year event are used to compute the expected value of the total cumulative damage between 2014 and 2050 if the event were to occur every 20, 33, 50, or 100 years. Table 7.6 demonstrates that, with random assignment of various return-period storms between 1-year and 100-year, return-period storms, the total additional cumulative damage (2014–50) for Dhaka from climate change would amount to Tk. 28,508 million with only planned improvements. With additional investments, that figure would fall to Tk. 12,562 million, reducing the extent of climate change damage by Tk. 15,946 million. Table 7.6 also shows cumulative damage estimates under alternative assumptions of increased frequency of 100-year storms. Overall Damage from Climate Change The study examined damage-estimate changes under a number of scenarios with varying levels of infrastructure improvements. As these improvements ­ continue to provide benefits over a long period of time, it is more useful to examine the cumulative damage resulting from climate change between 2014 and 2050 in Dhaka under the A1FI scenario since this may help in comparing the costs and Table 7.5  Damage Differences from Climate Change for Future Scenarios Using Alternative Probabilities Increased damage in 2050 (Tk. million) Return period based on probability Scenario 100 year RP20 RP33 RP50 RP100 SE + PI + CC 2,033 1,935 1,563 1,358 1,083 SE + PI + AI + CC 888 781 650 578 482 Reduced damage in 2050 due to climate change adaptation (Tk. million) SE + PI + AI + CC + AD 651 409 384 371 353 Source: Roy 2014. a. SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change. b. RP = return period (years). Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 170 Assessment of Dhaka City’s Flood Resilience Figure 7.3 Total Cumulative Damage for Dhaka, 2014–50 160,000 120,000 Million Tk. 80,000 40,000 0 RP20 RP33 RP50 RP100 Assumptiona SE + PI SE + PI + AI SE + PI + AI +CC + AD SE + PI + CC SE + PI + AI +CC Source: Roy 2014. Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; AD = adaptation to climate change. RP = return period (years). a. 100-year event = actual damage from a 100-year event occurring in 2050 with certainty; RP20, 33, 50, and 100 = damage from a 100-year event occurring once every 20, 33, 50, or 100 years, respectively. Table 7.6  Differences in Cumulative Damage (2014–50) from Climate Change for Future Scenarios Using Alternative Probabilities Increased cumulative damage (Tk. million) Using random numbers for each return period Scenario RP20 RP33 RP50 RP100 SE + PI + CC 32,886 32,054 29,594 28,508 SE + PI + AI + CC 14,441 14,086 13,021 12,562 Reduced cumulative damage due to climate change adaptation (Tk. million) SE + PI + AI + CC + AD 10,572 10,310 9,530 9,192 Source: Roy 2014. Note: SE = socioeconomic changes; PI = planned improvements; AI = additional investments; CC = climate change; RP = return period (years). benefits from these infrastructure changes in the future. For estimating the damage caused by climate change, this study used the 100-year, return-period storm as the most severe event and the scenario with the probabilistic estimate of various return-period storms occurring in the years between 2014 and 2050. With only planned improvements (PI), climate change causes a cumulative expected damage of Tk. 32,886 million. The climate change impact falls rapidly Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Assessment of Dhaka City’s Flood Resilience 171 Figure 7.4 Cumulative Climate Change Damage for Total Dhaka, Using PPP and Man-Days a. Million $US, using PPP b. Million man-days 6,000 400 300 4,000 200 2,000 100 0 0 Without climate change With climate change Without climate change With climate change With planned improvements With planned improvements With additional improvements With additional improvements Source: Roy 2014. under the additional investments (AI) scenario when the current adaptation defi- cit is addressed, with cumulative damage amounting to Tk. 14,441 million. A more meaningful comparison is gained by converting the damage estimates from local currency to U.S. dollars, using the purchasing power parity (PPP) index. Because of the uncertainty about the likely PPP index in 2050 and given that losses for the period between 2014 and 2050 are expressed in 2013 prices, the PPP adjustment is made, based on the PPP index for Bangladesh in 2005 of 2.84.2 Using PPP U.S. dollars, the additional economic loss from climate change effects under the A1FI scenario with PI amounts to US$1,197 million in 2050 and decreases to $556 million with AI undertaken to address the deficit in adap- tation to the current climate (figure 7.4a). Another useful method to estimate the extent of damage from climate change effects is to assess the number of man-days needed to recover from the economic loss. This is based on the average daily wage of urban workers in Dhaka. For this analysis, the average wage of an urban worker is assumed at Tk. 10,000 a month, for a daily income of about Tk. 400. Figure 7.4b shows that, under the A1FI scenario, an additional 82 million man-days are lost due to the effects of climate change between 2014 and 2050 with PI only. With AI, this extra loss decreases to 36 million. Cost and Benefits of Adaptation Separating the costs of meeting the current adaptation deficit and adapting to climate change gives one a better understanding of the total costs required to lessen people’s suffering in the years ahead. Table 7.7 shows that, once the Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 172 Assessment of Dhaka City’s Flood Resilience Table 7.7 Total Estimated Cost of Adaptation for Detailed Study Area Current Climate Damage Cumulative damage adaptation deficit change deficit Total cost avoided in 2050 avoided to 2050 Area modeled a (million Tk.) (million Tk.) (million Tk.) (million Tk.) (million Tk.) Old Dhaka 64.8 34.3 99.1 1,002 13,915 Central Dhaka 1,449.5 91.1 1,540.6 3,888 53,387 Kallyanpur 345.4 16.2 361.6 1,410 19,562 Goranchatbari 302.4 50.8 353.2 359 5,062 DND 544.6 1,105.4 1,650.0 352 4,868 Total cost 2,706.7 1,297.8 4,004.5 7,011 96,794 Source: Huq 2014. Note: These cost estimates are not based on detailed engineering estimates, which were beyond the scope of this study. Depending on their source of origin, equipment costs may rise sharply. DND = Dhaka-Narayanganj-Demra. a. Eastern Dhaka and Narayanganj are excluded since the structural measures recommended in the hydrological modeling exercise apply only to the five areas listed. current adaptation deficit of approximately Tk. 2.7 billion is addressed, the added cost of closing the climate change gap would require another Tk. 1.3 bil- lion, for a total estimated cost of about Tk. 4.0 billion. The damage avoided in 2050 would total more than Tk. 7.0 billion, with a cumulative (2014–50) damage savings of nearly Tk. 96.8 billion. Scope for Further Improvement In considering these cost figures, one should keep in mind that additional expen- ditures would be required to achieve a flood-free Dhaka. This study’s modeling exercise has assumed that pipes, khals, and culverts will perform according to their designated capacity and that stormwater will reach them without undue difficulty. However, it appears these assumption are not in accordance with Dhaka city’s current situation, and further studies are needed that focus on iden- tifying and removing these impediments. All khals and water bodies need to be protected from encroachment and accumulation of solid waste, and appropriate waste management is urgently needed. Because rainfall patterns appear to be changing temporally and spatially, more rain gauges are needed around Dhaka. The two main stations at Banani and Agargoan are not sufficient for collecting the required detailed data. Also, rainfall data should be collated centrally, and forecasts should be readily communicated to pump operators and drainage engi- neers as soon as a large event is expected. Furthermore, water-level data at the outfalls should be regularly recorded to know the response of drainage systems. An updated, detailed land-level survey of the Greater Dhaka region that considers the confining effects of buildings and other roadside structures on surface flows is required for more accurate, location-specific inundation modeling and mapping. Because local-level changes in land use and land levels are likely to have a greater impact on urban flooding than current estimated changes in rainfall intensities—a major finding of this study—the updated DAP and DMDP need to ensure that effective monitoring-and-enforcement measures are put in Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Assessment of Dhaka City’s Flood Resilience 173 place so that the drainage system is not adversely affected by future urban devel- opment. Recent conveyance improvements in some areas (e.g., Central Dhaka) tend to be ad hoc, meaning that flooding problems are often shifted from one location to another. In the DND area, for example, urban flooding problems persist or are worsening. With new urban development projects in Uttara (phase three) and Purbachal and private-housing projects in eastern Dhaka catchments, there is scope to introduce improved urban drainage practices. In this context, the ongoing Preparation/Updating of Stormwater Drainage Master Plan for Dhaka City can play a vital role (DWASA 2014). Not only will the master plan help to reduce the impacts of urban flooding caused by annual monsoon rainfall; it can also help to address the impacts of extreme rainfall events, which may occur with more frequency due to climate change processes. Accounting for the added expenditures needed to achieve a flood-free Dhaka would raise the total cost of adaptation substantially. Thus, a careful study incor- porating such expenditures should be undertaken to provide local policymakers a comprehensive cost estimate. The next chapter considers the non-structural flood-mitigation measures that are most urgently needed. Notes 1. This number refers only to households within DNCC and DSCC and thus excludes portions of the DND area and Narayanganj. 2. International Monetary Fund figure. References DWASA (Dhaka Water Supply and Sewerage Authority). 2014. “Preparation/Updating of Stormwater Drainage Master Plan for Dhaka City.” Prepared by Jurutera Perunding Zaaba Sdn. Bhd. (Malaysia), Farhat Consulting Engineers and Architects Limited, and SARM Associates Limited for the World Bank–funded Dhaka Water Supply and Sanitation Project, Dhaka. Huq, Mainul. 2014. Cost of Adaptation to a Changing Climate. Background report prepared for this study, World Bank, Washington, DC. Jahan, Sarwar. 2014. Assessment of Local Capacity to Deal with Flooding and Water Logging. Background report prepared for this study, World Bank, Washington, DC. JICA (Japan International Cooperation Agency). 2000. Baseline Information Study of the Dhaka Combined Flood Control cum Eastern Bypass Road Project: Final Report. Dhaka: Bangladesh Water Development Board, Ministry of Water Resources, Government of Bangladesh. Roy, Subhendu. 2014. Urban Flooding of Greater Dhaka Area in a Changing Climate: Vulnerability, Adaptation and Potential Costs. Background report prepared for this study, World Bank, Washington, DC. Zaman, Asif M. 2014. Urban Flooding of Greater Dhaka Area in a Changing Climate: Vulnerability, Adaptation and Potential Costs. Background report prepared for this study, Bangladesh Climate Change Resilience Fund and Institute of Water Modelling, Dhaka. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Chapter 8 Complementary Flood-Mitigation Measures Introduction Given Dhaka’s land scarcity, growing population density, and other prevailing socioeconomic, environmental, and geophysical conditions, the recommended conveyance-centric approach to urban flood mitigation described in the previous chapters—featuring mainly pipes, khals, and pumps—appears sound. What ­ non-conveyance and non-structural flood-mitigation measures might then be designed to complement these conveyance-centric solutions? This chapter explores the secondary measures most urgently needed and how key institu- tional, technical, and financial challenges to their implementation can be effec- tively addressed (Huq 2014). Institutional Challenges to Stormwater Management To date, multiple agencies have assumed shared responsibility for the planning, design, construction, and maintenance of Dhaka’s drainage system. These mainly include the Dhaka Water Supply and Sewerage Authority (DWASA) and Dhaka City Corporations (DCCs), as well as the Bangladesh Water Development Board (BWDB), Rajdhani Unnayan Kartripakkha (RAJUK), and Cantonment Board. Coordination among these agencies—a key ingredient for a well-functioning drainage system—is a commonly acknowledged challenge for Dhaka that has often been exacerbated by scarce resources and unplanned development. Results of the Climate Disaster Resilience Index (CDRI) analysis confirm that the city has only moderate resilience to climate disasters in terms of institutional collabo- ration and ranks even lower in terms of good governance, knowledge dissemina- tion, effectiveness of crisis management, and mainstreaming of disaster risk reduction and climate change adaptation (Jahan 2014). It has frequently been suggested that a single coordinating agency should be established with overriding authority for Dhaka’s drainage system. However, past efforts to create such a coordinating agency—including an initiative in the 1990s Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7   175   176 Complementary Flood-Mitigation Measures that involved the minister, mayor, and other high-ranking officials—fell short of expectations. Senior practitioners and officials agree that, given the city’s institu- tional track record, forming yet another coordinating agency would likely yield similar results. Instead, they suggest making a single authority responsible for stormwater management. The existing institutions best qualified to shoulder this key role would be the DCCs or DWASA. Dhaka’s two DCCs—Dhaka North City Corporation (DNCC) and Dhaka South City Corporation (DSCC)—currently have jurisdiction over tertiary/­ surface drains and roadside inlets. Recently, in response to larger volumes of rainwater runoff, they have begun to build secondary drains. Because the DCCs are responsible for surface drains, a portion of the stormwater drainage system is under their jurisdiction. However, they lack adequate experience in managing stormwater drainage systems, which differ markedly from surface systems. Also, the DCCs lack a separate drainage division or unit, and their engineers do not follow a formal design manual for solving problems. Such organizational weak- nesses and lower level of operational efficiency weaken the case for assigning the DCCs sole authority for stormwater management. Moreover, because the city is divided into two DCCs, institutional coordination would be further complicated. By comparison, DWASA appears better qualified to assume overall responsi- bility for the operation and maintenance of Dhaka’s drainage infrastructure. DWASA has the requisite capacity and experience in planning, development, and operation and maintenance of the city’s stormwater drainage system. DWASA was mandated under the Water Supply and Sewerage Authority Act of 1996 to provide stormwater drainage services to Dhaka City, having already assumed this responsibility in 1989 and having had its jurisdiction extended the following year to include Narayanganj township. DWASA has a mandated coverage area of about 360 km2, representing 12.5 million people; however, only 140 km2 is covered by stormwater drainage facilities (DWASA 2013). Given that only 39 percent of Dhaka is covered by the stormwater drainage system and that DWASA is not responsible for surface or tertiary drains, assigning DWASA full responsibility for stormwater management would require expand- ing its institutional capacity substantially. In addition, it would need to coordinate the installation and widening of drains with the DCCs, which are responsible for road development, maintenance, and improvement within the city limits. Also, a more extensive stormwater drainage system will need to be installed in the future to accommodate the city’s expansion and growing population density, emphasizing the need for strengthened institutional capacity. If current institutional arrangements remain unchanged in the foreseeable future, policy and organizational changes will be needed within the responsible agencies. For example, it will be imperative to establish separate drainage divi- sions or units within the DCCs. The DCCs’ drainage network is an integral part of the overall drainage system, and separate divisions can strengthen these corpo- rations’ capacity to handle their current responsibilities. In addition, RAJUK should include a stormwater drainage plan as a requirement for approving land Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Complementary Flood-Mitigation Measures 177 development plans. Before clearance is given, DWASA should approve the drain- age plan. If the approved drainage plan is not followed, utility services to the developed area should not be extended. Also, RAJUK should strictly follow its own area plan; that is, it should not approve any development plan in an explic- itly stated wetland area or flood-free zone. Improved Approach to Solid Waste Management A sound system for managing solid waste disposal is a critical determinant of the success of any city’s drainage system. In Dhaka, the responsibility for solid waste disposal currently lies with the DCCs. However, for both the DNCC and the DSCC, the extent of uncollected waste is quite substantial. For example, according to DSCC officials, Dhaka South generates an estimated 3,300 tons of solid waste per day, of which only two-thirds (2,200 tons) are collected; of this collected amount, 1,900 tons are processed in landfills, while the remaining 300 tons are recycled. The situation for Dhaka North is quite similar. Results of the CDRI analysis show that the physical parameter of sanitation and solid waste management has important implications for inter-zonal varia- tions in physical resilience to climate disasters. The findings show that the poorer performance of Narayanganj township and Western Dhaka (i.e., Central Dhaka, Goranchatbari, and Kallyanpur) results mainly from poor collection of waste and lack of waste treatment and recycling facilities (Jahan 2014). The clogging of stormwater drains and manholes with solid waste is a primary reason for frequent, localized flooding in Dhaka. Uncollected solid waste often finds its way into the drainage system with rainwater, restricting its passage and thus slowing the evacuation of rainwater runoff from the affected areas. In addi- tion, citizens frequently use drains or khals as though they were designated gar- bage dumps. Such practices reduce the drainage capacity of box-culverts and khals and severely worsen the drainage system’s water quality. Even worse, the DCCs lack adequate capacity to satisfactorily perform their solid waste manage- ment jobs. To adequately address the issue of solid waste management, a three-pronged approach is suggested, involving specific roles and responsibilities for the DCCs and Dhaka residents, along with a recycling strategy to reduce the need for ­ primary and secondary solid waste collection (photo 8.1). Role of Dhaka City Corporations and Local Community Achieving a more livable, less flooded city requires enhancing the DCCs’ solid waste collection and disposal capacity substantially. Of primary concern is their lack of capacity to collect the total amount of generated waste. Resolving this issue requires increasing the number of compactors in existing fleets and replac- ing outdated trucks and compactors. The DSCC requires smaller-sized compac- tors since roads and lanes in South Dhaka are narrower. A second concern is constricted passages in the stormwater drainage system from accumulated sedi- ment caused by dust and finer particles being swept into the drainage system. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 178 Complementary Flood-Mitigation Measures Photo 8.1  DWASA sign next to Segunbagicha Khal warning against encroachment, connecting wastewater line to the khal, and garbage dumping Credit: © Asif Zaman / World Bank. Used with permission. Further permission required for reuse. Resolving this problem requires raising awareness among street sweepers, who have the daily responsibility of piling and moving solid waste to the nearest ­ secondary transfer point. It is also necessary to monitor whether the street sweepers follow instructions properly and perhaps provide them incentives to do so. A third issue is resolving the mismatch between primary collection equip- ment and the secondary collection system. The current practice of dumping waste from the primary collection equipment on the ground and then transfer- ring it to the secondary collection unit is quite inefficient. Dhaka’s local community clubs and other neighborhood organizations have a critical role to play in minimizing inappropriate solid-waste disposal behavior Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Complementary Flood-Mitigation Measures 179 among local residents. Findings of the CDRI analysis show that local people’s education and awareness tend to be lower among disadvantaged groups and people living in hazard-prone zones. Massive awareness-raising campaigns involv- ing local organizations should be launched, particularly in poorer areas (e.g., South Dhaka) to help educate residents on the link between their disposal behavior and localized inundation. Local residents can also play a key role in reducing the contribution of sand at house construction sites to sedimentation of the drainage system. Most houses in Dhaka are built by either plot owners or small developers. Typically, a portion of the sand and dug-up soil stocks maintained by builders during construction finds its way into the drainage system when it rains. By monitoring construction sites, the local community can persuade contractors to properly maintain their stock to avoid its contributing to drainage system sedimentation. In addition, the DCCs should launch awareness-raising campaigns so that local residents learn appropriate ways to maintain construction sites. Minimizing Waste Collection and Transfer: Compost Plants In addition to the above measures, the DCCs should aim to reduce the amount of solid waste that must be collected, transported, and safely disposed of. To a great extent, the primary collection of solid waste has become unintentionally privatized; that is, local community based organizations (CBOs) and nongovern- mental organizations (NGOs) collect solid waste door-to-door, while the DCCs cover the cost of transport from secondary collection points to landfills. The pres- ence of readily perishable, organic waste necessitates frequent trips by the DCCs, putting even more pressure on their limited capacity. Constructing many decen- tralized plants that convert the generated organic waste from nearby areas into compost would go a long way toward reducing the amount that must be col- lected, transported, and safely disposed of. The aim would be to cover as many households, kitchen markets, and restaurants as possible. Overcoming Barriers to Composting Land scarcity is a primary obstacle to setting up localized composting plants. It appears that both the DNCC and DSCC suffer from land crisis, even for utilizing a portion of secondary transfer points, some of which are situated on the main roads. Both DCCs own numerous plots across Dhaka that they cannot utilize owing to illegal occupation by entities with support from powerful quarters. Thus, it is imperative that a process be initiated to recover illegally occupied land. Evidently, in many past instances, authorities have successfully recovered public land from illegal occupants. Additional problems include the foul odor emitted from the composting plants and their unseemly sight, likely resulting in Not in My Back Yard (NIMBY) syndrome among local residents, as well as the potential for plant leakage in flooded areas due to intense rainfall. The first two obstacles can be overcome by carefully selecting composting technologies and other innovative ways to mini- mize odor, such as the use of rock phosphate, and by constructing high walls Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 180 Complementary Flood-Mitigation Measures around the plants to hide them from public view. The third obstacle can be overcome by installing the composting system on a raised platform to avoid any leakage to the surrounding environment. The primary collection of solid waste and composting activities should be considered separate responsibilities since assigning the same entity responsibility for both could invite trouble from the current primary collector. Because primary collection from households is a profitable activity, public sentiment against estab- lishing a composting plant in their locality could be used by the aggrieved exist- ing primary collector. Cost-Benefit Analysis A composting plant with a three-ton daily capacity could process approximately 900 tons of organic waste per year. According to Waste Concern (2014), one ton of organic waste, when aerobic composted, would produce one-quarter ton of compost and reduce greenhouse gases (GHGs) by one-half ton. This means that 225 tons of compost would be produced each year, with GHGs lowered by 450 tons. Assuming a compost price of Tk. 10 per kg, the annual sales proceed from composting would total about Tk. 2.25 million. The other major portion of return on investment in the composting plant would come from the reduced cost of secondary collection, transport, and environmentally sound disposal of biodegradable waste. For example, with 900 fewer tons of waste to transport and dispose of, the DSCC would save about Tk. 1.8 million per year.1 Evidently, the positive externality of manufactur- ing compost is substantial. It would be difficult to value the reduced GHGs, given the pricing uncertainty in the international market. However, if the market price for GHG emissions were to bounce back to US$5 per ton, the annual value of certified emission reductions (CERs) would total $2,250 ($5 × 450), equivalent to Tk. 175,500 per year. When positive externalities in the form of reduced transport and disposal costs are internalized, the financial benefit from a three-ton-per-day composting plant would total Tk. 4.05 million per year, even without considering CER benefits. On the cost side, such a plant would entail annual fixed and variable ­ costs estimated at Tk. 1.16 million and Tk. 0.38 million, respectively.2 Fixed capi- tal is assumed to have a life span of five years, after which residual value would be equivalent to zero. Clearly, the extent of net benefit calls for special attention from the DCC authorities. It is in their interest to provide the composting entities land free of leasing cost and even provide them financial and logistical ­ support to set up the system. Increasing Market Demand Composting plants frequently face marketing problems since the quality of the produced compost is often at variance with acceptable standards, resulting in a relatively lower price. The presence of heavy metals and other impurities in the lower-quality compost is due to inorganic waste not having been separated from biodegradable waste in the raw materials used to produce the compost. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Complementary Flood-Mitigation Measures 181 This inefficiency can be reduced by using the source-separation method after primary collection. Extensive household awareness-raising campaigns, along with some form of payment for inorganic waste, would go a long way in instilling the practice of source separation. In some areas of the Philippines, for example, households are paid by inorganic waste collectors, who then separate out and sell the recyclables to generate revenue (Gozun and Palomata 2000). In Dhaka, the purchase of inorganic waste from households would further reduce the amount of waste the DCCs must handle. The lower price of compost in Dhaka is also due to weak demand, suggesting the need to popularize the benefits of composting. It is observed that raw com- post has less user appeal than products mixed with nutrients. Widespread public- ity should be undertaken on how composting increases the soil’s water-retention capacity and slows the process of soil erosion. With vermicomposting, the land’s water-retention capacity would be higher, reducing the amount of required ­ irrigation, which, in turn, would save farmers money. Conveyance System Maintenance Even with the best solid waste management system in place, litter, debris, and sediment cannot be entirely prevented from entering the stormwater drainage system and thus impeding the passage of rainwater runoff. Considering Dhaka’s current and probable drainage system situation over the medium term, contin- ued clogging and sediment accumulation are quite likely. Clearly, DWASA’s urgent need for modern equipment and regular cleaning of khals will continue in the foreseeable future (photo 8.2). Routine Cleaning: Sludge Removal It is argued that one meter of sludge is deposited in Dhaka’s khals and box-­ culverts every year. Close attention must be paid to sediment deposits in the drainage system in order to keep it in good condition. Cutting, removal, and such processes as dewatering and dam construction for dewatering altogether cost nearly Tk. 1,000 per m3 of sludge. The advantage of cleaning khals versus box- culverts is offset by the khals’ relative inaccessibility. An average width of 5 m is assumed for both khals and box-culverts. The cleaning of pipes is less expensive, at an estimated cost of Tk. 600 per m3. Table 8.1 presents the length of Dhaka’s pipes, box-culverts, and khals by ­system and a tentative cost estimate for cleaning them. As shown, pipes and box-culverts are concentrated mainly in Central Dhaka, while open khals, in terms of length, are found mostly in the Goranchatbari system. Based on the existing length of the infrastructure, the annual cost of sludge removal is esti- mated at Tk. 425.5 million. For the Goranchatbari system alone, the annual cleaning cost is Tk. 175.5 million. Opting for biennial sediment removal would cost DWASA about Tk. 212.8 million per year, based on the existing length of the drainage system. During the Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 182 Complementary Flood-Mitigation Measures Photo 8.2 Motorized solid-waste collector belt at Hatir Jheel outlet of Panthapath Box-Culvert near Sonargaon Hotel (Central Dhaka system) Credit: © Asif Zaman / World Bank. Used with permission. Further permission required for reuse. Table 8.1 Length of Drainage System Infrastructure and Estimated Cleaning Cost by System Drainage component Kallyanpur Old Dhaka Central Dhaka Goranchatbari Infrastructure length (km), 2004 Round pipes 31.5 24.1 112.8 10.8 Box-culverts 0 0.3 9.7 0.5 Open khals 15 3.3 1.5 33.3 Annual cleaning cost (million Tk.)a Round pipes 18.9 14.5 67.7 6.5 Box-culverts 0.0 1.5 48.5 2.5 Open khals 75.0 16.5 7.5 166.5 Total cost 93.9 32.5 123.7 175.5 Sources: IWM and DWASA. a. Per-kilometer cleaning cost is Tk. 0.6 million for round pipes, Tk. 5 million for box-culverts, and Tk. 5 million for open khals. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Complementary Flood-Mitigation Measures 183 off-year, DWASA could simply remove floating garbage.Alternatively, a ­step-by-step process could be implemented, whereby initial cleaning and sludge removal would occur annually in Old Dhaka and Central Dhaka, along with Kallyanpur, at an estimated cost of about Tk. 250 million. Considering Goranchatbari’s compara- tively low population density and limited resources, that system could be added later to the program as resource flows increase. However, a better option would be to prioritize locations for annual cleaning, leaving less problematic areas for bien- nial cleaning. In fact, the slower rate of sludge accumulation in many segments of khals and box-culverts suggests that biennial cleaning would suffice for those segments. Financial Resources for Routine Cleaning This study proposes creating a fund to cover DWASA’s routine drainage cleaning needs by adding 1 percent of the assessed base of holding tax to the DCCs’ cur- rent tax rate. This would mean adding one percentage point to the current appli- cable rate of 12 percent—covering property tax, conservancy, and lighting rates—for a new total rate of 13 percent. Despite serious weaknesses in the col- lection system, the DCCs brought in Tk. 3.83 billion from holding taxes in fiscal year July 2012–June 2013; the DNCC collected more than Tk. 2.28 billion, while the DSCC brought in Tk. 1.55 billion (DNCC 2014; DSCC 2013).3 At this rate, an extra percentage point would generate about Tk. 319 million a year. The DCCs would transfer the collected amount to DWASA’s account. It is hoped that, with a better performing collection system and continued economic growth, revenue collection will be set to rise. If so, the augmented amount should generate sufficient resources to cover DWASA’s fund. Hydraulic Jetting and Vacuum/Suction Machines Water recycling trucks fitted with a hydraulic jetting system and vacuum/suction machine can be quite useful in removing sludge from Dhaka’s clogged pipes and manholes, which contribute significantly to localized urban flooding. Unlike a freshwater system, which requires resources for production, the water recycling system utilizes recycled sewer water as jet to loosen clogs. The process can be repeated many times until the sludge tank is full, after which the accumulated sludge can be safely disposed of. For convenience sake, the hydraulic jetting sys- tems and vacuum/suction machines can be purchased separately. Both DWASA and the DCCs are in need of such equipment to meet their drainage cleaning responsibilities. In the case of DWASA, only three sets may be needed, according to officials. Each set would cost about Tk. 120 million—Tk. 80 million for the hydraulic jetting system and Tk. 40 million for the vacuum/suc- tion machine—for a total cost of Tk. 360 million. DWASA could also use the recycling trucks to earn income by offering septic-tank cleaning services to pri- vate, commercial, and public buildings.4 Cleaning septic tanks would be in DWASA’s interest since partially treated waste would otherwise infiltrate the drainage network, severely deteriorating its water quality. In the case of the DCCs, experts suggest that the DNCC and DSCC would each need five sets. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 184 Complementary Flood-Mitigation Measures Given that the DCCs are responsible for tertiary drainage systems, the required equipment could be less powerful than that required by DWASA. Each set would cost about Tk. 80 million, for a total estimated cost of Tk. 800 million. Remote-Controlled Excavators Direct access to areas of the drainage system pinpointed for cleaning is usually difficult. In addition, manually cleaning box-culverts is an often risky process requiring dewatering, which, in turn, disrupts normal economic activities. More developed economies are increasingly utilizing robotic excavators, also known as remote-controlled excavators, to address this problem. Use of excavators protects operators from exposure to the toxic gases that accumulate inside box-culverts. Also, if used during the dry season, dewatering is not required. Currently, DWASA has a project under way demonstrating how to operate remote-­ controlled excavators and their usefulness in cleaning box-culverts. The cost of remote-controlled excavators is about Tk. 40 million per unit. This study’s calculations show that DWASA would need only three such machines to cover the entire city. Thus, the required investment would total Tk. 120 million. Privatization Private-sector participation could improve Dhaka’s capacity in stormwater evacuation and solid waste management in several critical ways. First, it would introduce competition and efficiency into service delivery. Second, because of its constant need to enhance profit margin, it could rapidly adopt international best practices on efficient management and new cost-effective technologies. Finally, it would be able to substitute public-sector investment with its own resources. Clearly, the chronic shortage of investable resources suffered by such agencies as DWASA and the DCCs demonstrates the urgent need for such assistance. In the case of DWASA, the equipment needed for appropriate service delivery would require substantial financial resources. Under a suitable arrangement, the private sector could share the burden with the public agency. In order to attract their interest, private-sector participants would need to be given a certain amount of assurance that they would be able to recover and make a reasonable profit on their initial investment. A well-designed contracting system would go a long way toward achieving this aim. In some cases, activities could be unbundled into smaller components, which would offer the private sector opportunities for participating in the service delivery of DWASA and ­ the DCCs. Shaving Peak Flow A major objective of stormwater management is to shave peak flow by delaying, diverting, or detaining runoff. Among the variety of ways that may be used to achieve this objective, discussion in this section is limited to (a) the potential role of green roofs or rooftop gardens, (b) management of impervious surfaces, and (c) using the city’s water bodies as temporary detention ponds. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Complementary Flood-Mitigation Measures 185 Green Roofs Since ancient times—from the hanging gardens of Babylon to modern-day rooftop gardens—human beings have been familiar with green roofs, and Dhaka ­ is no stranger to the concept. Any casual observation of the city’s buildings reveals abundant greenery on the rooftops. The most encouraging aspect of this phenomenon is that it does not result from governmental laws or regulations; rather, citizens inherently enjoy practicing rooftop gardening. Modern-day green roofs are categorized into two groups: (a) intensive and (b) extensive. Intensive ones usually have a thicker substrate layer, which can sustain larger trees, as well as shrubs. Extensive ones have a shallower substrate, suited only for smaller plants and vegetation. Intensive green roofs require a comparatively higher level of maintenance. Many researchers have shown that replacing an impermeable rooftop with a permeable substrate substantially delays the commencement of runoff and peak flow and reduces the runoff rate and volume (Getter and Rowe 2006). In addi- tion, a green rooftop can retain a certain amount of rainwater runoff. The extent of retention is a function of the nature of the substrate, its thickness, and level of moisture in the substrate before commencement of the rainfall event. For exam- ple, Beattie and Berghage (2004) found that, for each dry substrate of 25 mm thickness, a green roof can retain up to 10 mm of precipitation. Dividing the remainder between solids (40 percent of total volume) and aerated space (20 percent of total volume) would create an ideal balance for vegetation growth and water-retention capacity. Green roofs contribute to reducing and delaying runoff and shaving peak flow. However, as precipitation volume and intensity increase, a stand-alone green-roof system is a poor substitute for a well-designed stormwater drainage system with adequate capacity. For example, MacMillan (2004) found that storm-size ranges of 10–19 mm, 20–29 mm, 30–39 mm, and 40 mm had average peak-flow reduc- tions of 85, 82, 68, and 46 percent, respectively. It is expected that increased storm size will cause peak-flow reduction to drop dramatically. Clearly, in Dhaka’s worst-case scenario (396 mm of precipitation in 24 hours), a roof garden would be of little use as a stand-alone system. However, in response to smaller- sized storm events, green roofs can be quite useful. The relative importance of green roofs in addressing Dhaka’s stormwater ­ runoff problem can be illustrated by making a rough calculation for Shankhari Bazar, one of Old Dhaka’s most densely populated areas. In 2006, 13.29 percent of Shankhari Bazar’s total land area was used either for the communications network or not at all. The remaining 86.71 percent was used for residential, com- mercial, religious, recreational, and other purposes. It was found that buildings covered an average of 68.9 percent of each plot; that is, the building-coverage ratio was 68.9 percent, meaning that 59.74 percent of its total land area was covered by buildings (under roof). Converting half of Shankhari Bazar’s roof area into green roofs would equal 29.84 percent of the locality’s area. Following Beattie and Berghage (2004), one can assume that a green roof has an average of 25 cm of substrate that can retain Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 186 Complementary Flood-Mitigation Measures up to 10 cm of precipitation. Considering that an equivalent of 29.84 percent of the locality would be under green roof, the area’s roof gardens could retain nearly 3 cm (2.98 cm) of the area’s total rainfall, provided that the substrate is dry when precipitation starts. In the case of a rainfall event of 300 mm in 24 hours,5 a green-roof system could retain about 10 percent of total rainwater volume.6 Clearly, a green-roof system can help to reduce runoff volume and shave peak flow, and can be quite effective when used in conjunction with a well-designed conveyance system. The effectiveness of roof gardens is primarily a function of the overall proportion of roof-covered land; proportion of roof covered by ­ substrate; and nature, thickness, and dryness of the media; among others. One caveat is that, during the dry season, green roofs must be watered using socially expensive DWASA resources, putting further burden on Dhaka’s water supply system. The argument for deploying a rainwater harvesting system, using structures built specifically for that purpose, is an expensive proposition for Dhaka. From an economic standpoint, rainwater harvesting—an inherently sound concept—is not the most appropriate tool for combating the city’s urban flooding problems. Management of Impervious Surfaces As discussed in previous chapters, much of Dhaka is being paved with impervi- ous surfaces (e.g., buildings, roads, and parking lots) to accommodate its growing population density. As mentioned, paved surfaces prevent rainwater from being absorbed into the ground, increasing a given area’s runoff volume. With the city’s future development, the proportion of paved surfaces is expected to continue increasing, generating more stormwater runoff. The proportion of rain that forms runoff is known as runoff coefficient. Table 8.2 shows the runoff coefficient values used by the Public Utilities Board of Singapore for designing that city’s surface drainage system. Given that Dhaka is situated on deltaic land, there is a natural tendency for its citizens to pave over the land. However, a recent trend among developers to pave whole plots on which they construct buildings must be discouraged to pre- vent the runoff situation from worsening. Building permits should clearly state that non-permeable surface is allowable only on the constructed area of the plot and that paving over vacant portions will incur tax for generating stormwater runoff. In Washington, DC, for example, such a policy has been implemented, Table 8.2 Values of Runoff Coefficient for Various Catchments Catchment characteristics when fully developed Runoff coefficient value Roads, highways, airport runways, and paved areas 1.00 Urban areas, closely and fully built-up 0.90 Residential and industrial areas, densely built-up 0.80 Residential and industrial, not densely built-up 0.65 Rural areas with fish ponds and vegetable gardens 0.45 Source: Public Utilities Board of Singapore 2011. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Complementary Flood-Mitigation Measures 187 whereby impervious surface results in a stormwater generation fee. Dhaka could consider adopting a similar system to discourage the conversion of permeable surfaces (e.g., the imposed fee could be collected along with the holding tax). Permeable pavements are frequently cited as a desirable measure. However, their porous material is prone to clogging by fine dust particles, which would necessitate occasional vacuuming. For a poor, deltaic country like Bangladesh, this costly option would work as a negative incentive and thus appears inappro- priate. Alternatively, using interlocking bricks with gaps between them would allow rainwater to permeate the ground. Some of Dhaka’s pavements already utilize interlocking bricks. It is suggested that an interlocking brick system be used for all new pavements and replacement of paved surfaces. Using Water Bodies as Detention Ponds One way to alleviate the sudden onrush of stormwater during heavy rainfall is to detain water temporarily, say for few days, to curtail peak flow. After peak flow, the water can be released in a staggered manner to avoid congestion. Fortunately, the Detailed Area Plan (DAP) has identified and proposed many water-retention areas and water bodies, such as ponds within the city (map 8.1). Most of Dhaka’s wards have some water bodies located within their boundaries that could be used as temporary detention ponds to minimize urban inundation resulting from heavy rainfall (appendix D). In the selected water bodies, water should normally be kept at a certain level so that it serves the purpose of city beautification. It should be evacuated in response to the Meteorological Department’s weather forecasts. If it is reported that an extreme storm is on the way, the water bodies should have enough room to accept at least a portion of the total incoming deluge.7 Once peak stormwater flow has passed, truck-mounted pumps can be utilized, if needed, to pump out the extra water into the nearest stormwater drain. At the same time, water in the drainage system should be pumped out until it reaches a normal level. Once the rainfall event ends and peak water flow subsides, it is critical to promptly relieve the water bodies of extra water in preparation for the next heavy storm. To illustrate, if an extreme rainfall event with precipitation of 244 mm per day were to occur (JICA 2000), 10 percent of total precipitation would be lost either to evapotranspiration or due to small local depressions, assuming a reduc- factor of 0.9. Thus, 10 percent of precipitation would not contribute to tion ­ runoff volume. The remaining 90 percent could potentially form the surface flow. If one assumes that only 10 percent of this portion permeates the ground, then 81 percent of total precipitation would reach the drainage system.8 For example, in 2004, Central Dhaka’s total area of approximately 39.2 km2 consisted of 36.8 km2 land and 2.39 km2 water bodies (i.e., 6.1 percent of the total area). For an extreme rainfall event with 244 mm of daily precipitation, the volume of stormwater flow reaching the stormwater drainage system would be calculated as (36.8 × 106) (244/1,000) × (0.9 × 0.9) = 36.8 × 244 × 103 × 0.81 = 7.27 × 106 m3. Similarly, for a rainfall event with 160 mm of precipitation per day, the amount of rainwater reaching the stormwater Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 188 Complementary Flood-Mitigation Measures Map 8.1  Water Bodies of Dhaka Source: IWM 2014. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Complementary Flood-Mitigation Measures 189 drainage system would be calculated as (36.8 × 106) (160/1,000) × (0.9 × 0.9) = 36.8 × 160 × 103 × 0.81 = 4.77 × 106 m3. Compared to the 160 mm per-day event, the 244 mm per-day event would see an additional 84 mm rise in the level of the water bodies, even without stormwater runoff from the land area. The difference in rainwater runoff volume from land area between the two events is 2.5 × 106 m3. Since the total area of the water body is 2.39 × 106 m2, the water level will rise by another 1.04 m. In the case of the 244 mm per-day event, the level of the water bodies will rise by an additional 1.13 m (1.04 m + 0.084 m), compared to the 160 mm per-day event. Thus, if 1.13 m of water were pumped out of Central Dhaka’s water bodies before the rainfall event,9 a 244 mm precipi- tation event would be downgraded to a 160 mm event in terms of flood impact. One caveat, however, is that the ponds would need to be situated at the down- ward end of the drains in order to be able to capture runoff. These findings suggest that a rigorous study should be undertaken to identify which water bodies have maximum impact on reducing water congestion in vari- ous areas of Dhaka. DWASA should acquire the identified ones that are not state-owned properties as soon as possible. The landscape design should prevent rainwater of lighter precipitation events (say, less than 30-year, return-period storms) from entering the water bodies. To deter encroachment, the water bodies should be lined with walkways. In addition, the DCCs should be encouraged to transform the walkway-lined water bodies into urgently needed, passive recre- ation zones, where city residents can enjoy open-air activities. Last Resort: Truck-Mounted Water Pumps Despite meticulous planning, Dhaka’s land-elevation pattern will remain unpre- dictable for a variety of reasons. The rainfall-generated surface flow may move in unexpected directions, and rainwater may accumulate in areas that cannot be foreseen. Thus, it would be prudent to prepare for addressing emergency cir- cumstances regarding urban flooding. One way to prepare is to build water- evacuation capacity that is quickly deployable, easy to operate, and of sufficient strength to be effective. Truck-mounted water pumps or mobile pump stations could be useful for this purpose. Such equipment is quickly deployable and can be operated without an external power supply or generator. Truck-mounted pumps can dewater at a rate above 1 m3 per second, and pumps with even higher capacities are available. Truck-mounted, mobile pump stations will be particu- larly useful in dewatering areas considered sensitive, such as key business districts, diplomatic zones, areas with critical infrastructure facilities, and densely popu- lated areas. At this stage, it is suggested that DWASA acquire a few such pumps and add more as its financial resources improve. Stormwater Quality Control Modern stormwater management systems rely not only on conveyance-centric structural methods. Increasingly, detention ponds, as well as non-structural and source-control measures, are being utilized. Control of stormwater quality, Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 190 Complementary Flood-Mitigation Measures as well as quantity, is being considered. Currently, a major objective of urban floodwater management is control of non-point source pollution. Grassed swales, wetlands, and gross pollution traps, among other measures, are frequently used to address this problem. In the case of Dhaka, storm-generated floodwater is of poor quality. Frequently black in color, it emits a foul odor and is full of decaying solid waste. The primary sources of pollution are uncollected solid waste, raw or par- tially treated sewage, partially or wholly untreated industrial effluents, and oil and grease. The subsections below discuss how pollution from these sources can be alleviated. Reducing Uncollected Solid Waste Rainwater runoff carries a variety of solids (e.g., litter, debris, and coarse sediment) into the stormwater drainage system (photo 8.3). Removal of gross ­ pollutants from stormwater runoff can be achieved using both structural and non-structural measures. Non-structural measures, discussed earlier in this chapter, include awareness-raising, source control, and better strategies for solid ­ waste management. The various structural measures that can be used include gully baskets, trash racks, and mechanical screens. Gross pollutant traps are meant to remove solids and coarse sediment from rainfall runoff. A mechanical screen installed near Sonargaon crossing could be replicated in various parts of the drainage system as required. A prerequisite for such a system is the timely evacu- ation of accumulated solids; otherwise, accumulated gross pollutants would wash back into the drainage system when the next rainfall event occurs. Photo 8.3 Hazaribagh Khal outlet (looking toward river) on the Western Flood Embankment near Rayerbazar Credit: © Subhendu Roy / World Bank. Used with permission. Further permission required for reuse. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Complementary Flood-Mitigation Measures 191 Improving Domestic Wastewater Quality Only 30 percent of Dhaka currently enjoys a formal sewerage system. In sur- rounding slums, the sanitation system is such that a substantial portion of human waste is held in situ; during flooding, it washes up, contaminating stormwater. Other households directly discharge their sewage into the drainage system or use a septic tank system. Because most septic tanks are not regularly desludged, the sewage is only partially treated before entering the drainage system. Even when sludge is removed, it is frequently dumped into the drainage system (photo 8.4). Clearly, this situation calls for regular desludging of septic tanks. Vacuum/suc- tion machines, as previously mentioned, could be quite useful in performing such a service for payment. The removed sludge could then be transported to sewer- age treatment facilities for safe disposal, in which case, many more hydraulic jetting systems and vacuum/suction machines would be needed. Slums should be targeted for a vigorous campaign to install septic systems, including the extension of subsidized loans for communities to cover upfront costs. Those residents who can afford the septic systems but continue to dis- charge their sewage directly into the drainage system should be encouraged to promptly install the systems or otherwise face steep fines. Even the DCCs are found to direct wastewater from surface drains into nearby ditches or lowlands when development of a new area is initiated. Before surface drains are connected to the appropriate drainage system, wastewater from these ditches and lowlands gets mixed with stormwater as water levels rise, resulting in contaminated, foul-smelling floodwater. Until a sewerage and Photo 8.4 Low-lying area of Kallyanpur, showing floodwater mixed with wastewater Credit: © Subhendu Roy / World Bank. Used with permission. Further permission required for reuse. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 192 Complementary Flood-Mitigation Measures drainage system is in place that covers the entire city, solutions can only be ad hoc. Even so, effective interim measures should be undertaken to minimize the problem to the extent possible. Minimizing Industrial Effluents Untreated industrial effluents discharged by thousands of variously sized process- ing units currently in operation within Dhaka’s city limits are another major source of stormwater-quality deterioration. Along with the garment industry, chemical, leather, hide, and other types of processing units continuously dis- charge their raw or partially treated effluents into the drainage system. In 2008, the Institute of Water Modelling estimated the daily discharge of industrial wastewater at 1.3 million m3—nearly three times more than the volume of domestic wastewater discharge, at 0.5 million m3 per day. A substantial portion of industrial wastewater finds its way into Dhaka’s drainage system. The Department of Environment, which is responsible for addressing industrial pol- lution in Bangladesh, must take effective measures to minimize the problem. The enormous rise in Dhaka’s land prices has greatly reduced the feasibility of implementing land-intensive measures, including grassed swales, wetlands, and construction of new detention ponds. Results of the CDRI analysis echo the fact that natural aspects of the city, particularly ecosystem services, have been deterio- rating, while implementation of environmental conversation policies are unsatis- factory (Jahan 2014). Thus, the conveyance-centric system, as recommended in previous chapters, appears to be the best solution for mitigating urban flooding. However, Eastern Dhaka and Narayanganj can opt for such land-intensive mea- sures in the future. The relevant authorities should apply lessons learned from the city’s rain-induced, flood-management predicament. Removing Oil and Grease Like many other large, motorized, and partially industrialized cities, Dhaka ­ suffers from the presence of oil and grease in its stormwater conveyance system and floodwater (photo 8.5). To minimize their presence, oil and grease traps/ separators should be installed in the stormwater drainage system. Another, perhaps cheaper option would be to install absorbents, such as polymethane foam, in sewer inlets. Absorbents are inexpensive, easy to maintain, and simple to replace. Such areas as Gulistan, Mohakhali, Syedabad, and Gabtoli—the city’s major bus-truck transportation hubs—should be targeted initially. In addition, the stormwater drainage systems of areas that feature numerous vehicle repair garages due to economic agglomeration (e.g., Tejgaon and Rasulbagh) should be fitted with oil-and-grease removal systems. Summing Up At this stage of development, the aim of controlling urban floodwater pollution can only be limited. The immediate objective should be for floodwater to appear clean and free of unpleasant odor, even though it may still contain heavy metals, Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Complementary Flood-Mitigation Measures 193 Photo 8.5  Discharge from Kallyanpur Retention Pond into Turag River Credit: © Subhendu Roy / World Bank. Used with permission. Further permission required for reuse. other toxic chemicals, and bacteria. Given Dhaka’s current socioeconomic situation, only the above-described, limited measures appear feasible. However, ­ implementing such policies would not protect downstream communities from having to face the adverse impacts of polluted, albeit clean-looking, stormwater generated upstream in Dhaka. With regard to khals, the focus should be to enhance their aesthetic value for pedestrians using walkways along them; that is, they should be maintained as passive recreation areas. Khals of Dhaka Dhaka’s extensive network of khals forms a vital component of the city’s storm- water drainage system. The khals carry rainwater to pump stations, where it is pumped out to the other side of the embankments. The subsections that follow describe the present status of Dhaka’s khals and highlight areas that require ­priority attention. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 194 Complementary Flood-Mitigation Measures Water-Flow Status At present, Dhaka has 26 khals under DWASA’s jurisdiction. Except for two khals—Debdulai and Kashaibari—these khals are maintaining their flow. Water flow in Debdulai is meager, while Kashaibari requires re-excavation and recovery of encroached land in order for flow to restart. A few other khals need widening. Appendix E provides a complete list of the 26 khals and their status. Encroachment While most khals are currently free from encroachment, several have been encroached on by land grabbers and public entities. For example, the Local Government Engineering Department (LGED) has constructed roads using a portion of Debdulai Khal, while RAJUK has encroached on Digun Khal and filled in 2 km of Baunia Khal. Segunbagicha, Rupnagar (main khal), and Kashaibari Khal face encroachment from the private sector, and portions of Kallyanpur “ka,” Journalist Colony, and Baishteki khals are on private property (photo 8.6). Steps are being taken to acquire those pieces of land. A simulation exercise was undertaken to investigate the impacts of khal encroachment in the DND area. The foreshore areas of the proposed khals were Photo 8.6 Segunbagicha Khal, Central Dhaka system Credit: © Asif Zaman / World Bank. Used with permission. Further permission required for reuse. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Complementary Flood-Mitigation Measures 195 removed in the model (representing encroachment), with just the main khal section left for water flow and storage. The results showed that, if the 2004 ­ extreme rainfall event were to occur in 2050, most of the area would be inun- dated at a depth of 0.75–1.5 m. Adding the effect of climate change would mean that nearly all of the area would be under more than 1.5 m of water (map 8.2). Need for Pedestrian Walkways The World Bank is financing a number of projects to construct walkways along Dhaka’s khals, open channels, and U-channels in the khals. Building walkways along both sides of the khals should be done as soon as possible to demarcate their proprietary boundaries. Dhaka’s walkways serve many purposes, some of which have strong positive externalities; these include the following: • The clear demarcation of khal borders makes urban residents aware of the boundaries between public and private property (some khals are situated on private property). Having clear boundaries gives neighbors more confidence in reporting encroachment on public property to the authorities and media, making it quite difficult for encroachment to occur. Map 8.2  Flood Maps for Extreme Event in 2050 with Planned Improvements and Khal Bank Encroachments Source: IWM 2014. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 196 Complementary Flood-Mitigation Measures • For encroachers, it becomes problematic to justify claims that their land continues beyond the walkways. ­ • The popularity of Dhaka’s existing walkways among pedestrians can help to reverse the city’s “pedestrian-unfriendly” reputation and, in the process, save the city many rickshaw- and passenger-hours, in turn, reducing traffic conges- tion and pollution, even if only by a small portion. • If enough space is available, benches can be added so that the walkways become passive recreational areas. Before that can occur, however, water quality must be improved dramatically. The walkways should be exclusively for pedestrian use.10 To prevent plying by rickshaws, the walkways should not exceed 1.2 m in width. Small bushes or trees planted in the middle of the walkways can deter rickshaw plying. To pre- vent plying by motorcycles, rotating gates can be installed at various points along the walkways. Road Building Building roads along khals is problematic for two major reasons and should be strongly discouraged. First, it is well-known that khals have little land availability; thus, it would be unwise to use scarce, spare land for any purpose other than drainage. Second, building any needed roads along khals using public land may not be justified on economic grounds; other locations would perhaps better serve the purpose. Closing Remarks This chapter has discussed key secondary flood-mitigation measures that can complement the primary conveyance-centric, structural solutions recommended in this book to combat Dhaka’s urban flooding problem. As discussed, better waste management and use of existing water bodies as detention ponds can sub- stantially alleviate the city’s flooding problem, while investment in green-roof systems can play an important complementary role. However, rainwater harvest- ing infrastructure—an inherently useful and cost-effective approach—would be of little use in minimizing the extent of flooding in the context of extreme intense rainfall in Dhaka. A variety of options are available for better maintenance of the conveyance system, including routine sludge removal, hydraulic jetting and vacuum/suction machines, and remote-controlled evacuators. Wetlands and other common mea- sures for controlling floodwater pollution no longer appear feasible owing to rising land prices, but future provisions should be made for implementing such solutions in Narayanganj and Eastern Dhaka. Designing and implementing an appropriate human-waste disposal system is a priority for improving the city’s floodwater quality, along with developing effective disposal systems for solid waste and industrial effluents. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Complementary Flood-Mitigation Measures 197 Notes 1. The cost of transport and landfill disposal is assumed at Tk. 2,000 per ton, based on estimates of DSCC officials. 2. These cost estimates are based on quite dated information gleaned from Enayetullah and Sinha (2000). The cost figures have been adjusted for the current price level. Even if the actual cost is twice the amount presented here, the benefit clearly outweighs the cost of establishing and running the composting plants. 3. In some cases, the assessed amount was no more than 5 percent of what should have been collected. 4. Cleaning a septic tank can cost up to Tk. 10,000. 5. According to JICA analyses (JICA 2000), a 100-year, return-period rainfall event has 244 mm per-day precipitation. 6. Given these assumptions, the 3-cm retention capacity approximates the highest limit; in reality, however, it would be substantially lower. 7. Fear of encroachment from keeping water levels too low can be minimized by pump- ing water out only after being warned that a severe rainfall event is imminent. 8. To be conservative, a low infiltration rate has been intentionally assumed. 9. Due to sloping ground, the water level of water bodies must be lowered by slightly more than 1.13 m. 10. It has been observed that walkways in Katasur are also used by rickshaws, whereas those in Gulshan are used only by pedestrians. References Beattie, D. J., and R. Berghage. 2004. “Green Roof Media Characteristics: The Basics.” In  Proceedings of 2nd North American Green Roof Conference: Greening Rooftops for Sustainable Communities, 411–16. Toronto: The Cardinal Group. DNCC (Dhaka North City Corporation). 2014. “Budget; Financial Year 2013–2014. Dhaka North City Corporation” (in Bengali). http://www.dncc.gov.bd/images/pdf​ /­Budget_-2013_-2014.pdf. DSCC (Dhaka South City Corporation). 2013. Budget, 2013–2014. Dhaka: DSCC. DWASA (Dhaka Water Supply and Sewerage Authority). 2013. Annual Report 2012–13. Dhaka: WASA Bhaban. http://www.dwasa.org.bd. Enayetullah, I., and A. H. Md. M. Sinha. 2000. “Community-Based Decentralized Composting: Experience of Waste Concern in Dhaka.” In Community-Based Solid Waste Management: The Asian Experience, edited by I. Enayetullah and A. H. Md. M. Sinha, 63–77. Dhaka: Waste Concern. Getter, K. L., and D. B. Rowe. 2006. “The Role of Extensive Green Roofs in Sustainable Development.” HortScience 41 (5): 1276–85. Gozun, B. G., and M. J. M. Palomata. 2000. “Empowering Communities for Solid Waste Management: Strategies and Practices.” In Community-Based Solid Waste Management: The Asian Experience, edited by I. Enayetullah and A. H. Md. M. Sinha, 123–35. Dhaka: Waste Concern. Huq, Mainul. 2014. Mitigation Strategies for Urban Flooding in Dhaka: Non- Conveyance Centric Approach. Background report prepared for this study, World Bank, Washington, DC. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 198 Complementary Flood-Mitigation Measures Jahan, Sarwar. 2014. Climate and Disaster Resilience of Greater Dhaka Area: A Micro-Level Analysis. Background report prepared for this study, World Bank, Washington, DC. JICA (Japan International Cooperation Agency). 2000. Baseline Information Study of the Dhaka Combined Flood Control cum Eastern Bypass Road Project: Final Report. Dhaka: Bangladesh Water Development Board, Ministry of Water Resources. MacMillan, Glenn. 2004. York University Rooftop Garden Stormwater Quantity and Quality, Performance Monitoring Report. Toronto: Toronto and Region Conservation Authority. http://www2.epa.gov/sites/production/files/documents/3.4_MacMillan​ -monitoring.pdf. Public Utilities Board of Singapore. 2011. Code of Practice on Surface Water Drainage, 6th ed. Singapore: Public Utilities Board of Singapore. Waste Concern. 2014. http://www.wasteconcern.org/model/p-4.jpg (accessed December 15). Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Chapter 9 Toward an Integrated Adaptation Strategy Introduction Urban flooding and waterlogging caused by intense rainfall are a recurring phenomenon in the Dhaka metropolis. Bangladesh’s capital city has a long history ­ of flooding as a natural hazard and is regularly inundated during the ­monsoon season, when about 70 percent of annual rainfall occurs. But recent floods have been worse in terms of inundation and duration of floodwater in the city’s fringe frequency. Climate areas, and extreme rainfall events appear to be increasing in ­ science predicts that extreme rainfall events will become more frequent over time. Thus, Dhaka may experience a number of intense rainfall events between now and 2050, meaning that the return-period of the September 2004 baseline event (341 mm per day) may be less than 100 years in the future. The Challenge of Managing Urban Flooding Dhaka has experienced three extreme rainfall events since the year 2000, highlighting the growing challenge of managing urban flooding in the future. ­ Since 2004, urban flooding conditions have improved in parts of Central Dhaka due to key investments. DWASA has worked hard to lay new drainage pipes, reclaim encroached khals, and improve conveyance in pipes and khals through desilting. Despite these efforts, problems persist or are worsening in other parts of the city (e.g., DND area). The identified causes are combined sewer and stormwater drainage, disposal of solid waste in the drainage system, illegal encroachment of khals, filling of lowlands, and lack of coordination among key stakeholder institutions. As a result, city dwellers, family assets, the urban-built environment, and vital public infrastructure are more vulnerable to the risks of urban flooding and waterlogging. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7   199   200 Toward an Integrated Adaptation Strategy Climate Change Impact The effects of climate change in the South Asia region, including greater rainfall intensity and increased frequency of extreme weather events, are expected to aggravate Dhaka’s flood vulnerability. The Intergovernmental Panel on Climate Change (IPCC) predicts that South Asia will experience an increase in both aver- age and extreme rainfall during summer monsoon seasons of longer duration. By 2050, the average monsoon flows in the Brahmaputra (Jamuna) River at Chilmari could increase by up to 14 percent. In a changing climate, it is expected that extreme rainfall events in Greater Dhaka could increase by 16 percent by 2050. In the rivers around Dhaka, monsoon-season water levels may increase by about 30 cm. Even with increased upstream flow, internal rainfall increases, and increased river levels, the total flooded area of Greater Dhaka is expected to decrease. The main reason is the predicted increase in land levels resulting from urbanization, as identified in RAJUK’s Detailed Area Plan (DAP), highlighting how local-level planned improvements can offset the effects of climate change on urban flooding and thus the urgency of implementing them. With planned improvements—all improvements to Dhaka’s drainage infrastructure, including those already planned and proposed by the relevant ­ drainage-system authorities—the detailed study area can cope with medium- ­ sized rainfall events (200–250 mm per day), with a peak intensity of approxi- mately 100 mm per hour. However, coping with larger events—those exceeding 300 mm per day—will require additional investments. In this study, the addi- tional investments scenario is limited to the investment needed to meet the acceptable flooding criteria before the effect of climate change. With climate change, flooding once again increases over those limits, requiring adaptation investment to lower the flood level to meet the selected criteria. The findings of this study show that the extent of additional investments required to address the existing adaptation deficits, as well as the added investment needed to address the effect of climate change, varies among the modeled zones. Furthermore, the 100-year event that occurred in September 2004 (341 mm per day) will likely become more frequent in the future as a result of climate change processes. Analyses of rainfall data for Dhaka have shown that the return period of extreme rainfall events is lessening. Uncertainty analyses conducted by IWM on the magnitude and timing of climate change risk highlight the need to make required investments in structural measures that adapt the drainage system to more frequent extreme rainfall events. Given the likelihood of more frequent extreme rainfall events due to climate change, it is pragmatic for investment plans to include all adaptation measures to meet climate change. Recommendations This study recommends a prudent strategy that begins by addressing the current adaptation deficit that Dhaka’s residents already face in coping with rainfall extremes. Closing the current adaptation gap is an attractive, no-regrets option Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Toward an Integrated Adaptation Strategy 201 that can provide policymakers a solid foundation on which to prioritize and sequence additional measures to reduce potential future damage from climate risk. The design of this study, documented in Part I of this book (chapters 2–5), provides Dhaka’s local decision-makers such a planning approach. General General recommendations from the hydrological modeling study include pro- tecting all remaining khals and water bodies from encroachment, installing more rain gauges around Dhaka, and readily communicating rainfall forecasts to pump operators. An updated, detailed land survey of the Greater Dhaka region should be conducted. In light of changing rainfall frequencies and intensities, existing design practices should be reviewed and incorporated into the design manual being developed by DWASA for the Drainage Master Plan Study. Due to the potential for heavy-rainfall events in the pre-monsoon season, maintenance schedules should be reviewed. In addition, box-culverts should no longer be constructed in urban areas, drainage outlets to rivers should be regularly de-silted and protected from encroachment, and the foreshore of surrounding rivers should be protected to prevent river-water levels from rising significantly during the monsoon season. Furthermore, effective monitoring and enforcement of RAJUK’s DAP is required, and appropriate solid waste management is urgently needed. Finally, better coordination is required among all institutional stakeholders. Detailed Study Area For each of the seven modeled zones of the detailed study area—Old Dhaka, Central Dhaka, Kallyanpur, Goranchatbari, Eastern Dhaka, DND, and Narayanganj (chapter 3)—the hydrological modeling exercise recommends specific structural measures to address the current adaptation deficit (chapter 6). These additional investments include installing new permanent and temporary pump stations, increasing the capacities of existing pump stations, laying new drainage pipes, deepening existing water bodies, and installing automatic sluice gates to prevent backflow in box-culverts. For Old Dhaka, the study recommends increasing temporary pump capacities, removing sediment from Debdulai and Dholaikhal box-culverts, operating all three pumps at Dholaikhal Pump Station during extreme events, and protecting khals and drains from solid waste. For Central Dhaka, two new pump stations are recommended—one 12.5 m3/s station at the downstream end of Panthapath Box-Culvert and a 15 m3/s station and sluice gate near the Bashabo water pump. In addition, khals should be protected from encroachment and solid waste. The Kallyanpur drainage system requires deepening of the water body in ward 9, installing new pipe drains to drain low pockets to khals and temporary pump station locations, increasing temporary pump capacities, maintaining a 100-ha retention pond at the Kallyanpur Pump Station, and protecting khals and water bodies from encroachment and solid waste. For the Goranchatbari system, the Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 202 Toward an Integrated Adaptation Strategy study recommends connecting new drains in developing areas to the khal system, deepening ponds at Mirpur Zoo and Botanical Garden, and protecting khals from encroachment and solid waste. In the DND area, pump capacities need to be increased at Shimrail, Adamjinar, and Pagla, retention ponds identified in the BWDB feasibility study need to be incorporated into the updated DAP, and water bodies and khals need to be protected. In Narayanganj, all water bodies identified in the DAP need protection, and khals and drains should be protected from solid waste. In Eastern Dhaka, planned land use, as identified in the DAP, must be enforced, water retention ponds require protection, and khals should be protected from encroachment and solid waste. By implementing the additional investments recommended in chapter 6, all study areas, with the exception of rural parts of Narayanganj and the DND area, would meet the acceptable flooding criteria set forth in this study for the extreme rainfall event in 2050 without climate change (i.e., 341 mm per 24 hours). The effects of A1FI climate change by 2050 on the improved drainage system would have varying effects on the detailed study area. With a projected 16 percent increase in rainfall (i.e., 396 mm per day in 2050), the DND area would experience a 12 percent increase in flooded area, while the other six areas would remain within a 3 percent increase. Cost of Closing the Adaptation Gap The cost of meeting Dhaka’s current adaptation deficit without climate change would total approximately Tk. 2.7 billion, while the added cost of closing the climate change gap would require another Tk. 1.3 billion, for a total estimated cost of about Tk. 4.0 billion. Central Dhaka would require the largest investment to meet its current adaptation deficit, at about Tk 1.4 billion, while Old Dhaka would require the least, at just under Tk. 65 million (chapter 7). The study findings show that the cost of the damage caused by not taking these recommended measures would be quite significant for the modeled zones and Dhaka city overall (chapters 6 and 7). For example, if a single 100-year, return-period storm were to occur in 2050, the increased damage caused by climate change would amount to more than Tk. 2.0 billion under a scenario with planned improvements only, but would be reduced to just Tk. 09 billion if addi- tional investments had also been made. Such savings in damage of Tk. 1.1 billion in just one year reveal how quickly the investment of Tk. 2.7 billion in current adaptation deficit can be paid back. The results showed similar outcomes for increased climate change damage under a weighted probability of various return- period storms for both planned-improvement and additional-investment sce- narios. Similar cumulative effects of climate change (2014–50) were found with the random assignment of various return-period storms under alternative assumptions of increased frequency of a 100-year, return-period event. Thus, the cumulative damage caused by climate change between 2014 and 2050 would go down from Tk. 28.5 billion with only planned improvements to Tk. 12.6 billion, representing a savings of Tk. 15.9 billion—a much larger amount than the invest- ment of Tk. 2.7 billion in current adaptation deficit. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Toward an Integrated Adaptation Strategy 203 Study Assumptions This study’s hydrological modeling exercise for 2050 is based on a number of optimistic assumptions. For example, the study assumes that Dhaka’s land cover will change according to RAJUK’s DAP. It also assumes that all planned and proposed improvements to the city’s drainage infrastructure by the relevant drainage-system authorities will be implemented. In Bangladesh, however, there is skepticism about appropriate enforcement of these plans. As noted in chapter 7, it is assumed that drainage pipes, khals, and box-culverts will perform ­ according to their designated capacity and that rainwater will reach them with- out undue difficulty. But given that these assumptions somewhat contradict Dhaka’s current reality, a careful study accounting for the required added expen- ditures would need to be conducted, including comprehensive cost estimates. Complementary Adaptation Measures In addition to the conveyance-centric solutions recommended in chapters 6 and 7, achieving a flood-free Dhaka will require an array of non-conveyance and non- structural adaptation measures. As discussed in chapter 8, implementation of these “soft” adaptation measures will require overcoming key institutional, tech- nical, and financial barriers. Better solid waste management, including compost- ing, and the use of existing water bodies as detention ponds can substantially alleviate Dhaka’s flooding problem. Most of the city’s wards have some water bodies located within their boundaries that could be used as temporary deten- tion ponds to minimize urban inundation resulting from heavy rainfall. For example, it is estimated that pumping out 1.13 m of water from Central Dhaka’s water bodies before a rainfall event would result in downgrading a 244 mm precipitation event to a 160 mm event in terms of flood impact. Such findings ­ suggest the need for undertaking a rigorous study to identify which water bodies have maximum impact on water congestion in various areas of the city. To deter encroachment, the water bodies should be lined with pedestrian walkways. In addition, Dhaka City Corporations (DCCs) should be encouraged to transform walkway-lined water bodies into urgently needed, passive recreation zones, where city residents can enjoy open-air activities. Rainwater harvesting infrastructure—an inherently useful and cost-effective approach—would be of little use in minimizing the extent of flooding in the context of Dhaka. A variety of options are available for better maintenance of the conveyance system, including routine sludge removal, hydraulic jetting and vacuum/suction machines, and remote-controlled evacuators. A fund could be ­ established to cover DWASA’s routine drainage cleaning needs by adding percent of the assessed base of holding tax to the DCCs’ current tax rate. 1 ­ Owing to rising land prices, wetlands, grass swales, and other common mea- sures for controlling floodwater pollution no longer appear feasible in Dhaka; however, future provisions should be made for implementing such solutions in Eastern Dhaka and Narayanganj township. Designing and implementing an Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 204 Toward an Integrated Adaptation Strategy appropriate human-waste disposal system is a priority for improving the city’s floodwater quality, along with developing effective disposal systems for solid waste and industrial effluents. Toward an Integrated Strategy In addition to this book’s recommended primary and secondary flood-mitigation measures, results of the flood vulnerability assessment can provide local policy- makers a powerful tool for identifying wards and regions outside the DCCs that may require greater attention, while designing adaptation methods to cope with flooding resulting from extreme weather events (chapter 6). As discussed in chapter 7, some wards and regions have increased vulnerability owing to non- flood factors that may exacerbate the impact caused by even low levels of flooding. In such areas, adaptation may need to focus more on physical, social, ­ economic, and institutional factors. In other areas more prone to flooding, it may be important to focus on flood mitigation and reducing the direct impact of flooding, in addition to addressing the physical, social, economic, and institu- tional aspects. Equipped with a menu of options designed to close the current adaptation deficit and further climate-proof urban infrastructure, local decision-makers can better evaluate their choices and develop more effective strategies that prioritize and sequence activities as resources permit. The benefits of addressing the current adaptation deficit are significant, even without climate change, making ­ such investments an attractive no-regrets option. The “hard” infrastructure investments recommended in this book cannot yield the expected benefits if implemented in isolation. Primary structural solutions must be complemented by urgently needed, “soft” adaptation measures designed to overcome key insti- tutional, technical, and financial barriers within a framework of sound develop- ment policies. In this way, Dhaka will have a solid foundation on which to build and sustain local resilience to the risk of urban flooding now and in the future. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Appendixes A. Individuals Consulted in Dhaka to Estimate Adaptation Costs  207 B. Basin and Region Modeling and Rainfall Analyses  209 C. Dhaka City Rainfall Analyses: Previous Study Findings  219 D. DAP-Proposed Water Retention Areas and Water Bodies  221 E. Status of 26 Khals under DWASA Jurisdiction  223 Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7   205   Appendix A Individuals Consulted in Dhaka to Estimate Adaptation Costs This appendix lists key practitioners who were consulted in Dhaka to estimate the costs of implementing the structural measures recommended in this study. Dhaka Water Supply and Sewerage Authority Ching, Kyaw Sha. Superintending Engineer and Project Director, Removal of Water Logging in Dhaka City (Phase II), Drainage (R&D) Circle. Habib Ullah, Kazi. Executive Engineer, Dhaka Water Supply and Sanitation Project (World Bank–assisted). Rahman, Md. Shafiqur. Executive Engineer. Rahaman, Mostafizur. Sub-Divisional Engineer, Drainage (R&D) Division-1. Vries, Taco de. Resident Project Manager, Urban Dredging Demonstration Project, Drainage (O&M) Circle. Uddin, A. K. M. Shahid. Superintending Engineer. Institute of Water Modelling Bhuiyan, Md. Anwar Hossain. Senior Water Resource Planning and Design Specialist, Water Resources Planning Division. Khan, Md. Arzel Hossain. Senior Water Resources Management Specialist, Water Resources Planning Division. Bangladesh Water Development Board Alam, Choudhury Nazmul. Executive Engineer, Design Circle-3. Islam, K.M. Nurul. Superintending Engineer, Design Circle-3. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7   207   208 Individuals Consulted in Dhaka to Estimate Adaptation Costs Rahman, Md. Habibur. Ex. Addl. Director General, Deputy Team Leader, CEIP project. Other Practitioners Ali, Md. Liakath. Climate Change and Environment Adviser, Department for International Development, Bangladesh. Islam, Shariful. Assistant Manager (Sales), Milnars Pumps Ltd. Khondoker, Md. Ekramul Hoque. Assistant Engineer (Mechanical), Waste Management Division, Dhaka North City Corporation. Rahman, Md. Shafiqur. Chief Executive Officer, Megatech Engineers, Uttara. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 appendix B Basin and Region Modeling and Rainfall Analyses The study area is under the influence of the water level of the Brahmaputra, as mentioned in chapter 2. Therefore, the river flooding situation depends on the water level of the Brahmaputra (Jamuna) River and its distributaries in the north-central region of Bangladesh during the monsoon months. But urban ­ ­ flooding is influenced by the amount of precipitation over the city area, as well as the prevailing water level in the surrounding rivers that act as outfall channels. Change in Monsoon River Flows: Basin Model Output As discussed in chapter 2, Institute of Water Modeling (IWM)’s GBM basin model was used to estimate the changes in flows of the Brahmaputra (Jamuna) River due to future climate change. The results showed significant increases in monthly river flows, especially during July–August, for the A1FI scenario in 2050 (figure B.1). The change in average monsoon flows—the average change in flow from June to September—was 5.8 percent for the A1FI scenario in 2025. By 2050, the average monsoon flow for this scenario increased by 13.6 percent, compared to ­ the 2004 base period (figure B.2). The monsoon flow changes of the Brahmaputra (Jamuna) River were used in the upstream boundary point of IWM’s North Central Region Hydrodynamic (NCRHD) model to simulate the monsoon season flows and water levels of the river network in the extended study area. Change in Sea Levels Based on recent assessments by the Intergovernmental Panel on Climate Change (IPCC), this study considered the effect of sea-level rise as negligible for the future scenarios modeled for the year 2050. Citing Vermeer and Rahmstorf Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7   209   210 Basin and Region Modeling and Rainfall Analyses Figure B.1  Average Monthly Flow of Brahmaputra (Jamuna) River at Chilmari 70,000 65,000 60,000 Average discharge (m3/s) 55,000 50,000 45,000 40,000 35,000 June July August September A1FI_2050 B1_2050 Base scenario Source: Zaman 2014. Note: A1FI = Intergovernmental Panel on Climate Change (IPCC) scenario with future based on fossil fuel– intensive development; B1 = IPPC scenario with future based on a high level of environmental and social consciousness combined with a globally coherent approach to a more sustainable development. Figure B.2 Change in Monsoon Discharge of Brahmaputra (Jamuna) River at Chilmari 15 12 9 Percentage 6 3 0 A1FI_2050 B1_2050 Scenario Source: Zaman 2014. Note: A1FI = Intergovernmental Panel on Climate Change (IPCC) scenario with future based on fossil fuel– intensive development; B1 = IPPC scenario with future based on a high level of environmental and social consciousness combined with a globally coherent approach to a more sustainable development. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Basin and Region Modeling and Rainfall Analyses 211 (2009), the IPCC (2012) indicates that sea-level rise would amount to 1.90 m for the A1FI scenario by 2100. Further, the IPCC (2013), citing Pfeffer, Harper, and O’Neel (2008), finds that sea-level rise of more than 2 m by 2100 is physi- cally implausible. An estimate of 0.8 m by 2100 that includes increased ice dynamics is considered the most plausible. Change in Land Use and Land Levels Figure B.3 shows the frequency distribution of land levels for Greater Dhaka. In the 2004 base scenario, the main land-level category is 3–4 m Public Works Department (PWD). However, due to population growth, subsequent land-filling, and density increases, the proportion of higher-elevation land levels increases in 2025 (i.e., the Figure B.3  Frequency Distribution of Land Levels for Greater Dhaka 0–1 1–2 2–3 3–4 4–5 5–6 6–7 Elevation (m PWD) 7–8 8–9 9–10 10–11 11–12 12–13 13–14 14–15 15–16 0 10 20 30 Percent of area 2004 2025 2050 Source: Zaman 2014. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 212 Basin and Region Modeling and Rainfall Analyses ­ ategory is 7–8 m distribution shifts to the right). By 2050, the main land-level c PWD. Such changes can already be seen in land development works undertaken for Uttara Phase 3 developments in northwestern Dhaka and also land-level changes in Badda Thana, located in the eastern part of the city. Given the further planned urbanization in the DAP’s proposed land use and expected population growth, this dramatic change in land use could be seen by 2025. Change in River Water Level: Regional Model Output The NCRHD model used the basin-level output, along with climate and river network data to simulate river flows and water levels of key rivers in north- central Bangladesh. The simulated water levels, in turn, were used to generate indicative flood maps for the DMDP area. The 2004 base-run simulation of hydrological and hydrodynamic conditions showed that most parts of central Dhaka, Savar core area, Gazipur Sadar core area, and Purbachal were nearly flood free. Other areas were flooded to various depths. The total flood free area (<0.25 m depth) was 22 percent, while deep- flooded areas (>1.5 m depth) comprised 64 percent. Based on the 2050 DEM, flood-depth maps were generated for the 2050 A1FI climate change scenario. The flood-free area was about 39 percent, while deep-flooded area was about 60 percent. The increase in flood-free area was due to local people’s initiative to raise the plinth level of their houses above anticipated inundation levels. Compared to 2004 monsoon flooding, the total flooded area (>0.25 m depth) decreased from 78 percent to 61 percent in the 2050 A1FI scenario. This occurred despite increased upstream flow from Chilmari, internal rainfall increases of 9.9 percent, and increased water levels (about 30 cm) of rivers around Dhaka. The main reason for the decrease in flooded areas is the projected increase in land levels in proposed built-up areas, as identified in the DAP. Changes in peak water levels of rivers around Dhaka City were simulated for the worst-case (A1FI) scenario in 2050. The simulated increase was less than 30 cm, suggesting that climate change is unlikely to have a major impact on peak water levels compared to the 2004 level of flooding (figure B.4). Change in Rainfall Heavy rainfall is one of the main factors that cause urban flooding. Therefore, changes in rainfall intensities need to be studied in order to model urban flooding and its impacts. As mentioned in chapter 3, rainfall data from the Bangladesh Water Development Board (BWDB) and Bangladesh Meteorological Department (BMD) were analyzed. Analyses of BWDB Data The study analyzed daily rainfall data from several BWDB stations around Dhaka. The Banani Station covered the longest period (1957–2011). The time-series record showed that two recent extreme events have broken previous records: Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Basin and Region Modeling and Rainfall Analyses 213 Figure B.4 Changes in Peak River Water Levels around Dhaka City Due to Climate Change 7.6 7.4 7.2 Meters (PWD) 7.0 6.8 6.6 6.4 Pubail Mirpur Mill Barrack Demra Station Base scenario 2050 A1FI Source: Zaman 2014. Note: The gauge points (i.e., locations of water-level gauges) for these Bangladesh Water Development Board–maintained stations are as follows: Balu River (Pubail), Turag River (Mirpur), Buriganga River (Mill Barrack), and Lakhya River (Demra). PWD = Public Works Department. September 13, 2004 (341 mm per day) and July 28, 2009 (448 mm per day). The previous record was set on May 12, 1965 (222 mm per day). These findings high- light a trend of increasing variability in daily rainfall (figure B.5). Frequency analyses were conducted on the Dhaka (Banani) station data and compared to JICA (2000) findings. According to the JICA analyses, the September 2004 event of 341 mm per day would be classified as more than a 1,000-year, return-period event. However, the IWFM (2005) classified the event as a 100-year return-period event. The inclusion of more recent data—including the extreme event of July 2009—further reduced the return period to 50 years. These findings clearly show that return periods are decreasing; that is, larger events are occurring more frequently (figure B.6). Analyses of BMD Data The study also analyzed BMD Dhaka station data, including more detailed, sub-daily rainfall data, which are required for urban flood modeling. An interest- ­ ing observation is that, compared to the BWDB data, the 2004 event is recorded as larger than the 2009 event (figure B.7a). The reason is that the two agencies collate daily rainfall data at different times of the day. BMD’s three-hourly rain- fall data reveals that the 2009 event was much more intense compared to other rainfall events since 2003 (figure B.7b). Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 214 Basin and Region Modeling and Rainfall Analyses Figure B.5  Daily Rainfall Time Series at BWDB Dhaka Station 500 400 Millimeters per day 300 200 100 0 4/ 61 4/ 63 4/ 71 4/ 81 4/ 91 4/ 93 4/ 99 4/ 01 4/ 03 11 4/ 59 4/ 65 4/ 67 4/ 69 4/ 73 4/ 79 4/ 83 4/ 85 4/ 89 4/ 95 4/ 97 4/ 09 4/ 57 4/ 75 4/ 77 4/ 87 4/ 05 4/ 07 19 19 19 19 19 19 19 20 20 20 19 19 19 19 19 19 19 19 19 19 19 20 19 19 19 19 20 20 4/ Source: Zaman 2014. Note: BWDB = Bangladesh Water Development Board. Figure B.6 Change in Return Periods of Daily Rainfall Events at BWDB Dhaka Station 600 500 One-day rainfall (mm/day) 400 300 200 100 0 2 5 10 25 50 100 500 1,000 Return period (years) Up to 1999 (JICA 2000) Up to 2009 Source: Zaman 2014. Note: BWDB = Bangladesh Water Development Board. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Basin and Region Modeling and Rainfall Analyses 215 Figure B.7 Rainfall Data at BMD Dhaka Station, 2003–09 a. Daily 350 300 250 Millimeters per day 200 150 100 50 0 2003 2004 2005 2006 2007 2008 2009 b. Three-hourly 250 200 Millimeters per three hours 150 100 50 0 2003 2004 2005 2006 2007 2008 2009 Source: Zaman 2014. Note: BMD = Bangladesh Meteorological Department. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 216 Basin and Region Modeling and Rainfall Analyses These findings reveal how the extreme rainfall events of 2004 and 2009 ­ iffered. The 2004 event was a high-volume event of relatively low intensity. d Over a four-day period, about 550 mm rainfall was recorded, with 341 mm maximum over a 24-hour period. This event is of particular interest for this study as it caused extensive flooding in Dhaka at the time and corresponds well with the availability of other datasets. Therefore, 2004 was chosen as the base year for the modeling simulations. The changes in three-hourly rainfall due to climate change in 2050 were then analyzed for the high-emissions (A1FI) scenario. Figure B.8 shows that the intensities increase to 16 percent by 2050. Design Storms This study adopted the widely accepted JICA (2000) estimates for one-day rainfall amounts for different return-period events. For the purpose of urban drainage modeling, these one-day rainfall amounts were distributed into an hourly synthetic storm pattern (i.e., design storm). JICA (1987) was followed Figure B.8 Three-Hourly Rainfall for September 2004 Event at BMD Dhaka Station 100 Overall rainfall: 550 mm Maximum 24-hours: 341mm 09/13/2004 80 Millimeters every three hours 60 40 20 0 0 0 6:00 9:00 0 0 0 0 0 0:00 3:00 6:00 9:00 0 3:00 0 0 0 0:00 3:00 6:00 0 9:00 0:00 6:00 9:00 0 0 3:00 0 0 0:00 12:0 21:0 12:0 15:0 18:0 21:0 12:0 15:0 18:0 21:0 15:0 18:0 12:0 15:0 18:0 21:0 09/1 2004 04 04 09/1 2004 09/1 2004 09/1 2004 09/1 2004 09/1 2004 09/1 2004 09/1 2004 09/1 2004 04 04 09/1 2004 09/1 2004 04 09/1 2004 04 09/1 2004 09/1 2004 04 04 09/1 2004 09/1 2004 09/1 2004 04 09/1 2004 09/1 2004 04 09/1 2004 04 09/1 2004 2/20 2/20 4/20 4/20 3/20 2/20 3/20 3/20 4/20 1/20 1/20 2/ 2/ 2/ 4/ 4/ 4/ 4/ 2/ 2/ 3/ 3/ 3/ 3/ 3/ 4/ 1/ 1/ 1/ 1/ 1/ 1/ 09/1 09/1 09/1 09/1 09/1 09/1 09/1 09/1 09/1 09/1 09/1 Time period Observed 2025 B1 2025 A1FI 2050 B1 2050 A1FI Total rainfall (mm) 550 582 589 611 638 Change (%) 106 107 111 116 Observed 2050 A1FI 2050 B1 Source: Zaman 2014. Note: A1FI = Intergovernmental Panel on Climate Change (IPCC) scenario with future based on fossil fuel–intensive development; B1 = IPPC scenario with future based on a high level of environmental and social consciousness combined with a globally coherent approach to a more sustainable development; BMD = Bangladesh Meteorological Department. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Basin and Region Modeling and Rainfall Analyses 217 Figure B.9 Hourly Fractions of Six-Hour Design Storm Event 50 40 Percent of total rainfall 30 20 10 0 1 2 3 4 5 6 Hours from start of rainfall Source: Zaman 2014. for the first day’s rainfall pattern of the two-day, five-year design storm. Figure B.9 shows that the pattern consists of a peak intensity three hours after the rain starts, when 44 percent of the total rainfall falls within an hour. Based on the above pattern, theoretical storms were developed for 10-, 20-, and 30-year return periods with and without climate change in 2025 and 2050. The drainage models were used to simulate the impacts of these design storms with and without climate change. Summary Remarks The study area of Greater Dhaka lies in a delta region, where the flooding pat- tern is dominated by spills from the Brahmaputra River, as well as intense local rainfall. Simulations of climate change effects with the Brahmaputra Basin model provided estimates of average monsoon-flow increases of up to about 14 percent by 2050 at Chilmari (northern Bangladesh) for the A1FI scenario. The knock-on effect on regional rivers around Dhaka, combined with increased monsoon-­ season rainfall, was simulated using IWM’s NCRHD model. In the worst-case scenario, an increase of approximately 30 cm was found in the peripheral rivers. As a result, in some areas of Greater Dhaka, the depth of flooding increased. However, in such areas as Eastern Dhaka, where urbaniza- land-raising, flood depths decreased or even became flood free.1 tion is driving ­ This finding highlights the importance of local effects on urban flooding, which sometimes can exceed global/regional climate change effects. The finding also highlights that land-raising is not driven primarily by climate change risks, but it does reduce such risks as a co-benefit. This complex interaction has been investigated further for the detailed study area. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 218 Basin and Region Modeling and Rainfall Analyses Note 1. In Eastern Dhaka, which is vulnerable to flooding from the Balu River, land is being raised before houses and office buildings are constructed; it is possible that the entire area will be raised by land filling before the planned embankment along the Balu River is completed. References IPCC (Intergovernmental Panel on Climate Change). 2012. Managing the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation. A Special Report of Working Groups I and II of the Intergovernmental Panel on Climate Change, edited by C. B. Field, V. Barros, T. F. Stocker, D. Qin, D. J. Dokken, K. L. Ebi, M. D. Mastrandrea, K. J. Mach, G.-K. Plattner, S. K. Allen, M. Tignor, and P. M.  Midgley. Cambridge, U.K.: Cambridge University Press. ———. 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, edited by T. F. Stocker, G.-K. Plattner, M. Tignor, S. K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex, and P. M. Midgley. Cambridge, U.K.: Cambridge University Press. IWFM (Institute of Water and Flood Management). 2005. Investigation of Hydrological Aspects of 2004 Flood, with Special Emphasis on Dhaka City. Dhaka: Institute of Water and Flood Management, Bangladesh University of Engineering and Technology. JICA (Japan International Cooperation Agency). 1987. Study on Storm Water Drainage System Improvement Project in Dhaka City. Tokyo: JICA. ———. 2000. Baseline Information Study of the Dhaka Combined Flood Control cum Eastern Bypass Road Project: Final Report. Dhaka: Bangladesh Water Development Board, Ministry of Water Resources, Government of Bangladesh. Pfeffer, W. T., J. T. Harper, and S. O’Neel. 2008. “Kinematic Constraints on Glacier Contributions to 21st-Century Sea-Level Rise.” Science 321: 1340–43. Vermeer, M., and S. Rahmstorf. 2009. “Global Sea Level Linked to Global Temperature.” Proceedings of the National Academy of Sciences 106 (51): 21527–32. Zaman, Asif M. 2014. Urban Flooding of Greater Dhaka Area in a Changing Climate: Vulnerability, Adaptation and Potential Costs. Background report prepared for this study, Bangladesh Climate Change Resilience Fund and Institute of Water Modelling, Dhaka. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Appendix C Dhaka City Rainfall Analyses: Previous Study Findings This study reviewed previous research that investigated trends and changes in rainfall frequencies in and around Dhaka city. The Climate Change Cell (CCC) (2009) analyzed rainfall trends at 8 BMD and 11 BWDB stations around Bangladesh, including Dhaka, for 1960–2001. The findings showed that mean monsoon rainfall in Dhaka decreased marginally for the 1960–80 and 1981–2001 periods. However, for the latter period, the standard deviation doubled. The decade trend for monsoon rainfall (June–October) was decreasing, at −27 mm per decade; however, it was increasing for the other seasons: 13 mm per decade in winter (November–February), 28 mm per decade in summer (March–May), and 16 mm per decade in the critical period (March 11–May 10). The decade trend for September alone was also positive, at 11 mm per decade. Trends in 10-day rainfalls were analyzed, and findings similar to the monthly analyses were reported. The research also found that the trend in annual maximum rainfall for Dhaka was –10 mm per decade. The study did not examine changes in return period for rainfall events. Murshed, Islam, and Khan (2011) also investigated the impact of climate change on rainfall intensity in Bangladesh. For the 1953–2009 period, it was found that the annual maximum rainfall at BMD’s Dhaka station was decreasing at a rate of 0.0154 mm per year. However, over a more recent period (1979–2009), the trend was increasing at 2.7 mm a year. The main factors iden- tified for the increase were the two quite large events that occurred during the most recent 10-year period. Omitting these two events results in a declining trend of 1.06 mm per year. The study also compared observed mean daily ­ rainfall data with similar output from PRECIS, a regional climate model. The observed data showed an increasing trend of 0.0103 mm per year for the 1953–2009 period, and the model showed a trend of 0.014 mm per year. The model was then used to predict the trend in monsoon rainfall up to the year 2100 for the high-emissions scenario (A1FI) of the Intergovernmental Panel on Climate Change (IPCC). The simulated results showed an increasing trend of Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7   219   220 Dhaka City Rainfall Analyses: Previous Study Findings 0.121 mm per year for monsoon rainfall. But the study did not investigate pos- sible changes in return period for different-size rainfall events. Ahammed and Hewa (2012) analyzed rainfall data from BMD’s Dhaka station for the 1953–2009 period. Various frequency analyses of daily rainfall ­ data were conducted. For rainfall events with return periods of 3, 20, 45, and 100 years, the respective daily rainfall amounts were 200 mm, 325 mm, 375 mm, and 425 mm. The study also developed intensity-duration-frequency diagrams based on BMD rainfall data. Hossain et al. (2013) studied Dhaka’s rainfall trends in the context of the city’s expansion over the past few decades. The analysis covered 20 years of rain- fall data (1985–2005) from 254 BMD and BWDB stations. The findings showed that urban areas tended to receive less rainfall, in terms of frequency and inten- sity, than did rural areas. They further showed that areas downwind of urban centers tended to receive more rainfall than upwind areas. References Ahammed, F., and G. A. Hewa. 2012. “Development of Hydrological Tools Using Extreme Rainfall Events for Dhaka, Bangladesh.” Water International 37 (1): 43–52. CCC (Climate Change Cell). 2009. Characterizing Long-Term Changes of Bangladesh Climate in Context of Agriculture and Irrigation. Dhaka: Climate Change Cell, Department of Environment, Ministry of Environment and Forestry and Component 4B, Comprehensive Disaster Management Programme, Ministry of Food and Disaster Management. Hossain, F., A. M. Degu, A. T. Woldemichael, W. Yigzaw, C. Mitra, J. M. Shepherd, and A. H. M. Siddique-E-Akbor. 2013. “Water Resources Vulnerability in the Context of Rapid Urbanization of Dhaka City (a South Asian Megacity).” In Climate Vulnerability: Understanding and Addressing Threats to Essential Resources, edited by R. Pielke, Vol. 5. Amsterdam: Elsevier. Murshed, S. B., A. K. M. S. Islam, and M. S. A. Khan. 2011. “Impact of Climate Change on Rainfall Intensity in Bangladesh.” Paper presented at 3rd International Conference on Water and Flood Management organized by Institute of Water and Flood Management, Bangladesh University of Engineering and Technology, Dhaka. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Appendix D DAP-Proposed Water Retention Areas and Water Bodies Table D.1  DAP-Proposed Water Retention Areas and Water Bodies Proposed land use Proposed land use and Proposed land use and area (km2) area (km2) and area (km2) Water Water Water Ward no. retention Water Ward no. retention Water Ward no. retention Water or thana area body or thana area body or thana area body 90 0.00 0.02 20 0.00 0.16 40 0.00 0.14 89 0.00 0.01 19 0.00 0.89 39 0.00 0.02 88 0.00 0.01 18 0.00 0.07 38 0.00 0.02 87 0.00 0.09 17 0.67 0.03 37 0.00 0.59 86 0.00 0.09 16 0.00 0.18 36 0.00 0.03 85 0.03 0.00 15 0.00 0.56 35 0.00 0.02 84 0.04 0.01 14 0.00 0.03 34 0.00 0.01 83 0.00 0.01 13 0.00 0.04 33 0.00 0.00 82 0.00 0.02 12 0.00 0.05 32 0.00 0.00 81 0.00 0.02 11 0.00 0.11 31 0.00 0.01 80 0.00 0.02 10 0.31 0.14 30 0.00 0.00 79 0.00 0.01 09 0.58 0.17 29 0.06 0.02 77 0.00 0.00 08 0.10 0.32 28 0.04 0.07 76 0.00 0.01 07 0.00 0.03 27 0.00 0.11 75 0.00 0.01 06 0.50 0.17 26 0.09 0.20 73 0.00 0.01 05 0.00 0.09 15, Narayanganj 0.00 0.02 71 0.00 0.00 04 0.00 0.19 14, Narayanganj 0.00 0.03 70 0.00 0.00 03 0.00 0.02 13, Narayanganj 0.00 0.04 68 0.00 0.00 02 0.00 0.20 12, Narayanganj 0.00 0.03 66 0.00 0.01 01 0.00 0.05 11, Narayanganj 0.01 0.03 65 0.00 0.01 Uttara 1.85 1.13 10, Narayanganj 0.02 0.03 63 0.00 0.00 Savar 0.01 0.00 09, Narayanganj 0.33 0.02 62 0.00 0.02 Sabujbagh 0.14 0.07 08, Narayanganj 0.16 0.00 60 0.00 0.01 Pallabi 1.75 0.15 07, Narayanganj 0.15 0.01 table continues next page Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7   221   222 DAP-Proposed Water Retention Areas and Water Bodies Table D.1  DAP-Proposed Water Retention Areas and Water Bodies (continued) Proposed land use Proposed land use and Proposed land use and area (km2) area (km2) and area (km2) Water Water Water Ward no. retention Water Ward no. retention Water Ward no. retention Water or thana area body or thana area body or thana area body 59 0.00 0.00 18, Narayanganj 0.00 0.12 06, Narayanganj 0.03 0.00 58 0.00 0.04 17, Narayanganj 0.00 0.02 05, Narayanganj 0.01 0.00 57 0.00 0.02 16, Narayanganj 0.00 0.02 04, Narayanganj 0.08 0.00 56 0.00 0.03 51 0.00 0.00 03, Narayanganj 0.06 0.01 55 0.00 0.07 50 0.00 0.00 02, Narayanganj 0.16 0.00 54 0.00 0.14 49 0.00 0.18 01, Narayanganj 0.17 0.00 53 0.00 0.01 48 0.00 0.04 Narayanganj 1.24 1.88 52 0.00 0.03 47 0.00 0.09 Khilgaon 3.88 0.02 25 0.00 0.05 46 0.00 0.09 Demra 1.03 0.44 24 0.00 0.09 45 0.00 0.01 Cantonment 0.00 0.06 23 0.00 0.15 43 0.01 0.09 Badda 6.79 0.11 22 0.02 0.17 42 0.00 0.01 Airport 0.00 0.14 21 0.00 0.09 41 0.00 0.04 Total 20.32 10.95 Sources: DAP and IWM. Note: DAP = Detailed Area Plan. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 appendix E Status of 26 Khals under DWASA Jurisdiction Table E.1 Status of 26 Khals under DWASA Jurisdiction Khal size Serial no. Khal name Length (m) Width (m) Status (updated to May 31, 2013) 1 Katasur 1,722 12 No encroachment; uninterrupted flow 2 Ramchandrapur 2,940 18–30 No encroachment; uninterrupted flow 3 Khilgaon-Bashabo 2,540 9–18 No encroachment; uninterrupted flow 4 Begunbari 10,940 20–60 No encroachment; uninterrupted flow 5 Shahjadpur 1,910 9 Uninterrupted flow 6 Sutibhola 3,680 15–20 Uninterrupted flow 7 Abdullahpur 3,160 18–36 Uninterrupted flow 8 Segun Bagicha 3,887 9–24 Encroachment in a few places; uninterrupted flow 9 Mohakhali 2,320 9–18 No encroachment; uninterrupted flow 10 Ibrahimpur 1,260 9–18 No encroachment; uninterrupted flow 11 Digun 4,588 22–60 RAJUK encroached on 700 m; uninterrupted flow using alternate route 12 Baunia 8,828 18–30 RAJUK filled in 2 km and re-excavation needed; but continuous flow 13 Kallyanpur A 1,540 10 740 m on private property; being acquired 14 Kallyanpur B 2,408 8–12 No encroachment; uninterrupted flow 15 Kallyanpur D 1,564 9–12 No encroachment; uninterrupted flow 16 Kallyanpur E 288 9 No encroachment; uninterrupted flow 17 Kallyanpur F 982 9–24 300 m under PWD; being transferred to DWASA; uninterrupted flow 18 Kallyanpur Main 3,460 18–36 No encroachment; uninterrupted flow 19 Debdulai 2,000 9–15 Encroachment; LGED built road using a portion; little flow. 20 Rupnagar Main 2,430 18 Eastern Housing problem; some encroachment by others; inadequate width Rupnagar Branch-1 (Zoo) 897 3 No encroachment; uninterrupted flow table continues next page Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7   223   224 Status of 26 Khals under DWASA Jurisdiction Table E.1  Status of 26 Khals under DWASA Jurisdiction (continued) Khal size Serial no. Khal name Length (m) Width (m) Status (updated to May 31, 2013) Rupnagar Branch-2 (Arambagh) 1,160 12 No encroachment; uninterrupted flow Rupnagar Branch-3 (Duaripara) 823 18 No encroachment; uninterrupted flow 21 Jirani 5,500 9–24 Uninterrupted flow 22 Manda 5,000 9–15 Uninterrupted flow; proposal made to acquire for widening 23 Hazaribagh (including Kalunagar 3,000 10–18 Encroachment in Kalunagar section; section) uninterrupted flow 24 Kashaibari 8,000 10–12 Encroachment; needs to be recovered and re-excavated to facilitate flow 25 Journalist Colony 1,157 9 669 m on private property; proposal sent to acquire land; uninterrupted flow 26 Baishteki 1,280 12 853 m on private property; needs to be acquired; uninterrupted flow Source: DWASA. Note: DWASA = Dhaka Water Supply and Sewerage Authority; LGED = Local Government Engineering Department; PWD = Public Works Department; RAJUK = Rajdhani Unnayan Kartripakkha (Capital Development Authority of Bangladesh). Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Environmental Benefits Statement The World Bank Group is committed to reducing its environmental footprint. In support of this commitment, the Publishing and Knowledge Division lever- ages electronic publishing options and print-on-demand technology, which is located in regional hubs worldwide. Together, these initiatives enable print runs to be lowered and shipping distances decreased, resulting in reduced paper consumption, chemical use, greenhouse gas emissions, and waste. ­ The Publishing and Knowledge Division follows the recommended standards for paper use set by the Green Press Initiative. Whenever possible, books are printed on 50 percent to 100 percent postconsumer recycled paper, and at least 50 percent of the fiber in our book paper is either unbleached or bleached using Totally Chlorine Free (TCF), Processed Chlorine Free (PCF), or Enhanced Elemental Chlorine Free (EECF) processes. More information about the Bank’s environmental philosophy can be found at http://crinfo.worldbank.org/wbcrinfo/node/4. Urban Flooding of Greater Dhaka in a Changing Climate  •  http://dx.doi.org/10.1596/978-1-4648-0710-7 Intense rainfall floods Dhaka, Bangladesh, one of the world’s fastest-growing megacities, year in and year out, and those in the city’s slums and shanties are usually the most affected. Low-lying flood plains, rivers, and canals that once drained water are gradually filling up as a result of indiscriminate urbanization and now magnify rather than help solve the problem. The climate outlook for South Asia in the 21st century is heavier and more erratic rainfall during the monsoon season, according to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change and the World Meteorological Organization. Thus, climate change may further aggravate Dhaka’s flood vulnerability. As such, Dhaka needs to better manage its flood drainage infrastructure and strengthen the city’s climate- disaster resilience and adaptive capacity. But climate-smart policies require local planners to better understand the likely damage from current flooding, potential damage from climate-related risks, and measures that can be taken to cope with current and future flooding and adaptation costs. Urban Flooding of Greater Dhaka in a Changing Climate: Building Local Resilience to Disaster Risk assesses Dhaka’s preparedness for urban flood emergencies, estimates probable damage from extreme rainfall events by 2050 with or without climate change, develops structural adaptation measures to cope with current and future flooding, evaluates the reduction in economic damage resulting from implementing these measures, and estimates their cost. The analysis in this book will help Bangladesh’s policy makers take targeted steps to mitigate urban flooding in Dhaka and improve the city’s resilience in the face of climate change and variability. Equipped with a host of investment options designed to address current flooding and further climate-proof urban infrastructure, local decision makers will be able to develop realistic, yet effective, strategies that prioritize interventions and sequence activities. “ The topic of this book is right on target and quite timely, particularly in the context of the frequent flooding that Dhaka has been facing. The authors’ data-intensive multidisciplinary analysis of the megacity’s urban flooding problem with and without climate change, deep insights into the flood-damage risks it already faces, and practical policy recommendations that local decision makers can act on are most welcome. This book will make a real difference.” —Professor Jamilur Reza Choudhury, Vice Chancellor, University of Asia Pacific “ This book meticulously synthesizes hydrological modeling and socioeconomic analysis—a rare practice—to offer a deeper understanding of how to tackle the multifaceted problem of urban flooding. It will be useful for practitioners and academics alike, not only in Dhaka, but also in other megacities of the region now facing similar challenges.” —Professor M. Monowar Hossain, Executive Director, Institute of Water Modelling “ This is the most comprehensive technical analysis of Dhaka’s urban flooding problem yet. The authors’ down-to-earth recommendations need to be taken seriously. As someone who handles the megacity’s flooding and waterlogging issues at the highest level of responsibility, I feel this book is a must-read for government officials and practitioners working to achieve our dream of a flood-free Dhaka.” —Taqsem A. Khan, Managing Director, Dhaka Water Supply and Sewerage Authority (DWASA) ISBN 978-1-4648-0710-7 SKU 210710