The World Bank Flood Risk Assessment for the Ganges Basin in South Asia For the attention of: Dr. Bill Young Lead Water Resources Specialist South Asia Water Initiative The World Bank, The Hindustan Times House (Press Block) 18-20, Kasturba Gandhi Marg New Delhi - 110001 Email: wyoung@worldbank.org Company Information: Name RMSI Private Limited CIN U74899DL1992PTC047149 Registered Office Address Seating 3, Unit No. 119, First Floor, Vardhman Star Citi Mall, Sector-7, Dwarka New Delhi Delhi-110075 India Corporate Office Address A-8, Sector-16 Noida, 201 301 India Tel:+91 120 251 1102, 251 2101 Fax:+91 120 251 1109, 251 0963 E-mail: info@rmsi.com Contact: RMSI Private Limited A-8, Sector 16 Noida 201301, India Tel: +91-120-251-1102, 2101 Fax: +91-120-251-1109, 0963 www.rmsi.com Email - info@rmsi.com Final Report: Volume I Confidential Page 2 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Acknowledgements The Final Report (Volume-I and Volume –II) is an output of the South Asia Water Initiative (SAWI); a partnership between the World Bank and the governments of United Kingdom, Australia and Norway. The program is designed to support countries improve and deepen transboundary dialog, enhance the basin and water resources knowledge base, strengthen water institutions, and support investments that lead to sustainable, fair and inclusive development of the Himalayan Rivers. RMSI is thankful to Dr. Bill Young, Lead Water Resources Specialist, World Bank, Washington DC for the continuous support provided throughout the project. We are thankful for the support and suggestions provided by Dr. Satya Priya, Senior Water Resources Specialist, World Bank, New Delhi. Important contributions were made by Dr. Anju Gaur, Senior Water Resources Specialist World Bank, New Delhi. The core team immensely benefitted from the advice of World Bank consultants Dr Raj Chabungbam, Dr Debashish Goswami and Ms Priyanka Chaturvedi. The work was led by Dr. MVRL Murthy, Vice President, RMSI Private Limited. The core team included Pratul Srivastava, Ankit Avasthi, Deshraj Singh, Sanwar Bajiya, Divyanshi Agarwal, and Sumit Dhand of RMSI Private Limited. We would also like to extend our sincere gratitude to various institutions like Central Water Commission, India; Dartmouth Flood Observatory; Department of Hydrology and Meteorology, Nepal; India Meteorological Department; and Water Resource Information System of India (WRIS) whose various public source datasets, publications, and other information have been extensively used and referred in the report at various stages of the project. The team gratefully acknowledges support from the World Bank and SAWI and sincerely thanks all individuals and institutions met during the development of this work. Final Report: Volume I Confidential Page 3 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Table of Contents Acknowledgements ............................................................................................................... 3 Table of Contents.................................................................................................................. 4 List of Figures ....................................................................................................................... 8 List of Tables ...................................................................................................................... 16 Abbreviations Used ............................................................................................................. 21 Executive Summary ............................................................................................................ 22 1 Introduction .................................................................................................................. 29 1.1 Background........................................................................................................... 29 1.2 Topography........................................................................................................... 32 1.3 Climate ................................................................................................................. 32 1.4 Meteorological causes of heavy rainfall over Ganges River Basin ........................ 32 1.5 Rainfall pattern...................................................................................................... 33 1.6 Objectives of the study.......................................................................................... 33 1.7 Scope of the study ................................................................................................ 34 1.8 About the fourth deliverable .................................................................................. 34 2 Exposure Data Collection and Management ................................................................ 36 2.1 Methodology for exposure data development........................................................ 36 2.1.1 Data Collection .............................................................................................. 36 2.1.2 Data Processing And Data Gaps ................................................................... 36 2.1.3 Data Quality Checks And Quality Assurances................................................ 37 2.1.4 Development Of Exposure Data And Generation Of Gis Data Layers ............ 38 2.2 Analysis of exposure elements ............................................................................. 38 2.2.1 Demographic Analysis ................................................................................... 38 2.2.2 Housing ......................................................................................................... 43 2.2.2.1 Overall Housing Occupancy.................................................................... 43 2.2.2.2 Residential houses ................................................................................. 46 2.2.2.3 Commercial houses ................................................................................ 49 2.2.2.4 Industrial Establishments ........................................................................ 52 2.2.2.5 Educational Institutions ........................................................................... 55 2.2.2.6 Health Facilities ...................................................................................... 58 2.2.2.7 Other houses .......................................................................................... 60 2.2.2.8 Housing Structural Types ........................................................................ 63 2.2.3 Infrastructure Data ......................................................................................... 64 2.2.3.1 Road Network ......................................................................................... 65 2.2.3.2 Rail Network ........................................................................................... 67 Final Report: Volume I Confidential Page 4 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 2.2.4 Agricultural Data ............................................................................................ 69 2.3 Estimation of exposure values .............................................................................. 77 2.3.1 Estimation Of Built-Up Area ........................................................................... 77 2.3.2 Calculation Of Total Housing Exposure Values .............................................. 80 2.3.3 Transportation................................................................................................ 82 2.3.3.1 Roads ..................................................................................................... 82 2.3.3.2 Railways ................................................................................................. 84 2.3.4 Agriculture ..................................................................................................... 86 3 Flood Hazard Assessment ........................................................................................... 93 3.1 Methodology overview .......................................................................................... 93 3.2 Data availability ..................................................................................................... 93 3.2.1 Meteorological Data ....................................................................................... 93 3.2.2 Hydrological Data .......................................................................................... 95 3.2.3 Topographical Information ............................................................................. 97 3.2.4 Land Use Land Cover (LULC) ........................................................................ 98 3.2.5 Soil Map......................................................................................................... 99 3.2.6 Flood History................................................................................................ 100 3.3 Basin and River Delineation ................................................................................ 106 3.4 Hydraulic modeling ............................................................................................. 107 3.4.1 Model Set Up ............................................................................................... 107 3.4.2 Model calibration and validation ................................................................... 109 3.4.2.1 Historical Flood Extent .......................................................................... 115 3.4.3 Estimation of return period flows .................................................................. 118 3.4.3.1 Annual peak discharge data:................................................................. 118 3.4.3.2 Return period flow estimation: ............................................................... 118 3.4.4 Flood hazard mapping for return period flows .............................................. 120 4 Risk Assessment ....................................................................................................... 124 4.1 Vulnerability functions ......................................................................................... 126 4.1.1 Methodology ................................................................................................ 126 4.1.2 Review of Vulnerability (Depth – Damage) Functions .................................. 127 4.1.2.1 Asian Institute of Technology Depth-Damage Functions ....................... 127 4.1.2.2 U.S. Army Corps of Engineers (USACE) Depth-Damage Functions ..... 128 4.1.3 Final vulnerability functions .......................................................................... 129 4.2 Risk analysis - loss module ................................................................................. 132 4.2.1 Methodology ................................................................................................ 132 4.2.2 Aggregated Exposure .................................................................................. 134 4.2.3 Site specific exposure .................................................................................. 135 Final Report: Volume I Confidential Page 5 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 4.2.4 Historical and modeled Flood damage ......................................................... 136 4.2.4.1 Introduction ........................................................................................... 136 4.2.4.2 Factors Influencing Flood Losses.......................................................... 136 4.2.4.3 Validation of flood losses ...................................................................... 138 5 Findings – Basin Level ............................................................................................... 147 5.1 Demography ....................................................................................................... 147 5.2 Economic Losses ................................................................................................ 150 5.2.1 Buildings ...................................................................................................... 152 5.2.1.1 Residential ............................................................................................ 152 5.2.1.2 Commercial .......................................................................................... 155 5.2.1.3 Industrial ............................................................................................... 157 5.2.2 Educational institutions ................................................................................ 159 5.2.3 Health facilities............................................................................................. 161 5.2.4 Other buildings............................................................................................. 164 5.2.5 Infrastructure................................................................................................ 166 5.2.5.1 Road Network ....................................................................................... 166 5.2.5.2 Rail Network ......................................................................................... 168 5.2.6 Agriculture ................................................................................................... 171 5.2.6.1 Rice ...................................................................................................... 171 5.2.6.2 Wheat ................................................................................................... 173 5.2.6.3 Maize .................................................................................................... 176 6 Findings – Sub-basin Level ........................................................................................ 179 6.1.1 Lower Ganges Sub-basin ............................................................................ 179 6.1.2 Bagmati Sub-basin ...................................................................................... 183 6.1.3 Kosi Sub-basin............................................................................................. 187 6.1.4 Gandak Sub-basin ....................................................................................... 191 6.1.5 Mahananda Sub-basin ................................................................................. 195 6.1.6 Middle Ganges Sub-basin ............................................................................ 199 6.1.7 Ghagra Sub-basin ........................................................................................ 203 6.1.8 Kamla-Balan Sub-basin ............................................................................... 207 6.1.9 Yamuna Sub-basin ...................................................................................... 211 6.1.10 Ramganga Sub-basin .................................................................................. 215 6.1.11 Gomti Sub-basin .......................................................................................... 219 6.1.12 Sone Sub-basin ........................................................................................... 223 6.1.13 Upper Ganges Sub-basin ............................................................................ 227 6.1.14 Tons Sub-basin............................................................................................ 231 6.1.15 Sind Sub-basin ............................................................................................ 235 Final Report: Volume I Confidential Page 6 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 6.1.16 Ken Sub-basin ............................................................................................. 239 6.1.17 Betwa Sub-basin .......................................................................................... 243 6.1.18 Chambal Sub-basin ..................................................................................... 247 7 Conclusions and Recommendations .......................................................................... 251 8 Appendix A: Hydrological Station Network of Ganges Basin ...................................... 257 References ....................................................................................................................... 261 Final Report: Volume I Confidential Page 7 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia List of Figures Figure 1-1: Ganges Basin in India, Nepal, Bangladesh and China ...................................... 31 Figure 2-1: State level population distribution for the basin in India ..................................... 39 Figure 2-2: Gender Distribution for the basin in India .......................................................... 39 Figure 2-3: State level gender distribution for the basin in India. ......................................... 39 Figure 2-4: Literacy rate for the basin in India ..................................................................... 40 Figure 2-5: State wise literacy distribution for the basin in India .......................................... 40 Figure 2-6: Province level population distribution for the basin in Nepal .............................. 40 Figure 2-7: Overall gender ratio for the basin in Nepal ........................................................ 41 Figure 2-8: Province level gender distribution for the basin in Nepal ................................... 41 Figure 2-9: District level distribution of total population for the basin in Bangladesh ............ 41 Figure 2-10: Overall gender ratio for the basin in Bangladesh ............................................. 42 Figure 2-11: District level distribution of gender for the basin in Bangladesh ....................... 42 Figure 2-12: Population distribution in Ganges Basin .......................................................... 43 Figure 2-13: Housing distribution in India ............................................................................ 44 Figure 2-14: Occupancy of houses in India ......................................................................... 44 Figure 2-15: Province wise distribution of households for the basin in Nepal ...................... 44 Figure 2-16: Housing occupancy for the basin in Nepal ...................................................... 44 Figure 2-17: District level distribution of households for the basin in Bangladesh ................ 45 Figure 2-18: Housing distribution for the basin in Bangladesh ............................................. 45 Figure 2-19: Distribution of total houses in the Ganges Basin ............................................. 45 Figure 2-20: State level distribution of residential houses for the basin in India ................... 46 Figure 2-21: Types of residential houses in India ................................................................ 47 Figure 2-22: Province level distribution of residential houses for the basin in Nepal ............ 48 Figure 2-23: District level distribution of residential houses for the basin in Bangladesh ..... 48 Figure 2-24: Distribution of residential houses in Ganges Basin .......................................... 49 Figure 2-25: State level distribution of commercial houses for the basin in India ................. 50 Figure 2-26: Different types of commercial houses in India ................................................. 50 Figure 2-27: Province level distribution of commercial houses for the basin in Nepal .......... 51 Figure 2-28: District level distribution of commercial houses for the basin in Bangladesh ... 51 Figure 2-29: Distribution of commercial houses for the Ganges Basin ................................ 52 Figure 2-30: State level distribution of industrial establishments for the basin in India ......... 53 Figure 2-31: Examples of industrial establishments in India ................................................ 53 Figure 2-32: Province level distribution of industrial establishments for the basin in Nepal .. 54 Figure 2-33: District level distribution of industrial establishments for the basin in Bangladesh ................................................................................................................................ 54 Final Report: Volume I Confidential Page 8 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Figure 2-34: Distribution of industrial establishments for Ganges Basin .............................. 55 Figure 2-35: State level distribution of educational institutions for the Ganges Basin in India ................................................................................................................................ 56 Figure 2-36: Province level distribution of educational institutions in Nepal. ........................ 56 Figure 2-37: District level distribution of educational institutions for the basin in Bangladesh ................................................................................................................................ 57 Figure 2-38: Distribution of educational institutions across the Ganges Basin ..................... 57 Figure 2-39: State level distribution of health facilities for the basin in India ........................ 58 Figure 2-40: Province level distribution of health facilities in Nepal...................................... 58 Figure 2-41: District level distribution of health facilities for the basin in Bangladesh ........... 59 Figure 2-42: Distribution for health facilities for Ganges Basin ............................................ 60 Figure 2-43: State level distribution of other houses in India ............................................... 61 Figure 2-44: Province level distribution of other houses in Nepal ........................................ 61 Figure 2-45: District level distribution of other houses in Bangladesh .................................. 62 Figure 2-46: Distribution for other houses for Ganges Basin ............................................... 62 Figure 2-47: Distribution of structural classes used in India for housing .............................. 64 Figure 2-48: Distribution of structural classes used in Nepal for housing ............................. 64 Figure 2-49: Distribution of structural classes used in Bangladesh for housing ................... 64 Figure 2-50: State-level distribution of road lengths for the basin in India ............................ 65 Figure 2-51: Province level distribution of road lengths for the basin in Nepal ..................... 65 Figure 2-52: District level distribution of road lengths for the basin in Bangadesh ............... 66 Figure 2-53: Road network for the Ganges Basin................................................................ 66 Figure 2-54: State level distribution of railway tracks for the Ganges basin in India ............. 67 Figure 2-55: Province level distribution of railway track lengths for the basin in Nepal ........ 67 Figure 2-56: District level distribution of railway track lengths for the basin in Bangadesh ... 68 Figure 2-57: Rail network for the study area ........................................................................ 68 Figure 2-58: State level distribution of area under Rice cultivation for the basin in India (2010/11) ................................................................................................................. 69 Figure 2-59: State level distribution of Rice production for the basin in India (2010/11) ....... 69 Figure 2-60: State level distribution of area under Wheat cultivation for the basin in India (2010/11) ................................................................................................................. 70 Figure 2-61: State level distribution of Wheat production for the basin in India (2010/11) .... 70 Figure 2-62: State level distribution of area under Maize cultivation for the basin in India (2010/11) ................................................................................................................. 71 Figure 2-63: State level distribution of Maize production for the basin in India (2010/11) .... 71 Figure 2-64: Province level distribution of area under Rice cultivation for the basin in Nepal (2012/13) ................................................................................................................. 72 Figure 2-65: Province level distribution of Rice production for the basin in Nepal (2012/13) 72 Final Report: Volume I Confidential Page 9 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Figure 2-66: Province level distribution of area under Wheat cultivation for the basin in Nepal (2012/13) ................................................................................................................. 73 Figure 2-67: Province level distribution of Wheat production for the basin in Nepal (2012/13) ................................................................................................................................ 73 Figure 2-68: Province level distribution of area under Maize cultivation for the basin in Nepal (2012/13) ................................................................................................................. 74 Figure 2-69: Province level distribution of Maize production for the basin in Nepal (2012/13) ................................................................................................................................ 74 Figure 2-70: District level distribution of area under Rice cultivation for the basin in Bangladesh (2009/10) ............................................................................................. 75 Figure 2-71: District level distribution of Rice production for the basin in Bangladesh (2009/10) ................................................................................................................. 75 Figure 2-72: District level distribution of area under Wheat cultivation for the basin in Bangladesh (2009/10) ............................................................................................. 76 Figure 2-73: District level distribution of Wheat production for the basin in Bangladesh (2009/10) ................................................................................................................. 76 Figure 2-74: State level distribution of total exposure value of houses for the basin in India 81 Figure 2-75: Province level distribution of total exposure value of houses in Nepal ............. 81 Figure 2-76: District level distribution of total exposure value of houses in Bangladesh ...... 82 Figure 2-77: State level distribution of exposure value of roads in India .............................. 83 Figure 2-78: Province level distribution of exposure value for roads in Nepal ...................... 84 Figure 2-79: District level distribution of road exposure values for Bangladesh ................... 84 Figure 2-80: State level exposure values of railway tracks in India ...................................... 85 Figure 2-81: Province level distribution of exposure values of railway tracks in Nepal. ........ 86 Figure 2-82: District level railway exposure values of Bangladesh ...................................... 86 Figure 2-83: Exposure value of Rice sown in India during 2010/11 ..................................... 88 Figure 2-84: Exposure value of Wheat sown in India during 2010/11 .................................. 89 Figure 2-85: Exposure value of Maize sown in India during 2010/11 ................................... 89 Figure 2-86: Exposure value of Rice sown in Nepal during 2012/13.................................... 90 Figure 2-87: Exposure value of Wheat sown in Nepal during 2012/13 ................................ 90 Figure 2-88: Exposure value of Maize sown in Nepal during 2012/13 ................................. 91 Figure 2-89: Exposure value of Rice sown in Bangladesh during 2009/10 .......................... 91 Figure 2-90: Exposure value of Wheat sown in Bangladesh during 2009/10 ....................... 92 Figure 3-1: Flood hazard assessment framework ................................................................ 93 Figure 3-2: Location of rainfall stations used in the study .................................................... 95 Figure 3-3: Flow gauge station locations ............................................................................. 97 Figure 3-4: Elevation map of the study area ........................................................................ 98 Figure 3-5: LULC map of Ganges River basin ..................................................................... 99 Figure 3-6: Soil map for the study area ............................................................................. 100 Final Report: Volume I Confidential Page 10 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Figure 3-7: Flood Footprint: 2007 flood ............................................................................. 104 Figure 3-8: Delineated catchments and major sub basins of Ganges River Basin ............. 107 Figure 3-9: HEC RAS model set up for the study area ...................................................... 109 Figure 3-10: Location map of the gauge stations used in HEC RAS model ....................... 110 Figure 3-11: Location map of the gauge stations used in calibration-validation of HEC RAS model .................................................................................................................... 111 Figure 3-12: Deviation between simulated and observed water levels at various gauge stations for the calibration event of 1978 ............................................................... 112 Figure 3-13: Deviation between simulated and observed water levels at various gauge stations for the calibration event of 1981 ............................................................... 112 Figure 3-14: Deviation between simulated and observed water levels at various gauge stations for the calibration event of 1988 ............................................................... 113 Figure 3-15: Deviation between simulated and observed water levels at various gauge stations for the calibration event of 1996 ............................................................... 113 Figure 3-16: Deviation between simulated and observed water levels at various gauge stations for the calibration event of 1998 ............................................................... 114 Figure 3-17: Deviation between simulated and observed water levels at various gauge stations for the validation event of 2004................................................................. 114 Figure 3-18: Deviation between simulated and observed water levels at various gauge stations for the validation event of 2008................................................................. 115 Figure 3-19: Flood Extent Map of July, 2004 flood event (Source: DFO) ........................... 116 Figure 3-20: Flood Extent Map of September, 2008 flood event (Source: DFO) ................ 116 Figure 3-21: Comparison between observed (left) and modeled (right) flood extents of July, 2004 event ............................................................................................................ 117 Figure 3-22: Comparison between observed (left) and modeled (right) flood extents of September, 2008 event ......................................................................................... 117 Figure 3-23: L moment ratio diagram ................................................................................ 120 Figure 3-24: Flood hazard map for 2-year return period for Ganges basin ........................ 121 Figure 3-25: Flood hazard map for 5-year return period for Ganges basin ........................ 121 Figure 3-26: Flood hazard map for 10-year return period for Ganges basin ...................... 122 Figure 3-27: Flood hazard map for 25-year return period for Ganges basin ...................... 122 Figure 3-28: Flood hazard map for 50-year return period for Ganges basin ...................... 123 Figure 3-29: Flood hazard map for 100-year return period for Ganges basin .................... 123 Figure 4-1: Basic building blocks of risk assessment methodology ................................... 124 Figure 4-2: Various types of buildings affected due to flood in Ganges Basin ................... 125 Figure 4-3: Development methodology for vulnerability functions ...................................... 126 Figure 4-4 : Example USACE depth – damage function - one storey no basement ........... 128 Figure 4-5: Vulnerability function for buildings ................................................................... 130 Figure 4-6: Vulnerability function for Infrastructure ............................................................ 130 Figure 4-7: Vulnerability function for Agriculture ................................................................ 132 Final Report: Volume I Confidential Page 11 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Figure 4-8: Sample of AAL ................................................................................................ 133 Figure 4-9: Sample of LEC ................................................................................................ 133 Figure 4-10: District/sub-district overlap with flood hazard ................................................ 134 Figure 4-11: Railway lines affected with flood hazard ........................................................ 135 Figure 4-12: Affected population reported by Giriraj and Amarnath, IWMI (left) and RMSI’s modeled affected population (right) ....................................................................... 139 Figure 4-13: Validation of affected population for Supaul district ....................................... 140 Figure 4-14: Validation of affected population for Burhi Gandak sub-basin ....................... 140 Figure 4-15: State wise affected area reported by GFCC annual report of year 2010-11 .. 141 Figure 4-16: State wise maximum affected area from modeled result ............................... 141 Figure 4-17: Validation of residential buildings losses for Supaul district ........................... 143 Figure 4-18: Validation of residential buildings losses for Burhi Gandak Basin .................. 144 Figure 4-19: Validation of agriculture losses for Supaul district ......................................... 145 Figure 4-20: Validation of agriculture losses for Burhi Gandak Basin ................................ 145 Figure 5-1: Total number of percentage affected persons in Ganges Basin for various return period flood events ................................................................................................ 148 Figure 5-2: Sub-basin level distribution of percentage affected persons in Ganges Basin for 100-year return period flood event ......................................................................... 148 Figure 5-3: Spatial distribution of total affected persons (in thousands) for a 100-Year Return Period Flood .......................................................................................................... 149 Figure 5-4: Percentage distribution of AAL for various assests class ................................ 151 Figure 5-5: Exceedance probability curve showing the total losses due to flood for all exposure types – Ganges Basin ............................................................................ 151 Figure 5-6: Percentage distribution of AAL for various occupancy types ........................... 152 Figure 5-7: Exceedance probability curve showing the total losses due to flood for Buildings – Residential.......................................................................................................... 153 Figure 5-8: Sub-basin level AAL (million INR) for Buildings – Residential.......................... 154 Figure 5-9: AAL (million INR) due to flood for Buildings – Residential ............................... 154 Figure 5-10: Exceedance probability curve showing the total losses due to flood for Buildings – Commercial ........................................................................................................ 155 Figure 5-11: Sub-basin level AAL (million INR) for Buildings – Commercial ...................... 156 Figure 5-12: AAL (million INR) due to flood for Buildings – Commercial ............................ 156 Figure 5-13: Exceedance probability curve showing the total losses due to flood for Buildings – Industrial............................................................................................................. 157 Figure 5-14: Sub-basin level AAL (million INR) for Buildings – Industrial ........................... 158 Figure 5-15: AAL(thousand INR) due to flood for Buildings – Industrial ............................. 158 Figure 5-16: Exceedance probability curve showing the total losses due to flood for Buildings - Educational Institutions ....................................................................................... 159 Figure 5-17: Sub-basin level AAL (million INR) for Buildings – Educational Institutions ..... 160 Figure 5-18: AAL (thousand INR) due to flood for Buildings – Educational Institutions ...... 161 Final Report: Volume I Confidential Page 12 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Figure 5-19: Exceedance probability curve showing the total losses due to flood for Buildings - Health Facilities ................................................................................................... 162 Figure 5-20: Sub-basin level AAL (million INR) for Buildings – Health Facilities ................ 163 Figure 5-21: AAL (thousand INR) due to flood for Buildings – Health Facilities ................. 163 Figure 5-22: Exceedance probability curve showing the total losses due to flood for Buildings – Others ................................................................................................................ 164 Figure 5-23: Sub-basin level AAL (million INR) for Buildings – Others .............................. 165 Figure 5-24: AAL (million INR) due to flood for Buildings – Others .................................... 165 Figure 5-25: Exceedance probability curve showing the total losses due to flood for Infrastructure – Road network ............................................................................... 167 Figure 5-26: Sub-basin level AAL (million INR) for Infrastructure – Road network ............. 167 Figure 5-27: AAL (thousand INR) due to flood for Infrastructure – Road network .............. 168 Figure 5-28: Exceedance probability curve showing the total losses due to flood for Infrastructure – Rail network .................................................................................. 169 Figure 5-29: Sub-basin level AAL (million INR) for Infrastructure – Rail network ............... 170 Figure 5-30: AAL (thousand INR) due to flood for Infrastructure – Rail network ................ 170 Figure 5-31: Exceedance probability curve showing the total losses due to flood for Agriculture – Rice .................................................................................................. 172 Figure 5-32: Sub-basin level AAL (million INR) for Agriculture – Rice ............................... 172 Figure 5-33: AAL (thousand INR) due to flood for Agriculture – Rice................................. 173 Figure 5-34: Exceedance probability curve showing the total losses due to flood for Agriculture – Wheat ............................................................................................... 174 Figure 5-35: Sub-basin level AAL (million INR) for Agriculture – Wheat ............................ 175 Figure 5-36: AAL (thousand INR) due to flood for Agriculture – Wheat ............................. 175 Figure 5-37: Exceedance probability curve showing the total losses due to flood for Agriculture – Maize ................................................................................................ 177 Figure 5-38: Sub-basin level AAL (million INR) for Agriculture – Maize ............................. 177 Figure 5-39: AAL (thousand INR) due to Flood for Agriculture – Maize ............................. 178 Figure 6-1: Exceedance probability curve showing the total losses due to flood for all exposure types – Lower Ganges sub-basin ........................................................... 180 Figure 6-2: AAL (million INR) due to floods for buildings: Residential – Lower Ganges sub- basin ..................................................................................................................... 182 Figure 6-3: Exceedance probability curve showing the total losses due to flood for all exposure types – Bagmati sub-basin ..................................................................... 184 Figure 6-4: AAL (million INR) due to floods for buildings: Residential – Bagmati sub-basin .............................................................................................................................. 186 Figure 6-5: Exceedance probability curve showing the total losses due to flood for all exposure types – Kosi sub-basin ........................................................................... 188 Figure 6-6: AAL (million INR) due to floods for buildings: Residential – Kosi sub-basin .... 190 Figure 6-7: Exceedance probability curve showing the total losses due to flood for all exposure types – Gandak sub-basin ..................................................................... 192 Final Report: Volume I Confidential Page 13 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Figure 6-8: AAL (million INR) due to floods for buildings: Residential – Gandak sub-basin .............................................................................................................................. 194 Figure 6-9: Exceedance probability curve showing the total losses due to flood for all exposure types – Mahananda sub-basin ............................................................... 196 Figure 6-10: AAL (million INR) due to floods for buildings: Residential – Mahananda sub- basin ..................................................................................................................... 198 Figure 6-11: Exceedance probability curve showing the total losses due to flood for all exposure types – Middle Ganges sub-basin .......................................................... 200 Figure 6-12: AAL (million INR) due to floods for buildings: Residential – Middle Ganges sub- basin ..................................................................................................................... 202 Figure 6-13: Exceedance probability curve showing the total losses due to flood for all exposure types – Ghagra sub-basin ...................................................................... 204 Figure 6-14: AAL (million INR) due to floods for buildings: Residential – Ghagra sub-basin .............................................................................................................................. 206 Figure 6-15: Exceedance probability curve showing the total losses due to flood for all exposure types – Kamla-Balan sub-basin.............................................................. 208 Figure 6-16: AAL (million INR) due to floods for buildings: Residential – Kamla-Balan sub- basin ..................................................................................................................... 210 Figure 6-17: Exceedance probability curve showing the total losses due to flood for all exposure types – Yamuna sub-basin ..................................................................... 212 Figure 6-18: AAL (million INR) due to floods for buildings: Residential – Yamuna sub-basin .............................................................................................................................. 214 Figure 6-19: Exceedance probability curve showing the total losses due to flood for all exposure types – Ramganga sub-basin................................................................. 216 Figure 6-20: AAL (million INR) due to floods for buildings: Residential – Ramganga sub- basin ..................................................................................................................... 218 Figure 6-21: Exceedance probability curve showing the total losses due to flood for all exposure types – Gomti sub-basin ........................................................................ 220 Figure 6-22: AAL (million INR) due to floods for buildings: Residential – Gomti sub-basin 222 Figure 6-23: Exceedance probability curve showing the total losses due to flood for all exposure types – Sone sub-basin.......................................................................... 224 Figure 6-24: AAL (million INR) due to floods for buildings: Residential – Sone sub-basin . 226 Figure 6-25: Exceedance probability curve showing the total losses due to flood for all exposure types – Upper Ganges sub-basin ........................................................... 228 Figure 6-26: AAL (million INR) due to floods for buildings: Residential – Upper Ganges sub- basin ..................................................................................................................... 230 Figure 6-27: Exceedance probability curve showing the total losses due to flood for all exposure types – Tons sub-basin .......................................................................... 232 Figure 6-28: AAL (million INR) due to floods for buildings: Residential – Tons sub-basin . 234 Figure 6-29: Exceedance probability curve showing the total losses due to flood for all exposure types – Sind sub-basin ........................................................................... 236 Figure 6-30: AAL (million INR) due to floods for buildings: Residential – Sind sub-basin .. 238 Final Report: Volume I Confidential Page 14 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Figure 6-31: Exceedance probability curve showing the total losses due to flood for all exposure types – Ken sub-basin............................................................................ 240 Figure 6-32: AAL (million INR) due to floods for buildings: Residential – Ken sub-basin .. 242 Figure 6-33: Exceedance probability curve showing the total losses due to flood for all exposure types – Betwa sub-basin ........................................................................ 244 Figure 6-34: AAL (million INR) due to floods for buildings: Residential – Betwa sub-basin246 Figure 6-35: Exceedance probability curve showing the total losses due to flood for all exposure types – Chambal sub-basin .................................................................... 248 Figure 6-36: AAL (million INR) due to floods for buildings: Residential – Chambal sub-basin .............................................................................................................................. 250 Figure 8-1: Hydrological Station network of Ganges Basin (Source: India-WRIS) ............. 257 Final Report: Volume I Confidential Page 15 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia List of Tables Table 1-1: Country wise composition of Ganges Basin ....................................................... 30 Table 1-2: Areas occupied by the basin in the four countries .............................................. 31 Table 2-1: Housing categories by construction materials and structural types ..................... 63 Table 2-2: Projects involving field work in Study Area ......................................................... 77 Table 2-3: Areas of various housing in India ....................................................................... 77 Table 2-4: Unit replacement costs for different types of structures in India .......................... 79 Table 2-5: Unit replacement costs for different types of structures in Nepal ........................ 80 Table 2-6: Unit replacement costs for different types of structures in Bangladesh ............... 80 Table 2-7: Types of roads and associated costs in India ..................................................... 83 Table 2-8: Weighted average unit cost calculation for roads ............................................... 83 Table 2-9: Exposure values for crops in India ..................................................................... 87 Table 2-10: Exposure values for crops in Nepal .................................................................. 87 Table 2-11: Exposure values for crops in Bangladesh......................................................... 88 Table 3-1: Station wise rainfall data records........................................................................ 94 Table 3-2: Availability of discharge data .............................................................................. 95 Table 3-3: Flood loss summary: Bihar, 1979 to 2006 ........................................................ 101 Table 3-4: Flood loss summary: Uttar Pradesh, 1973 flood ............................................... 102 Table 3-5: Flood loss summary: Bihar, 2007 flood ............................................................ 105 Table 3-6: Major sub basins considered in study............................................................... 106 Table 4-1: Classification of buildings based on the construction material .......................... 125 Table 4-2: Suitability of AIT damage-functions .................................................................. 128 Table 4-3: Suitability of the New Orleans District damage functions .................................. 129 Table 4-4: Coefficient of damage functions for Rice crop at different flood heights ............ 131 Table 4-5: Validation for affected population ..................................................................... 138 Table 4-6: Validation for affected area .............................................................................. 141 Table 4-7: Validation for residential housing units ............................................................. 142 Table 4-8: Validation for residential housing losses ........................................................... 142 Table 4-9: Validation for AAL ............................................................................................ 144 Table 4-10: Validation for Agriculture losses ..................................................................... 145 Table 4-11: Validation for public property losses ............................................................... 146 Table 5-1: Total number (in millions) of persons affected due to flood in Ganges Basin for various return period flood events.......................................................................... 147 Table 5-2: Total number (in thousands) of persons affected due to flood in Ganges Basin for various return periods flood events – Lower Ganges, Bagmati, and Kosi subbasins .............................................................................................................................. 149 Final Report: Volume I Confidential Page 16 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Table 5-3: Total number (in millions) of children (age 0-6 years), Scheduled Castes (SCs) and Scheduled Tribes (STs) affected due to floods in the Indian part of Ganges Basin for various return periods ....................................................................................... 150 Table 5-4: PML and AAL due to flood events of various return periods for various exposure classes .................................................................................................................. 150 Table 5-5: AAL due to flood for various occupancy types .................................................. 152 Table 5-6:PML and AAL due to flood events of various return periods for buildings - Residential ............................................................................................................ 153 Table 5-7: PML and AAL due to flood events of various return periods for buildings - Commercial ........................................................................................................... 155 Table 5-8: PML and AAL due to flood events of various return periods for buildings - Industrial................................................................................................................ 157 Table 5-9: PML and AAL due to flood events of various return periods for buildings - Educational Institutions.......................................................................................... 159 Table 5-10: PML and AAL due to flood events of various return periods for buildings - Health Facilities ................................................................................................................ 161 Table 5-11: PML and AAL due to flood events of various return periods for buildings - Others .............................................................................................................................. 164 Table 5-12: Affected length (km) due to flood events of various return periods for Infrastructure – Road network ............................................................................... 166 Table 5-13: PML and AAL due to flood events of various return periods for Infrastructure – Road network ........................................................................................................ 166 Table 5-14: Affected length (km) due to flood events of various return periods for Infrastructure – Rail network .................................................................................. 168 Table 5-15: PML and AAL due to flood events of various return periods for Infrastructure – Rail network........................................................................................................... 169 Table 5-16: Affected cultivated area (thousand hectare) due to flood events of various return periods for Agriculture – Rice ................................................................................ 171 Table 5-17: PML and AAL due to flood events of various return periods for Agriculture – Rice .............................................................................................................................. 171 Table 5-18: Affected cultivated area (thousand hectare) due to flood events of various return periods for Agriculture – Wheat ............................................................................. 173 Table 5-19: PML and AAL due to flood events of various return periods for Agriculture – Wheat .................................................................................................................... 174 Table 5-20: Affected cultivated area (thousand hectare) due to flood events of various return periods for Agriculture – Maize .............................................................................. 176 Table 5-21: PML and AAL due to flood events of various return periods for Agriculture – Maize .................................................................................................................... 176 Table 6-1: Total number (in thousands) of persons affected due to various return period flood events – Lower Ganges sub-basin ........................................................................ 179 Table 6-2: PML and AAL due to flood events of different return periods for various exposure classes – Lower Ganges sub-basin ....................................................................... 180 Table 6-3: State/Province wise breakup of PML and AAL – Lower Ganges sub-basin ...... 181 Table 6-4: Blocks/districts having maximum AAL – Lower Ganges sub-basin ................... 182 Final Report: Volume I Confidential Page 17 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Table 6-5: Total number (in thousands) of persons affected due to various return period flood events – Bagmati sub-basin .................................................................................. 183 Table 6-6: PML and AAL due to flood events of different return periods for various exposure classes – Bagmati sub-basin ................................................................................. 183 Table 6-7: State/Province wise breakup of PML and AAL – Bagmati sub-basin ................ 185 Table 6-8: Blocks/districts having maximum AAL – Bagmati sub-basin ............................. 186 Table 6-9: Total number (in thousands) of persons affected due to various return period flood events – Kosi sub-basin ........................................................................................ 187 Table 6-10: PML and AAL due to flood events of different return periods for various exposure classes – Kosi sub-basin ....................................................................................... 187 Table 6-11: State/Province wise breakup of PML and AAL – Kosi sub-basin .................... 189 Table 6-12: Total number (in thousands) of persons affected due to various return period flood events – Gandak sub-basin .......................................................................... 191 Table 6-13: PML and AAL due to flood events of different return periods for various exposure classes – Gandak sub-basin .................................................................................. 191 Table 6-14: State/Province wise breakup of PML and AAL – Gandak sub-basin ............... 193 Table 6-15: Total number (in thousands) of persons affected due to various return period flood events – Mahananda sub-basin .................................................................... 195 Table 6-16: PML and AAL due to flood events of different return periods for various exposure classes – Mahananda sub-basin ........................................................................... 195 Table 6-17: State/Province wise breakup of PML and AAL – Mahananda sub-basin ........ 197 Table 6-18: Total number (in thousands) of persons affected due to various return period flood events – Middle Ganges sub-basin ............................................................... 199 Table 6-19: PML and AAL due to flood events of different return periods for various exposure classes – Middle Ganges sub-basin ...................................................................... 199 Table 6-20: State/Province wise breakup of PML and AAL – Middle Ganges sub-basin ... 201 Table 6-21: Total number (in thousands) of persons affected due to various return period flood events – Ghagra sub-basin ........................................................................... 203 Table 6-22: PML and AAL due to flood events of different return periods for various exposure classes – Ghagra sub-basin .................................................................................. 203 Table 6-23: State/Province wise breakup of PML and AAL – Ghagra sub-basin ............... 205 Table 6-24: Total number (in thousands) of persons affected due to various return period flood events – Kamla-Balan sub-basin................................................................... 207 Table 6-25: PML and AAL due to flood events of different return periods for various exposure classes – Kamla-Balan sub-basin .......................................................................... 207 Table 6-26: State/Province wise breakup of PML and AAL – Kamla-Balan sub-basin ....... 209 Table 6-27: Total number (in thousands) of persons affected due to various return period flood events – Yamuna sub-basin.......................................................................... 211 Table 6-28: PML and AAL due to flood events of different return periods for various exposure classes – Yamuna sub-basin ................................................................................. 211 Table 6-29: State/Province wise breakup of PML and AAL – Yamuna sub-basin .............. 213 Final Report: Volume I Confidential Page 18 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Table 6-30: Total number (in thousands) of persons affected due to various return period flood events – Ramganga sub-basin ..................................................................... 215 Table 6-31: PML and AAL due to flood events of different return periods for various exposure classes – Ramganga sub-basin ............................................................................. 215 Table 6-32: State/Province wise breakup of PML and AAL – Ramganga sub-basin .......... 217 Table 6-33: Total number (in thousands) of persons affected due to various return period flood events – Gomti sub-basin ............................................................................. 219 Table 6-34: PML and AAL due to flood events of different return periods for various exposure classes – Gomti sub-basin..................................................................................... 219 Table 6-35: State/Province wise breakup of PML and AAL – Gomti sub-basin .................. 221 Table 6-36: Total number (in thousands) of persons affected due to various return period flood events – Sone sub-basin .............................................................................. 223 Table 6-37: PML and AAL due to flood events of different return periods for various exposure classes – Sone sub-basin ...................................................................................... 223 Table 6-38: State/Province wise breakup of PML and AAL – Sone sub-basin ................... 225 Table 6-39: Total number (in thousands) of persons affected due to various return period flood events – Upper Ganges sub-basin ................................................................ 227 Table 6-40: PML and AAL due to flood events of different return periods for various exposure classes – Upper Ganges sub-basin ....................................................................... 227 Table 6-41: State/Province wise breakup of PML and AAL – Upper Ganges sub-basin .... 229 Table 6-42: Total number (in thousands) of persons affected due to various return period flood events – Tons sub-basin ............................................................................... 231 Table 6-43: PML and AAL due to flood events of different return periods for various exposure classes – Tons sub-basin ...................................................................................... 231 Table 6-44: State/Province wise breakup of PML and AAL – Tons sub-basin ................... 233 Table 6-45: Total number (in thousands) of persons affected due to various return period flood events – Sind sub-basin ................................................................................ 235 Table 6-46: PML and AAL due to flood events of different return periods for various exposure classes – Sind sub-basin ....................................................................................... 235 Table 6-47: State/Province wise breakup of PML and AAL – Sind sub-basin .................... 237 Table 6-48: Total number (in thousands) of persons affected due to various return period flood events – Ken sub-basin ................................................................................ 239 Table 6-49: PML and AAL due to flood events of different return periods for various exposure classes – Ken sub-basin ........................................................................................ 239 Table 6-50: State/Province wise breakup of PML and AAL – Ken sub-basin ..................... 241 Table 6-51: Total number (in thousands) of persons affected due to various return period flood events – Betwa sub-basin ............................................................................. 243 Table 6-52: PML and AAL due to flood events of different return periods for various exposure classes – Betwa sub-basin .................................................................................... 243 Table 6-53: State/Province wise breakup of PML and AAL – Betwa sub-basin ................. 245 Table 6-54: Total number (in thousands) of persons affected due to various return period flood events – Chambal sub-basin......................................................................... 247 Final Report: Volume I Confidential Page 19 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Table 6-55: PML and AAL due to flood events of different return periods for various exposure classes – Chambal sub-basin ................................................................................ 247 Table 6-56: State/Province wise breakup of PML and AAL – Chambal sub-basin ............. 249 Table 7-1: Districts/sub-districts having maximum affected persons.................................. 252 Table 7-2: Districts/sub-districts having maximum residential building losses ................... 252 Table 7-3: Districts/sub-districts having maximum commercial building losses ................. 253 Table 7-4: Districts/sub-districts having maximum industrial building losses ..................... 253 Table 7-5: Districts/sub-districts having maximum road losses.......................................... 254 Table 7-6: Districts/sub-districts having maximum railway losses ...................................... 254 Table 8-1: List of Hydrological Station lying in Ganges Basin ............................................ 258 Final Report: Volume I Confidential Page 20 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Abbreviations Used Abbreviation/Acronym Expanded Form AAL Average Annual Loss CWC Central Water Commission DEM Digital Elevation Model DFO Dartmouth Flood Observatory DHM Department of Hydrology and Meteorology ELT Event Loss Table EP Exceedance Probability FAO Food and Agriculture Organization ft feet GIS Geographic Information System GPS Global Positioning System Ha Hectares HEC Hydrologic Engineering Center HEC-RAS Hydrologic Engineering Centre’s River Analysis System HVRA Hazard Vulnerability Risk Assessment INR Indian Rupees IT Information Technology km Kilometer LEC Loss Exceedance Curve LULC Land use land cover m meter MDR Mean Damage Ratio mm millimeter MS Microsoft NSE Nash-Sutcliffe measure of efficiency pa Per Annum PML Probable Maximum Loss RAS River Analysis System SC Scheduled Caste ST Scheduled Tribe STDEV/SD Standard deviation SRTM Shuttle Radar Topographic Mission RMSE Root-mean square error UP Uttar Pradesh USGS U.S. Geological Survey WB West Bengal WRD Water Resources Department WRIS Water Resource Information System of India Final Report: Volume I Confidential Page 21 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Executive Summary Background The World Bank has contracted RMSI to carry out a risk assessment study for the entire Ganges Basin. The aim of this study is to help decision makers for better mitigation planning to reduce losses due to frequent floods in the basin. Ganges basin is spread across parts of four countries viz. India, Nepal, China, and Bangladesh. The total area of the basin is around 984,076 km2. Out of these four countries, India has the maximum share in area, which is around 80% of total basin area. Country wise area and population distribution is shown in the following table and figure. Area in Population in Country Ganges Basin Ganges basin 2 in km (Million) India 790,223 441.47 Nepal 147,706 26.25 China 39,133 0.37- Bangladesh 7,014 6.36 474.45 Total 984,076 % Area distribution Figure - 1: Country wise area and population in Ganges Basin Objective and Scope The main objectives of this assignment are to understand the geographical impacts of floods on various sectors (viz. building stock of various categories, infrastructure, and agriculture), exposed in the Ganges basin spread across the four countries. This will help the concerned authorities in identification of priority areas for flood mitigation work. Key activities include: 1. Development of exposure (assets at risk) database for various asset classes such as buildings, infrastructure, demography, and agriculture 2. Development of one-dimensional steady flow probabilistic flood hazard model for 2, 5, 10, 25, 50, and 100-year return periods using hydraulic modeling for the entire Ganges Basin up to its confluence with the Brahmaputra River in Bangladesh 3. Development of vulnerability curves (damage functions) for each of the exposure (asset) classes 4. Estimation of direct losses at sub district/district levels (AAL or Average Annual Loss and loss exceedance curve) Hazard Model Formulation The hazard model formulation consists of database development, flood hazard (hydraulic) modeling for various return periods, and study of the impact of flood hazard on various exposures. The flood hazard grids for all return periods of 2, 5, 10, 25, 50, and 100-years are overlaid on the sub-district/district boundaries to find the percent affected area and mean flood depth for each sub-district/district and for each return period. The mean damage ratio (MDR) value is selected from respective MDR curves based on the exposure class and flood depth. Final Report: Volume I Confidential Page 22 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia The affected exposure value for each class for each sub-district is multiplied by the respective MDR value to get the loss value. To calculate affected population for various return period flood events, the percentage of flood affected area of each sub-district/district is multiplied by the respective population numbers. This process has been carried out for all sub-districts/districts of the three countries viz. India, Nepal, and Bangladesh and totaled to arrive at the affected population for the entire Ganges Basin. Exposure Data Collection and Management Developing exposure data is a critical component of any risk assessment study. The datasets required for the study have been collected from various agencies during the initial part of the study. These datasets have been brought into a usable format. The data types include geo-spatial, meteorological, hydrological, demographic, infrastructure, and crop area production data. Exposure data constitutes population, the built environment, systems that support infrastructure and livelihood functions, or other elements present in the hazard zone, which are subjected to potential losses. Therefore, modeling vulnerability of a system to natural hazards involves establishing a relationship between the potential damageability of critical exposure elements and different levels of local flood intensity. An exposure database of population, buildings, infrastructure assets, and crops was developed at sub-district/district level for the three countries in the Ganges Basin. An inventory of vulnerable buildings, infrastructure, demographics, and any other asset elements (e.g., crops) present in hazard zones, constituting critical exposure elements, was developed, which were considered for risk assessment. The spatial distribution of exposure and their values are important to decision makers from the planning point of view. The sub-district/district level aggregated exposure values for residential, commercial, industrial, essential facilities, infrastructure, and agriculture exposure classes have been calculated. The total estimated value of exposure in all categories present in the study area is around INR 67,465 billion, out of which building exposure accounts for about 94.9% of the total value, infrastructure exposure (roads and rail) and agriculture (rice, wheat and maize) for about 2.6% and 2.5% respectively. Out of the total exposure value of the basin, India has the maximum share of INR 64,990 billion whereas Nepal and Bangladesh have INR 1,839 billion and INR 636 billion respectively. The total exposure if divided with the total population of the basin will give us a broad estimate of exposure per capita for the basin which is around USD 1970. According to International Monetary Fund World (IMF) estimates1, GDP per capita of India for 2016 is USD 1,942/- in comparison to RMSI estimates of USD 1,970/-. It can be noted that the RMSI estimates are at higher side which can be attributed to the fact that all the exposure values have been estimated, based on the information collected from various public source websites. For Nepal and Bangladesh even this much basic information was sparsely available. RMSI team has faced a lot of challenge to collect and collate all these information to develop the exposure database for the basin. Flood Hazard Assessment RMSI developed a database with regard to catchment area, river network, river cross sections at various locations, water levels at various river streams, and land-use land-cover classes, etc. required for the assessment. Digital elevation model (DEM), of 90-m resolution 1 http://www.imf.org/external/pubs/ft/weo/2015/01/weodata/weorept.aspx?pr.x=33&pr.y=13&sy=1980&e y=2020&scsm=1&ssd=1&sort=country&ds=.&br=1&c=534&s=NGDPDPC%2CPPPPC&grp=0&a Final Report: Volume I Confidential Page 23 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia from SRTM (Shuttle Radar Topographic Mission) has been used to extract the river network. This DEM has also been used to extract the elevations at various cross section locations for hydraulic modeling. Analysis has been done for the study area to identify the basin and catchment characteristics and determine water flows in each catchment for given rainfall in the entire Ganges Basin. Hydrologic Engineering Center’s River Analysis System (HEC-RAS) has been used for hydraulic modeling. Once the model has been set-up for the entire Ganges Basin, the model has been calibrated for five historical flood events of 1978, 1981, 1988, 1996, and 1998. This calibrated model has been, then, further validated for two historical flood events of 2004 and 2008. By using this validated model, flood hazard maps depicting flood extents and flood depths have been derived by performing one-dimensional hydraulic routing through the river system for 2, 5, 10, 25, 50, and 100-year return period discharges. The hazard analysis shows that around 75,000 sq. km. of the basin gets inundated due to a 100-year return period flood. Vulnerability The damage or vulnerability function relates hazard intensity to the economic loss potential. The vulnerability module is a critical component of any natural hazard risk model that links the hazard (natural) module to the exposure (manmade) to deduce expected probability of specific loss scenarios. Damage loss can be measured with respect to a building damage ratio defined as the repair or replacement cost divided by the total replacement cost for the building (structural damage) and infrastructure elements. This method has allowed vulnerability assessment to be carried out with limited loss data of historical events. For applying vulnerability, the building exposure in the basin is classified into typical seven classes of buildings. The building types are selected in such a way that each type reflects distinct vulnerability against flood. The categories are selected based on the main construction material used in roofs and walls. Similar vulnerability functions have been developed for infrastructure and agriculture. Flood Risk Assessment Risk is the uncertainty of future losses – if we perfectly know a future loss, it is simply a cost, not a risk. Risk is uncertain with regard to the causative hazard event (flood), and its location, date and time of occurrence, and the degree or amount of damage to assets caused by the flood event, and what losses accrue due to the damage. Since risks are uncertain, they must be stated probabilistically, which is expressed in terms of a Loss Exceedance Curve (LEC, also sometimes termed an Exceedance Probability, or EP curve). 2, 5, 10, 25, 50, and 100-year return-period flood hazard maps have been developed. Direct losses are calculated for these return period scenario events and for all types of exposures at risk like residential, commercial, industrial buildings, essential facilities (schools, hospitals), infrastructure (rail and road networks), and major crops. Using the LEC, losses have been estimated for the return period events. Furthermore, GIS based risk maps showing Annual Average Losses (AAL) and losses for various key return periods have been generated showing the areas likely to get affected at the sub- district/district level. In the present context, the direct loss is the hazard-induced losses in terms of financial losses for various structures based on their valuations. The spatial distribution of the modeled risk outputs are portrayed in the form of maps showing the hazard, exposure, and risk characteristics. The temporal characteristics of the modeled risk outputs are depicted in the form of LECs. Final Report: Volume I Confidential Page 24 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Findings One of the major outcomes of this study shows that the average annual loss due to flood is highest for India (approx INR 40,802 million), followed by Bangladesh (approx INR 1,298 million), and Nepal (approx INR 242 million). Summary of affected population and AAL for severely affected sub-basins has been shown in the following figure. F i g u r e - 2 : S u m m a r y o f a f f e c t e d p o p u l a t i o n a n d A AL f o r s e ve r e l y a f f e c t e d s u b - b a s i n s Affected Population The Indian state of Bihar is likely to have the highest affected population (approximate 18 million people) once in a two-year period. Uttar Pradesh follows with a likely affected population of around 17.3 million. In Bangladesh, the Rajshahi state is likely to be the worst affected with 0.99 million people, whereas in case of Nepal, Eastern province is likely to have an affected population of around 0.15 million. Final Report: Volume I Confidential Page 25 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Figure - 3: Population affected b y Flood (RP 2 Year) in Gang es basin Economic losses – Average Annual Loss (AAL) The Indian state of Bihar is expected to have maximum losses as most of the assets and the population in this state are in flood prone areas. These losses come from the various sub- basins that lie in the state, i.e. Lower Ganges, Bagmati, Kosi, Gandak, Mahananda, Ghagra, Kamla-Balan, and Sone. F i g u r e - 4 : A v e r a g e a n n u a l l o s s b y v a r i o u s s t a t e s a n d p r o vi n c e s Final Report: Volume I Confidential Page 26 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia The second most severely affected state in India is Uttar Pradesh (UP). It comprises almost all the sub-basins of the Ganges, i.e., Betwa, Chambal, Gandak, Ghagra, Gomti, Ken, Lower Ganges, Middle Ganges, Ramganga, Sind, Sone, Tons, Upper Ganges, and Yamuna. Ghagra and Middle Ganges sub-basins have the maximum share in losses and flood affected persons in Uttar Pradesh. The third most severely affected state in India is Delhi, which lies completely in the catchment area of the Yamuna sub-basin. The state of West Bengal in India also witnesses substantial losses due to floods in the Lower Ganges and Mahananda sub-basins. Madhya Pradesh and Rajasthan are the other major contributors in the losses due to floods in the Ganges basin. In Bangladesh, the states of Rajshahi and Kushtia are likely to bear the maximum flood losses among all the four states of Bangladesh in the Ganges Basin. F i g u r e - 5 : S u b b a s i n s h a vi n g m a x i m u m l o s s c o n t r i b u t i o n i n s t a t e / p r o vi n c e A A L In Nepal, the Mid Western province is at maximum risk followed by the Eastern province. They bear maximum flood losses in Nepal in the Ganges Basin. Priority areas for mitigation (structural and non-structural measures) Lower Ganges sub basin shows maximum average annual loss in the majority of exposure classes and could be a priority candidate sub-basin for future mitigation work. Flood protection measures should be provided along the banks of the river in the Lower Ganges sub-basin as an initial measure along with migration of population and assets from the flood plains to safer areas. A separate, detailed study needs to be carried out to find the optimal mitigation measures for this sub-basin. Worst affected areas at the State/Province level for the three countries have also been identified and are provided below. This study recommends that these areas should be taken up on priority for structural and non-structural mitigation measures to reduce flood losses due to future flood events. Final Report: Volume I Confidential Page 27 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia India 1. Bihar: Patna Rural, Dinapur-cum-khagaul, Munger, Chapra, and Sonepur blocks 2. Uttar Pradesh: Varanasi, Allahabad, Mirzapur, Kanpur and Ballia blocks 3. West Bengal: Manikchak, Suti – II, Raghunathganj - II, Farakka and Samserganj blocks Bangladesh 1. Rajshahi: Nawabganj, Rajshahi, Naogaon, and Natore districts 2. Kushtia: Kustia and Meherpur districts Nepal 1. Mid Western province: Dang, Banke, Surkhet, Bardiya and Pyuthan districts 2. Eastern province: Sunsari, Saptari, Udayapur, Jhapa and Bhojpur districts Concluding Remarks To summarize, the study indicates that Bihar and Uttar Pradesh states in India and areas of Bangladesh lying in the basin are highly vulnerable to the floods. At sub-basin level, the Lower Ganges is the most severely affected sub-basin. The findings of this study can be useful in the identifying the priority sub-basins and/or states which need immediate attention for flood mitigation measures. The study can also be treated as base study to identify the sub-basins for flood forecasting by the World Bank. These findings can help minimize the economic losses in the three countries and help reduce the impact of floods on the population. For further improvements, it is recommended to replicate the study for finer resolution at sub-basin level. Also, in order to estimate the probable damage to other vulnerable exposures, other remaining assets (bridge, pipelines, livestock, electric lines, other crops, etc.) at risk and loss due to business interruption should also be included in the study. Local consultants and experts from various fields should be hired in order to validate the various processes involved in the study. The social vulnerability of various ethnic groups and adaptive capacity of people in the flood prone areas of the three countries should also be surveyed and made part of a separate study. The study can be replicated for other basins and sub-basin of the region. Improvements in the data availability, employing process oriented hydro-agrological models, considering detailed aspects of socio economic vulnerability and adaptive capability of various community groups in the study area can further enhance the usefulness of such studies. The sector specific impact of hydro meteorological disasters on the economy of the countries can also be analyzed. Final Report: Volume I Confidential Page 28 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 1 Introduction 1.1 Background The Ganges basin is one of the largest river basins of the world. It lies in India, Nepal, China, and Bangladesh and drains an area of about 9,84,076 sq. km. In India, its catchment lies in the states of Uttar Pradesh, Uttarakhand, Madhya Pradesh, Chhattisgarh, Bihar, Jharkhand Rajasthan, West Bengal, Haryana, Himachal Pradesh and Delhi. It lies between latitude 22.45° N and 31.47° N and longitude 73.37° E and 89.31° E. The Ganges River originates as Bhagirathi from the Gangotri Glaciers in the Himalayas at an elevation of about 7,000 m above mean sea level, in the Uttarkashi district of Uttarakhand. The Bhagirathi is joined by the Alaknanda at Deoprayag and the combined stream under the name Ganges flowing through the mountain region debouches into the plains at Rishikesh. It is joined by a large number of tributaries on both the banks in the course of its total run of about 2,500 km before its outfall into the Bay of Bengal. The important tributaries are the Yamuna, the Ramaganga, the Gomti, the Ghagra, the Son, the Gandak, the Kosi and the Mahananda. At Farakka in West Bengal, the river divides into two arms namely the Padma which flows to Bangladesh and the Bhagirathi which flows through West Bengal. The Yamuna River is the biggest tributary of the Ganges River. It originates from Yamunotri Glacier near Banderpoonch peaks in the Mussourie range of the lower Himalayas at an elevation of about 6,300 m above the mean sea level in the district of Uttarkashi. The Himalayas exercise a dominating influence on climate in the northern region of the Upper Yamuna catchment. In this region, winters are very cold, while summers are moderate. The average annual rainfall varies between 400 mm to 1,500 mm. The entire catchment comes under the influence of the southwest monsoon and a major part of the rainfall is received between June and September. Winter rainfall is scanty and occurs between December and February. In the lower part of the Yamuna basin, temperatures are relatively moderate. In summer, temperatures frequently exceed 40° C. The Chambal River is the largest of the rivers flowing through Rajasthan State. This tributary of the Yamuna is about 900 km long. The total area drained by the Chambal up to its confluence with the Yamuna is about 1,43,000 sq. km out of which about 76,000 sq. km lies in Madhya Pradesh state, about 65,000 sq. km in Rajasthan state and about 1,100 sq. km in Uttar Pradesh. The Ramganga is the one of the major tributaries joining the Ganges River. It rises at an altitude of about 3,100 m in the lower Himalayas near the Lohba village in the Garhwal district of Uttaranchal. The length of the Ramganga River, from the source to the confluence with the Ganges, is about 600 km. During its course, the river flows through a mountainous terrain and has a number of falls and rapids. The river flows entirely in the states of Uttarakhand and Uttar Pradesh. The catchment area of the Ramganga is about 30,000 sq. km. The Gomti River originates near Mainkot, about 3 km east of Pilibhit town in Uttar Pradesh, at an elevation of 200 m. The river drains the area between Ramganga and Ghaghra systems. The total length of the river is about 900 km and it flows entirely in the state of Uttar Pradesh. The total drainage area of the river is 30,000 sq. km. The Gaghra River originates at an elevation of 4,800 m near Mansarover Lake. The river is also known as Manchu and Karnali in Nepal. After flowing for about 70 km in a southeasterly direction, the river enters Nepal. Gaghra enters into India at Kotia Ghat near Royal Bardia National Park, Nepal Ganj, where it is known as the river Girwa for about 25 km. The total catchment area of the Gaghra River is about 132,000 sq. km, out of which 45% falls in India. Final Report: Volume I Confidential Page 29 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia The total length of the Gaghra River before its confluence with the Ganges River is about 1,000 km. The Son River is an important right bank tributary of the Ganges River. The river originates at an elevation of 600 m at Sonbhadra in the Maikala range in Madhya Pradesh. The catchment area of the basin is about 68,000 sq. km. Total length of the river is about 800 km, out of which about 500 km lies in Madhya Pradesh, about 80 km in Uttar Pradesh and the remaining in Jharkhand and Bihar. The river meets the Ganges River about 16 km upstream of Dinapur in the Patna district of Bihar. The Gandak River originates near the Nepal-Tibet border at an altitude of about 7,600 m to the northeast of Dhaulagiri and flows about 100 km in a southeasterly direction in Nepal. After that it debauches into the plains of the Champarann district of Bihar at Trivani. The total length of the river from its source to outfall into the Ganges is about 600 km of which about 380 km lie in Nepal and Tibet. The total drainage area of the river is about 46,300 sq. km of which about 7,600 sq. km is in India. The Kosi River is a major tributary of the Ganges River, which originates at an altitude of 7,000 m in the Himalayas. The total catchment area of the Kosi River is 70,000 sq. km out of which 20,376 sq. km lie in India. Kamla Balan and Bagamati are two major tributaries on the right side of the river in Bihar. The delta of the Ganges is said to begin at the Farakka barrage. About 40 km downstream of Farakka, the river splits into two arms. The right arm, known as the Bhagirathi River, flows towards the south and enters the Bay of Bengal about 150 km downstream of Kolkata. The left arm, known as Padma, turns towards the east and enters Bangladesh. While flowing in Bangladesh, Padma meets the Brahmaputra River at Goalundo. The combined flow, still known as Padma, is joined by Meghna, at Chandpur, about 100 km downstream of Goalundo. Further down, the river flows into the Bay of Bengal. Note: Damodar River basin has not been included in the modeling and analysis under the study due to the reason that it does not originate from Ganges River System and even has a different catchment outlet than Ganges River. Thus, Damodar catchment boundary is shown in dotted line in the maps. The country wise spread of the basin across eleven states in India, ten districts in Bangladesh, five Provinces in Nepal, and seven Counties in China is depicted in Table 1-1. Table 1-1: Country wise composition of Ganges Basin Country State Country Provinces Country Districts Country Counties Uttar Central Dinajpur Dinggye Pradesh Madhya Eastern Kustia Nyalam Pradesh Far Rajasthan Meherpur Tingri Western Mid Bihar Naogaon Burang Western India Nepal Bangladesh China West Bengal Natore Gamba Uttarakhand Nawabgunj Gyirong Jharkhand Pabna Haryana Panchagarh Western Chhattisgarh Rajshahi Sa'gya Himachal Pradesh Thakurgaon Delhi Final Report: Volume I Confidential Page 30 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia The major part of the geographical area of the Ganges basin lies in India and it is the biggest river basin in the country draining an area of 790,223 sq.km, which is slightly more than one fourth (26.3 %) of the total geographical area of the country. The following table shows the share of geographical area of the Ganges Basin in the four countries. Table 1-2: Areas occupied b y the basin in the four countries 2 Country Area occupied by Ganges Basin in km India 790,223 Nepal 147,706 China 39,133 Bangladesh 7,014 Total 984,076 The following figure (Figure 1-1) shows the Ganges basin in the four countries. Figure 1-1: Ganges Basin in India, Nepal, Bangladesh and China Final Report: Volume I Confidential Page 31 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 1.2 Topography The Ganges basin is bounded on the north by the Himalayas, on the west by the Aravallis as well as the ridge separating it from Indus basin, on the south by the Vindhyas and Chhotanagpur Plateau and on the east by the Brahmaputra ridge. The main physical sub- divisions of the Ganges basin are the northern mountains, the Gangetic Plains and the central highlands. The northern mountains comprise of the Himalayan ranges including their foothills. The Gangetic plains, situated between the Himalayas and the Deccan plateau, constitute most of the sub-basin ideally suited for intensive cultivation. The central highlands lying to the south of the great plains consist of mountains, hills and plateaus intersected by valleys and river plains. They are largely covered by forests. Aravali uplands, Bundelkhand upland, Malwa plateau, Vindhyan ranges and Narmada valley lie in this region. 1.3 Climate The Ganges Basin has a tropical climate. The annual average rainfall in the basin varies between 390 mm to 2,000 mm, with an average of 1,000 mm. The climate in the Ganges basin is characterized by a distinct wet season during the period of southwest monsoon (June to October). The southwest monsoon makes landfall at the mouth of the Ganges around the first week of June and advances upstream. By the end of July, the monsoon reaches the western end of the Ganges basin. In the majority of the basin, the rainy season spreads over three months (July, August and September) and usually 80% of the total annual rainfall occurs during this period. In the eastern part of the basin, such as in West Bengal and Bihar, the wet season is longer, usually starting in June and continuing until the end of September or early October. The lowest precipitation in the Gangetic plains occur in Haryana (less than 500 mm per year), with the rainfall increasing downstream until reaching lower Bengal, where nearly 1,600 mm of rainfall occurs. Heavier rainfall continues in the upper Himalayan region, such as in Dehra Dun, where the rainfall is as high as 2,209 mm per year. Snow is also a significant part of precipitation in the higher reaches of the basin. The winter precipitation that occurs in the form of snow in hilly areas accumulates until summer. During summer, the melting of snow contributes to considerable runoff. The average temperature in the basin ranges between 9°C to 40°C. 1.4 Meteorological causes of heavy rainfall over Ganges River Basin The weather in the Ganges basin is characterized by a distinct wet season during the period of southwest monsoon (June to October). The air temperature starts falling with onset of the monsoon from June onwards, making the weather more humid. The southwest monsoon makes landfall at the mouth of the Ganges around the first week of June and advances upstream. By the end of July the monsoon reaches the western end of the Ganges basin. In the majority of the basin, the rainy season spreads over three months (July, August and September) and usually 70% to 80% of the total annual rainfall occurs during this period. In the eastern part of the basin, such as in West Bengal and Bihar, the wet season is longer, usually starting in June and continuing until the end of September or early October. Lowest rainfall occurs in Haryana (less than 500 mm per annum) with the rainfall increasing downstream until reaching lower Bengal, where nearly 1,600 mm of rainfall occurs. Heavier rainfall continues in the upper Himalayan region, such as in Dehra Dun, where the rainfall is as high as 2,209 mm per annum. As mentioned above, most of the rainfall over the Ganges River Basin occurs during the southwest monsoon season. During this season, cyclonic disturbances are important synoptic systems that cause heavy spells of rainfall in the Ganges River Basin. On an average seven cyclonic disturbances (mainly depressions) form in the Bay of Bengal during the 4 months from June to September. These disturbances generally move in a west- northwest direction after their formation at the head of the Bay of Bengal up to the central Final Report: Volume I Confidential Page 32 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia parts of the country before weakening. It is well known that heavy rainfall occurs in the southwestern sector of the monsoon depressions due to strong convergence in that sector. The part of Ganges River Basin comes in the southwest of monsoon depressions tracks and as such heavy to very heavy rainfall occurs over different parts of the Ganges River Basin. The topography of the basin including Himalayas also plays an important role in causing heavy rainfall in the parts of the basin during the southwest monsoon season. When monsoon depressions are formed in the Bay of Bengal, the Arabian Sea currents are strengthened and cause heavy rainfall over the parts of the basin. The main synoptic situations of the southwest monsoon system that produce heavy rainfall over the Ganges River Basin are formation and subsequent movement of monsoon depressions, low- pressure systems from the head Bay of Bengal and well marked seasonal trough. On average, five to seven western disturbances move over the northwestern region of India during the winter months. An equal number of induced low pressure zones form to their south and move east and northeastward, giving heavy to very heavy rains over the plains of Punjab, Haryana, and Uttar Pradesh. Normally, the southwest monsoon starts withdrawing from the northwest and adjoining regions of India from 15 September; it gets revived again in association with Bay of Bengal/Arabian Sea depressions, which move toward this region and interact with western disturbances, causing exceptionally heavy rainfall over the region, especially in J&K, sub-Himalayan regions of Himachal Pradesh, Punjab, and Uttarakhand Himalayas. The tracks of monsoon depressions or low-pressure areas from the Bay of Bengal sometimes re-curve in a northerly to northeasterly direction when they travel over the central parts of the country. This is mostly due to the movement of strong westerlies (western disturbances) over the extreme northwest to northeast Himalayas. The synchronization of movement of westerly waves in the extreme north with the passage of monsoon disturbances in the lower latitudes cause heavy to very heavy rainfall along the foothills of the Himalayas (Nandargi and Dhar, 2012). 1.5 Rainfall pattern The mean annual rainfall of the Ganges River Basin is about 1,019 mm with a CV of 89%. About 91% of the annual rainfall is received during the southwest monsoon season of June to October. The bulk of the remaining 9% occurs mostly in the winter, pre-monsoon and post-monsoon periods. However, mean annual rainfall over the individual sub-basins varies widely due to orographic influences and preferential occurrence of rain producing systems in certain parts of the basin. The mean annual rainfall of a catchment is as low as 575 mm in the western part of the basin. Rainfall gradually increases towards the east. 1.6 Objectives of the study The main objectives of this study is to understand the geographical impacts of floods on various sectors (viz. residential, commercial, industrial, essential facilities, infrastructure, and agriculture), exposed in the countries falling in the Ganges Basin. This will enhance the knowledge base for better understanding the socio-economic impacts of flooding in the basin and subsequently help the stake holders/decision makers for better mitigation planning. In view of the regular recurrence of losses due to floods in different parts of the basin, a comprehensive study for the entire Ganges system was considered essential to understand flood risks in the region. Keeping this need in mind, the World Bank has initiated a River Basin Approach to develop a shared knowledge base and analytical framework for flood risk assessment and has contracted RMSI as consultant to complete this assignment. Final Report: Volume I Confidential Page 33 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 1.7 Scope of the study Key activities include: 1. Development of exposure (assets at risk) database for various classes such as buildings, infrastructure, critical and essential facilities, demography, and agriculture 2. Development of one dimensional steady flow probabilistic flood hazard model for 2, 5, 10, 25, 50, and 100-years return periods, which includes:  Collection and compilation of existing data pertaining to hydro-meteorological parameters from state run meteorological bureaus and water resources departments for river flow data from respective countries available in public domain  Development of hydraulic models for the entire Ganges basin up to its confluence with the Brahmaputra River in Bangladesh for probabilistic flood hazard assessment 3. Vulnerability and risk/loss estimation  Development of vulnerability (damage function) curves for each of the exposure (assets) classes  Estimation of direct losses to derive various risk metrics (i.e. AAL or Average Annual Loss and loss exceedance curve) at basin, sub-basin, country, state, district and block (sub-district) level 4. Development of an open-source web GIS based Flood Risk Atlas to view the results of this study over the web 1.8 About the fourth deliverable The fourth deliverable of the Flood Risk Assessment for the Ganges Basin in South Asia is presented in this report, viz., Final Report. The report includes: ï‚· Exposure data compilation ï‚· Model development results, assumptions, and challenges ï‚· Initial results of hazard assessment ï‚· Risk assessment results ï‚· Priority areas, conclusions, and recommendations The chapter wise synopsis of the report is as follows: ï‚· Chapter 1: (present chapter) provides the reader with an overview of the project, its scope, and its objectives. It also provides the reader with an understanding of key concepts related to the project. ï‚· Chapter 2: describes exposure data collection and management, including the following:  The data sources and the methodology for exposure data management  Exposure data development for demographic data, general building stocks, essential facilities, and transportation systems and discusses their valuation. ï‚· Chapter 3: provides details of the Flood Hazard Assessment, including the following:  Provides details on data availability required for flood assessment  Provides initial results of the hydraulic model, development, calibration, and validation  Flood Hazard Mapping for the probabilistic events of 2, 5, 10, 25, 50, and 100-years ï‚· Chapter 4: provides details of the risk assessment methodology. This includes:  Methodology for vulnerability functions (depth-damage function) development for various exposure classes viz. Buildings, Infrastructure, and Agriculture etc.  Methodology for loss calculation for various types of exposures Final Report: Volume I Confidential Page 34 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia ï‚· Chapter 5: provides details on the findings of the risk assessment task at basin-level. This includes:  Details of affected demography (Male, female, children, Scheduled Castes, Scheduled Tribes, etc.) due to various return period flood events of 2, 5, 10, 25, 50, and 100-Years  Details of economic losses in terms of Average Annual Loss (AAL), Probable Maximum Loss (PML) and Loss Exceedance Curves due to various return period flood events of 2, 5, 10, 25, 50, and 100-Years. ï‚· Chapter 6: provides details on the findings of the risk assessment task at each of the 18 sub-basins of the Ganges Basin. This includes:  Details of affected population and economic losses due to various return period flood events of 2, 5, 10, 25, 50, and 100-Years  Details of losses at State level in the sub-basin  Details of loss and priority areas at sub-district (block) level ï‚· Chapter 7: provides the conclusions and recommendations based on the findings of the risk assessment study as a part of this assignment along with the areas to improve the study. Note: In addition to the Main Report, RMSI is also providing the Ganges Basin Risk Atlas – a compendium of the main risk maps at the basin and sub-basin levels. Final Report: Volume I Confidential Page 35 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 2 Exposure Data Collection and Management Exposure is a function of the geographic location of the elements at risk. It is expressed in terms of the number of human lives and the value of properties or assets that can potentially be affected by the hazard. Developing exposure data is a critical component of any risk assessment study. Exposure data constitutes population, the built environment, systems that support infrastructure and livelihood functions, or other elements present in the hazard zones, which are subjected to potential losses. Modeling vulnerability of a system to natural hazards involves establishing a relationship between the potential damageability of critical exposure elements and different levels of local hazard intensity for the hazard. Damage susceptibility associated with a given level of hazard is measured in terms of a Mean Damage Ratio (MDR), defined as the expected proportion of the monetary value of repairs needed to bring back the facility to pre-event condition, over the replacement value of the facility, as a consequence of the hazard. This chapter details the types of exposure data collected and their analysis for the geographical extent of the Ganges Basin occupying India, Nepal, Bangladesh, and China. Data pertaining to China was very sparse and hence details related to the country are not covered in this report. The following sub-sections present a detailed overview of the development of the exposure database of population, buildings, infrastructure assets, and crops developed for the Ganges Basin. 2.1 Methodology for exposure data development The methodology for exposure database development in this study is based on a "bottom- upâ€? approach that includes classifying different types of buildings and infrastructure elements into different categories, estimating the count/dimension under each category, and combining those building/infrastructure counts/dimensions with per unit built-up floor area/cost of construction. Thus, the output of exposure data is the total monetary value by asset category, called replacement cost. 2.1.1 DATA COLLECTION The team emphasized on acquiring latest high resolution data sets for the present study. Main steps in the data collection process include: 1. Preparing a comprehensive list of data required 2. Acquiring the required data 3. Creating a data inventory sheet 4. Identifying data gaps and alternate sources 2.1.2 DATA PROCESSING AND DATA GAPS The collected data was reviewed, assessed, and maintained in data inventory sheets along with information regarding the vintage, source, resolution, and other feature attributes. While the data, without any gaps, was considered for processing, appropriate sources/methods were identified to fill any gaps for the remaining data sets. To process the collected data, the RMSI team carried out an initial analysis of collected data and prepared a list of input data having vintage, source, resolution, and attribute information. After initial analysis, data was processed to bring them in usable formats following standard quality assurance norms. There were separate quality checks for aggregate and site-specific data. RMSI took care of the following points during data processing: Final Report: Volume I Confidential Page 36 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia ï‚· Resolution ï‚· Projection System ï‚· Positional Accuracy ï‚· Naming Convention: Naming convention of all files are the same to maintain consistency in the names of all the files. ï‚· Creation of Unique Id: RMSI maintained a unique id for each data set, which helped to analyze the data at any stage with respect to the initial one. ï‚· Standard Unit in all Dataset: Unit for each entity was decided to have consistency in the entire data set. ï‚· Essential Fields: RMSI maintained a few common fields in all the data sets to help link them. ï‚· Logical Checks: RMSI applied required logical checks to verify data consistency ï‚· Meta data: The team created metadata considering the various data limitations and the many assumptions RMSI makes while processing data. All the information stored in Meta data will help the World Bank and other stakeholders understand the data more clearly. After reviewing data and gaps, the exposure elements without any data gaps were processed in the required formats after applying due quality checks. Such datasets are used to estimate the exposure value by multiplying the total area/length/count with unit replacement costs. Mostly, it has been observed that once the data is processed, gaps have been identified and filled primarily in the areas of: 1. Missing Location Information: sources such as Google Earth/ Google Maps (after due verification) and data from past studies were used to fill these types of data gaps. It is not necessary that for every exposure element, location coordinates (latitude and longitude) are available. However, the exposure element in question must be associated with some administrative unit for analysis. 2. Missing structure or classification information: when detailed data associated with structural types are not available, the team used databases available from past studies and literature on the subject. 3. Missing replacement cost: the team relied on the literature surveys to extract and compile useful information on unit exposure values for features whose replacement costs could not be obtained 4. No data available/obtained: in such cases, data gaps are filled through use of appropriate literature and/ or GIS data available from open sources. The data outputs, after bridging the gaps, served as inputs for further processing (count estimation, area/length estimation, unit cost estimation) and finally for the total exposure values of different features in the study area. 2.1.3 DATA QUALITY CHECKS AND QUALITY ASSURANCES Data quality is an important aspect and is best achieved by preventing errors throughout the project life cycle. The Quality control system comprises of a set of routine technical activities, to measure and control the quality of the data divided into specific quality control stages. Therefore, the objective of the Quality Control (QC) and Quality Assurance (QA) processes is to ensure that the data is accurately captured, recorded, and saved. The data QA/QC focus is mainly concentrated in the areas of: 1. Resolution of data 2. Projection system 3. Positional accuracy 4. Topological error of the data 5. Creation of Unique Id (UID) 6. Essential fields Final Report: Volume I Confidential Page 37 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 7. Standard unit in all dataset 8. File naming convention and versioning 2.1.4 DEVELOPMENT OF EXPOSURE DATA AND GENERATION OF GIS DATA LAYERS Exposure data are categorized into ‘aggregate’ or ‘site specific’, to analyze the impact of hazards. Aggregate data are those where area and count are summed at a suitable administrative unit level, for instance here at State/Province/District-level, while the site- specific data are presented with specific geographic coordinates. A general rule for categorizing data as aggregated or site specific is based on the level at which the location information is available. For example, socio-economic/ demographic data (population by age groups, economically weaker sections etc.) are generally presented at an aggregated level. Depending on the format of the data received, various data processing steps are applied, that include cleaning, data standardization, geo-referencing, data development etc. The positional accuracy and completeness of the data is also taken care of in this activity. Structural details for buildings and their occupancies are considered as important elements of exposure in risk assessment studies. The Census 2011 building structural details (wall and roof material) are grouped into structural classes and joined with the respective polygon shape file. After compilation of exposure data, the exposure values are estimated based on the material of construction for the compiled data sets by applying per unit replacement costs. The unit replacement costs for various exposure elements are finalized using several sources (i.e., Government reports, and other relevant literature). Similar steps have been carried out for infrastructure elements such as transportation features. These exposure elements are site-specific and are presented as separate GIS layers in this study. For transportation features like roads and railways, the data has been processed as line/poly-line features with compiled attribute information attached to them. Based on built up area, average number of persons residing at each building is estimated. Such information is useful for estimating population affected due to the hazard. The other demographic and socio-economic data like population (age, sex, density, literacy etc.), crop details etc. were aggregated after required processing and analysis. After processing of tabular data, these data sets with metadata information were attached with corresponding polygons for risk assessment. 2.2 Analysis of exposure elements Detailed analysis of each exposure data such as demography, agriculture, housing, occupancy, and infrastructure that have been collected is presented in the following subsections. Each element has been described separately for the respective countries. 2.2.1 DEMOGRAPHIC ANALYSIS The source of demographic and household data is the Census report of the respective country. These reports provide population distribution by age, gender, education, and occupation. India In India, the total population of the study area is 441,469,944 (approx. 441.5 million) spread over 11 states. The data is presented at state level. Figure 2-1 presents the state level total population distribution in India. Uttar Pradesh state is the most populated and Himachal Pradesh is the least. Final Report: Volume I Confidential Page 38 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 2 - 1 : S t a t e l e ve l p o p u l a t i o n d i s t r i b u t i o n f o r t h e b a s i n i n I n d i a Amongst the population, 52% are males and 48% are females. The female population is similarly distributed in all the states and ranges from 46% to 49% of the total population of each state’s area lying within the Ganges Basin. Figure 2-2 presents the overall gender distribution of population. Figure 2-3 shows the state-level gender distribution. Figure 2-2: Gender Distribution for the F i g u r e 2 - 3 : S t a t e l e ve l g e n d e r d i s t r i b u t i o n basin in India for the basin in India. Figure 2-4 presents the literacy rate in India. About 57% of the total population is literate and 43% of the total population is illiterate. Delhi has the highest literacy rate and Jharkhand has the lowest as can be seen from Figure 2-5. It can also be inferred that literacy rates are higher in states in the western, northwestern, and central parts of the basin. Final Report: Volume I Confidential Page 39 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Figure 2-4: Literac y rate for Figure 2-5: State wise literac y distribution for the the basin in India basin in India Nepal The total population of the study area in Nepal is 26,253,828 (26.25 million) spread over five provinces. Figure 2-6 shows the province level populations. Central province has the most population and Far Western province has the least. Figure 2-6: Province level population distribution for the basin in Nepal The overall gender distribution for the basin in Nepal is exactly the reverse that of India. It is skewed towards females with 48% comprising males and 52% females. Figure 2-7 presents the overall gender ratio for the basin in Nepal. Final Report: Volume I Confidential Page 40 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 2 - 7 : O ve r a l l g e n d e r F i g u r e 2 - 8 : P r o v i n c e l e ve l g e n d e r d i s t r i b u t i o n f o r t h e ratio for the basin in Nepal basin in Nepal The province-level gender distribution for the basin is significantly skewed in four of the five provinces and is highest (about 54% females) in Western province. Figure 2-8 presents the gender distribution for the basin in Nepal. Only the Central province has a more equitable gender ratio. Bangladesh The total population of the study area in Bangladesh is 6,364,813 (about 6.3 Million) spread over 10 Districts. Figure 2-9 illustrates the district level distribution of the total population for the basin in Bangladesh. It can be seen that Nawabgunj has the maximum population amongst districts and Meherpur has the least. Figure 2-9: District le vel distribution of total population for the basin in Bangladesh Final Report: Volume I Confidential Page 41 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 2 - 1 0 : O ve r a l l g e n d e r r a t i o F i g u r e 2 - 1 1 : D i s t r i c t l e ve l d i s t r i b u t i o n o f for the basin in Bangladesh gender for the basin in Bangladesh Figure 2-10 presents the overall gender distribution for the basin in Bangladesh, which is equally balanced between males and females (both 50%). District wise gender distribution is also shown in Figure 2-11. Summarizing the above information, Figure 2-12 shows the distribution of population for the sub districts/districts of India, Nepal and Bangladesh falling within the Basin. Significantly, districts from Delhi, Haryana, and Uttar Pradesh falling within the basin have higher total populations. On the contrary, the mountainous districts of Himachal Pradesh, Uttarakhand, and Nepal and districts in the southern part of the basin have lower total populations. Final Report: Volume I Confidential Page 42 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Figure 2-12: Population distribution in Ganges Basin 2.2.2 HOUSING In the present study, housing classification was based on occupancy types and structure types. The occupancy-based classification differentiates housing into different categories viz. residential, commercial, industrial establishments, health facilities, educational institutes, and others. The final class denoted as others, comprises of places of worship and houses kept locked. Data was collected from census at sub district level for the total number of houses for residential occupancy which includes residential and residential cum others category while for the rest of the occupancy types data was available only at district level. This was divided into the above-mentioned classes at sub district level using sub district area weightage. The sub district area weightage is a function of the geographical area occupied by each sub district within a district. Further, housing has been also classified into occupied or vacant classes. The following subsections describe the distribution of these classifications for the portion of Ganges Basin occupying India, Nepal, and Bangladesh. 2.2.2.1 Overall Housing Occupancy India The occupancy distribution of housing shows that 94% of houses are occupied and 6 % are vacant for the basin in India as shown in Figure 2-13. Final Report: Volume I Confidential Page 43 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Figure 2-13: Housing distribution in India Figure 2-14: Occupancy of houses in India Figure 2-14 shows that majority of the houses (78%) for the basin in India are residential, about 19% are commercial, and rest of the types Education Institute, Health facilities, industrial establishments, and Others are either 1% or less than 1% approximately. Nepal The total number of households for the basin in Nepal is 5,400,000, which are spread over the basin in the five provinces. Central province has the maximum number and Far Western has the least number of households as can be seen in Figure 2-15. Figure 2-15: Province wise distribution of Figure 2-16: Housing occupanc y for the households for the basin in Nepal basin in Nepal Figure 2-16 illustrates the distribution based on occupancy. Here, as in India, the maximum percentage of houses (79%) are used for residential purposes while a sizable percentage (18%) is used for commercial purposes. Bangladesh The total number of households for the basin in Bangladesh is 1,483,686, which are spread over 10 districts. Nawabganj has the maximum number of households (24%) and Meherpur has the least (1%) as can be seen in Figure 2-17. Final Report: Volume I Confidential Page 44 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 2 - 1 7 : D i s t r i c t l e ve l d i s t r i b u t i o n o f Figure 2-18: Housing households for the basin in Bangladesh distribution for the basin in Bangladesh Figure 2-18 illustrates the distribution based on occupancy. Almost identical to Nepal, the maximum percentage(79%) of houses for the basin in Bangladesh are used for residential purposes while a sizable percentage (18%) is used for commercial purposes. From the above information a map was produced for the Ganges Basin, Figure 2-19 shows the distribution of total houses for the study area covering India, Nepal and Bangladesh. It can be seen that the maximum number of houses can be found in parts of the basin lying in Uttar Pradesh, Delhi, and Rajasthan, and in Central province of Nepal. Figure 2-19: Distribution of total houses in the Ganges Basin Final Report: Volume I Confidential Page 45 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 2.2.2.2 Residential houses India The Census of India (2011) provides the number of residential and residence-cum-other use houses. Both categories of data have been combined together to get the total number of residential houses for the basin in India. Figure 2-20 shows the state level distribution of residential houses in India. Uttar Pradesh state has the highest number of residential houses and Himachal Pradesh the least. F i g u r e 2 - 2 0 : S t a t e l e ve l d i s t r i b u t i o n o f r e s i d e n t i a l h o u s e s f o r t h e b a s i n i n I n d i a Illustrations of a few types of residential houses in India are shown in Figure 2-21. Final Report: Volume I Confidential Page 46 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Huts House Apartment Villa F i g u r e 2 - 2 1 : T yp e s o f r e s i d e n t i a l h o u s e s i n I n d i a Nepal The province level distribution of residential houses for the basin in Nepal is illustrated in Figure 2-22. It shows that the basin in Central province has the highest number of residential houses and Far Western has the least number. Final Report: Volume I Confidential Page 47 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 2 - 2 2 : P r o v i n c e l e ve l d i s t r i b u t i o n o f r e s i d e n t i a l h o u s e s f o r t h e b a s i n i n N e p a l Bangladesh The district level distribution of residential houses is given in Figure 2-23. It shows that the maximum number of residential houses can be found in parts of the basin lying in Nawabgunj district and the least in Meherpur district. F i g u r e 2 - 2 3 : D i s t r i c t l e ve l d i s t r i b u t i o n o f r e s i d e n t i a l h o u s e s f o r t h e b a s i n i n Bangladesh Summarizing the above information, a map was produced for the Ganges Basin. Figure 2-24 shows the distribution of residential houses for the entire basin. Final Report: Volume I Confidential Page 48 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Figure 2-24: Distribution of residential houses in Ganges Basin 2.2.2.3 Commercial houses India The Census provides the number of shops and offices, hotels, lodges, and guest houses at district level. All these categories of data have been combined together to get the total number of commercial houses for the basin in India. Figure 2-25 shows that the basin lying in Uttar Pradesh has the maximum number of commercial houses and that lying in Chhattisgarh has the least. Figure 2-26 displays a few images of the types of commercial houses in India. Final Report: Volume I Confidential Page 49 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 2 - 2 5 : S t a t e l e ve l d i s t r i b u t i o n o f c o m m e r c i a l h o u s e s f o r t h e b a s i n i n I n d i a F i g u r e 2 - 2 6 : D i f f e r e n t t yp e s o f c o m m e r c i a l h o u s e s i n I n d i a Final Report: Volume I Confidential Page 50 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Nepal The province level distribution of commercial houses for the basin in Nepal is shown in Figure 2-27. Central province has the most number of commercial houses and Far Western has the least. F i g u r e 2 - 2 7 : P r o v i n c e l e ve l d i s t r i b u t i o n o f c o m m e r c i a l h o u s e s f o r t h e b a s i n i n N e p a l Bangladesh Figure 2-28 shows the district level distribution of commercial houses for the basin in Bangladesh. We see that the basin lying in Nawabgunj has the maximum commercial houses and that lying in Meherpur has the least. F i g u r e 2 - 2 8 : D i s t r i c t l e ve l d i s t r i b u t i o n o f c o m m e r c i a l h o u s e s f o r t h e b a s i n i n Bangladesh Summarizing the above information, a map was produced for the Ganges Basin. Figure 2-29 shows the distribution of commercial houses for the study area. Central and Northern India along with Central Province of Nepal have more commercial houses as compared to other parts of the Ganges Basin. Final Report: Volume I Confidential Page 51 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Figure 2-29: Distribution of commercial houses for the Ganges Basin 2.2.2.4 Industrial Establishments India The acquired data for industrial establishments has been analyzed based on numbers available from the demographic data at state and district levels. Figure 2-30 shows that Uttar Pradesh state has the maximum number of industrial establishments and Chhattisgarh has the least. Figure 2-31 illustrates a few types of industrial establishments in India. Final Report: Volume I Confidential Page 52 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 2 - 3 0 : S t a t e l e ve l d i s t r i b u t i o n o f i n d u s t r i a l e s t a b l i s h m e n t s f o r t h e b a s i n i n India Figure 2-31: Examples of industrial establishments in India Nepal Figure 2-32 shows the distribution of industrial establishments for the basin in Nepal. Central province has the most number of industrial establishments in comparison to Far Western, which has the least. Final Report: Volume I Confidential Page 53 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 2 - 3 2 : P r o v i n c e l e ve l d i s t r i b u t i o n o f i n d u s t r i a l e s t a b l i s h m e n t s f o r t h e b a s i n i n Nepal Bangladesh Figure 2-33 illustrates the district level distribution of industrial establishments for the basin in Bangladesh. Nawabgunj has the maximum number of industrial establishments and Meherpur has the least. F i g u r e 2 - 3 3 : D i s t r i c t l e ve l d i s t r i b u t i o n o f i n d u s t r i a l e s t a b l i s h m e n t s f o r t h e b a s i n i n Bangladesh Summarizing the above information, a map was produced for the Ganges Basin. Figure 2-34 shows the distribution of industrial establishments for the study area. Delhi and the adjoining areas have the most industrial establishments. This is followed by parts of Haryana, Rajasthan, Uttar Pradesh, and Madhya Pradesh in India, along with Central province in Nepal and Dinajpur in Bangladesh, which have significant concentrations of industrial establishments as compared to the other parts of the Ganges Basin. Final Report: Volume I Confidential Page 54 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Figure 2-34: Distribution of industrial e stablishments for Ganges Basin 2.2.2.5 Educational Institutions India Educational institutions play a critical role in mitigation and recovery operations during and after disasters as these are generally used as shelters. Census of India provides district- level data for the number of schools of 2011 vintage. The acquired data was analyzed based on the household numbers available in the demographic data at district levels. The area lying in the basin for the state of Uttar Pradesh has the highest number of educational institutions and Himachal Pradesh has the least as seen in Figure 2-35. Final Report: Volume I Confidential Page 55 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 2 - 3 5 : S t a t e l e ve l d i s t r i b u t i o n o f e d u c a t i o n a l i n s t i t u t i o n s f o r t h e G a n g e s B a s i n in India Nepal Figure 2-36 shows the province level distribution of educational institutions in Nepal. The figure indicates that areas of the Central province lying in the basin have the most educational institutions as compared to the Far Western province, which has the least. F i g u r e 2 - 3 6 : P r o v i n c e l e ve l d i s t r i b u t i o n o f e d u c a t i o n a l i n s t i t u t i o n s i n N e p a l . Bangladesh Figure 2-37 shows the district level distribution of educational institutes in Bangladesh. The figure indicates that areas of Nawabgunj district lying in the Ganges Basin have the maximum number of educational institutes as compared to Meherpur, which has the least. Final Report: Volume I Confidential Page 56 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 2 - 3 7 : D i s t r i c t l e ve l d i s t r i b u t i o n o f e d u c a t i o n a l i n s t i t u t i o n s f o r t h e b a s i n i n Bangladesh Summarizing the above information, a map was produced for the Ganges Basin. Figure 2-38 shows the distribution for educational institutions for the study area. Uttar Pradesh, Delhi, and Madhya Pradesh appear to have more educational institutions as compared to other parts of the Ganges Basin. Figure 2-38: Distribution of educational institutions across th e Ganges Basin Final Report: Volume I Confidential Page 57 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 2.2.2.6 Health Facilities India Health facilities have been included under essential facilities. These facilities play a critical role in mitigation and recovery operations during and after disasters. The acquired data for health facilities was analyzed based on the household numbers available in the demographic data at district level. The areas of the state of Uttar Pradesh lying within the basin have the most health facilities and Chhattisgarh has the least as seen in Figure 2-39. F i g u r e 2 - 3 9 : S t a t e l e ve l d i s t r i b u t i o n o f h e a l t h f a c i l i t i e s f o r t h e b a s i n i n I n d i a Nepal Figure 2-40 shows the province level distribution of health facilities for the basin in Nepal. The Figure indicates that the maximum number of health facilities is in the Central province and the Far Western province has the least. F i g u r e 2 - 4 0 : P r o v i n c e l e ve l d i s t r i b u t i o n o f h e a l t h f a c i l i t i e s i n N e p a l Final Report: Volume I Confidential Page 58 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Bangladesh Figure 2-41 shows the district level distribution of health facilities for the basin in Bangladesh. It can be concluded that areas of Nawabganj lying in the basin have the most health facilities and Meherpur has the least. F i g u r e 2 - 4 1 : D i s t r i c t l e ve l d i s t r i b u t i o n o f h e a l t h f a c i l i t i e s f o r t h e b a s i n i n B a n g l a d e s h Summarizing the above information, a map was produced for the Ganges Basin. Figure 2-42 shows the distribution of health facilities for the study area. The map suggests that Madhya Pradesh, Delhi and parts of Uttar Pradesh along with Central province in Nepal have more health facilities as compared to other parts of the Ganges Basin. Final Report: Volume I Confidential Page 59 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Figure 2-42: Distribution for health facilities for Ganges Basin 2.2.2.7 Other houses India Places of Worship and houses kept locked have been classified as other houses for this study. The following Figure shows the state level distribution of these houses in India. Figure 2-43 shows that Uttar Pradesh has the maximum number of other houses and Chhattisgarh has the least. Final Report: Volume I Confidential Page 60 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 2 - 4 3 : S t a t e l e ve l d i s t r i b u t i o n o f o t h e r h o u s e s i n I n d i a Nepal Figure 2-44 illustrates the province level distribution of other houses in Nepal. It can be seen that the maximum number of other houses is in Central province and the least is in Far Western. F i g u r e 2 - 4 4 : P r o v i n c e l e ve l d i s t r i b u t i o n o f o t h e r h o u s e s i n N e p a l Bangladesh The district level distribution of other houses in Bangladesh is shown in Figure 2-45. We see that Nawabgunj has the maximum number of houses classified as other houses and Meherpur has the least. Final Report: Volume I Confidential Page 61 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 2 - 4 5 : D i s t r i c t l e ve l d i s t r i b u t i o n o f o t h e r h o u s e s i n B a n g l a d e s h Summarizing the above information a map was produced for the Ganges Basin. Figure 2-46 shows the distribution for other houses (places of worship and houses kept locked) for the study area. Other houses have a greater concentration in Delhi, Rajasthan, Madhya Pradesh, and Uttar Pradesh along with Central province in Nepal, as compared to the rest of the Ganges Basin. Figure 2-46: Distribution for other houses for Ganges Basin Final Report: Volume I Confidential Page 62 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 2.2.2.8 Housing Structural Types Structure based classification further categorizes the houses based on the construction material used. The structural classes were analyzed using block level data of occupancy. There are seven identified classes. From the vulnerability and risk to hazard perspective, structural classification of housing is a critical component of exposure data development. The vulnerability of houses to various hazards depends largely on their construction materials, structural types and heights, which have been categorized into different structural-types, based on their characteristics. The different categorizations represent elements that are distinctly vulnerable to the same level of hazard. The construction properties vary from country to country among different income groups. The broad structural types classification used for this study, based on the construction material, have been given in the Table 2-1. T a b l e 2 - 1 : H o u s i n g c a t e g o r i e s b y c o n s t r u c t i o n m a t e r i a l s a n d s t r u c t u r a l t yp e s Structural Composition of housing material Sub class Structural 1 Grass, Thatch, Bamboo, Wood, Mud, Plastic, etc. Structural 2 Mud/Unburnt Brick/Stone without mortar Structural 3 Light Metal Burnt Brick/ stone with mortar having temporary Roof (Tiles, wood, GI, Structural 4 slate, etc.) Structural 5 Masonry building /Reinforced concrete frame with brick infill Structural 6 RCC Structural 7 Any Other Residential houses have been classified into all the seven structural classes based on the roof and wall construction material. Usually the education institutes and health facility consist of all structural types. However, as there was no detailed information available of the construction material for educational institutes and health facility, the distribution ratio of residential houses have been applied on these two occupancy types to get the structural type distribution. Thus, educational institutes and health facilities have also been subdivided into these seven classes. Usually the commercial and industrial buildings are Reinforces Concrete Structures (RCC). Therefore, the commercial, industrial, and other houses have been assigned “Structural 6â€? – type classification as observed in Table 2-1. India In India, all seven structural types of houses as shown in Table 2-1 can be found. The “Structural 4â€? – type, classification (combination of burnt brick/ stone with mortar having temporary roof tiles, wood, GI, slate, etc.) is most common in India, and “Structural 3’ – type classification (combination of light metal) is least used for housing, around 0.1%, as can be seen from Figure 2-47. Final Report: Volume I Confidential Page 63 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Figure 2-47: Distribution of structural classes used in India for housing Figure 2-48: Distribution of structural Figure 2-49: Distribution of structural classes used in Nepal for housing classes used in Bangladesh for housing Nepal In Nepal, five types of structural classes are used as described in available literature. These are five of the seven classes as shown in Table 2-1. The most common material used is “Structural 1â€? - type classification (combination of grass, thatch, bamboo, wood, mud, plastic, etc.). Figure 2-48 shows the distribution of structural classes used for construction of houses in Nepal. Bangladesh In Bangladesh, four types of structural classes are used for construction, as available literature explains. These are four of the seven classes shown in Table 2-1. Figure 2-49 shows the distribution of the same. “Structural 2â€? - type classification is the most used, which is a combination of mud/un-burnt brick/stone without mortar. 2.2.3 INFRASTRUCTURE DATA In the present study, the inventory of infrastructure assets, comprising of the locations, types, measurement, replacement costs etc., were collated after necessary data processing and quality checks. Though the infrastructure database is sparse, it contains comprehensive details of major infrastructure facilities with a higher level of geo-locational details. It should Final Report: Volume I Confidential Page 64 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia be noted that the analysis presented below is for areas of respective countries falling within the study area. During disasters, the transportation network plays an important role in rescue and recovery operations in a district. The roads and railways data compiled are considered for exposure analysis in the sections presented below. 2.2.3.1 Road Network Road data constitutes of important attributes like types of roads, lengths, administrative areas and replacement costs. India Roads cover about 81,172 km in 11 states of India lying in the basin. Figure 2-50 displays road lengths in each state. Uttar Pradesh has the maximum length of roads among the states lying within the basin in India. F i g u r e 2 - 5 0 : S t a t e - l e ve l d i s t r i b u t i o n o f r o a d l e n g t h s f o r t h e b a s i n i n I n d i a Nepal Total road length for the basin in Nepal is 11,092 km. Figure 2-51 displays province level road lengths. F i g u r e 2 - 5 1 : P r o v i n c e l e ve l d i s t r i b u t i o n o f r o a d l e n g t h s f o r t h e b a s i n i n N e p a l Final Report: Volume I Confidential Page 65 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Bangladesh Roads cover about 773 km of the study area in Bangladesh. The district level distribution of road lengths is shown in Figure 2-52. F i g u r e 2 - 5 2 : D i s t r i c t l e ve l d i s t r i b u t i o n o f r o a d l e n g t h s f o r t h e b a s i n i n B a n g a d e s h The road network for the whole Ganges basin is mapped for the study. Figure 2-53 presents the road network for the whole Ganges Basin. Figure 2-53: Road network for the Ganges Basin Final Report: Volume I Confidential Page 66 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 2.2.3.2 Rail Network The railway data constitutes of important attributes like types of rail line, length, administrative area and replacement costs. India The total length of rail tracks for the Ganges basin in India is about 19,432 km spread across 11 states. Figure 2-54 presents the state level distribution of rail network for the Ganges Basin in India. F i g u r e 2 - 5 4 : S t a t e l e ve l d i s t r i b u t i o n o f r a i l w a y t r a c k s f o r t h e G a n g e s b a s i n i n I n d i a Nepal The total length of rail tracks is about 152 km for the basin in Nepal. Figure 2-55 shows the province level distribution of length of railway track lengths in Nepal. F i g u r e 2 - 5 5 : P r o v i n c e l e ve l d i s t r i b u t i o n o f r a i l w a y t r a c k l e n g t h s f o r t h e b a s i n i n N e p a l Bangladesh Rail tracks cover about 225 km of the basin in Bangladesh. The district level distribution is shown in Figure 2-56. Final Report: Volume I Confidential Page 67 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 2 - 5 6 : D i s t r i c t l e ve l d i s t r i b u t i o n o f r a i l w a y t r a c k l e n g t h s f o r t h e b a s i n i n Bangadesh The railway network for the whole Ganges basin is mapped for the study. Figure 2-57 presents the railway network for the whole Ganges basin. Figure 2-57: Rail network for the stud y area Final Report: Volume I Confidential Page 68 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 2.2.4 AGRICULTURAL DATA Exposure data for agriculture is given by sown area in hectares and production in tons. The major crops Rice, Maize, and Wheat have been considered in this study for each country depending on data availability. India Figure 2-58 shows the state level distribution of area under Rice cultivation for the year 2010/11. The total area sown was 11,920,356 (11.92 million ha) hectares. F i g u r e 2 - 5 8 : S t a t e l e ve l d i s t r i b u t i o n o f a r e a u n d e r R i c e c u l t i va t i o n f o r t h e b a s i n i n India (2010/11) Figure 2-59 shows the state level distribution of Rice production for the basin in India (2010/11). The production in the basin was 16,486,402 tons (approx. 16.5 m tons). F i g u r e 2 - 5 9 : S t a t e l e ve l d i s t r i b u t i o n o f R i c e p r o d u c t i o n f o r t h e b a s i n i n I n d i a (2010/11) Figure 2-60 shows the state level distribution of area under Wheat cultivation for the year 2010/11. The total area sown was 17,925,327 (17.92 million ha) hectares. Final Report: Volume I Confidential Page 69 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 2 - 6 0 : S t a t e l e ve l d i s t r i b u t i o n o f a r e a u n d e r W h e a t c u l t i va t i o n f o r t h e b a s i n i n India (2010/11) Figure 2-61 shows the state level distribution of Wheat production for the basin in India (2010/11). The production in the basin was 50,843,155 tons (approx. 51 m tons). F i g u r e 2 - 6 1 : S t a t e l e ve l d i s t r i b u t i o n o f W h e a t p r o d u c t i o n f o r t h e b a s i n i n I n d i a (2010/11) Figure 2-62 shows the state level distribution of area under Maize cultivation for the year 2010/11. The total area sown was 2,718,152 (2.71 million ha) hectares. Final Report: Volume I Confidential Page 70 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 2 - 6 2 : S t a t e l e ve l d i s t r i b u t i o n o f a r e a u n d e r M a i z e c u l t i va t i o n f o r t h e b a s i n i n India (2010/11) Figure 2-63 shows the state level distribution of Maize production for the basin in India (2010/11). The production in the basin was 4,936,936 tons (approx. 5 m tons). F i g u r e 2 - 6 3 : S t a t e l e ve l d i s t r i b u t i o n o f M a i z e p r o d u c t i o n f o r t h e b a s i n i n I n d i a (2010/11) Nepal Figure 2-64 shows the province level distribution of area under Rice cultivation for the basin in Nepal (2012/13). The total area sown was 1,420,570 (1.42 million ha) hectares. Final Report: Volume I Confidential Page 71 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 2 - 6 4 : P r o v i n c e l e ve l d i s t r i b u t i o n o f a r e a u n d e r R i c e c u l t i va t i o n f o r t h e b a s i n i n Nepal (2012/13) Figure 2-65 shows the province level distribution of Rice production for the basin in Nepal for the year 2012/13. The production in the basin was 4,504,503 (approx. 4.5 m) tons. F i g u r e 2 - 6 5 : P r o v i n c e l e ve l d i s t r i b u t i o n o f R i c e p r o d u c t i o n f o r t h e b a s i n i n N e p a l (2012/13) Figure 2-66 shows the province level distribution of area under Wheat cultivation for the basin in Nepal (2012/13). The total area sown was 759,843 (approx. 0.76 million ha) hectares. Final Report: Volume I Confidential Page 72 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 2 - 6 6 : P r o v i n c e l e ve l d i s t r i b u t i o n o f a r e a u n d e r W h e a t c u l t i va t i o n f o r t h e b a s i n in Nepal (2012/13) Figure 2-67 shows the province level distribution of Wheat production for the basin in Nepal for the year 2012/13. The production in the basin was 1,882,220 (approx. 1.9 m) tons. F i g u r e 2 - 6 7 : P r o v i n c e l e ve l d i s t r i b u t i o n o f W h e a t p r o d u c t i o n f o r t h e b a s i n i n N e p a l (2012/13) Figure 2-68 shows the province level distribution of area under Maize cultivation for the basin in Nepal (2012/13). The total area sown was 849,635 (approx. 0.85 million ha) hectares. Final Report: Volume I Confidential Page 73 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 2 - 6 8 : P r o v i n c e l e ve l d i s t r i b u t i o n o f a r e a u n d e r M a i z e c u l t i va t i o n f o r t h e b a s i n in Nepal (2012/13) Figure 2-69 shows the province level distribution of Maize production for the basin in Nepal for the year 2012/13. The production in the basin was 1,999,010 (approx. 2.0 m) tons. F i g u r e 2 - 6 9 : P r o v i n c e l e ve l d i s t r i b u t i o n o f M a i z e p r o d u c t i o n f o r t h e b a s i n i n N e p a l (2012/13) Bangladesh Figure 2-70 shows the district level distribution of area under Rice cultivation for the basin in Bangladesh (2009/10). The total area sown was 669,920 (approx. 0.67 million ha) hectares. Final Report: Volume I Confidential Page 74 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 2 - 7 0 : D i s t r i c t l e ve l d i s t r i b u t i o n o f a r e a u n d e r R i c e c u l t i va t i o n f o r t h e b a s i n i n Bangladesh (2009/10) Figure 2-71 shows the district level distribution of Rice production for the basin in Bangladesh for the year 2009/10. The production in the basin was 1,996,173 (approx. 2.0 m) tons. F i g u r e 2 - 7 1 : D i s t r i c t l e ve l d i s t r i b u t i o n o f R i c e p r o d u c t i o n f o r t h e b a s i n i n B a n g l a d e s h (2009/10) Figure 2-72 shows the district level distribution of area under Wheat cultivation for the basin in Bangladesh (2009/10). The total area sown was 86,502 hectares. Final Report: Volume I Confidential Page 75 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 2 - 7 2 : D i s t r i c t l e ve l d i s t r i b u t i o n o f a r e a u n d e r W h e a t c u l t i va t i o n f o r t h e b a s i n in Bangladesh (2009/10) Figure 2-73 shows the district level distribution of Wheat production for the basin in Bangladesh for the year 2009/10. The production in the basin was 227,762 (approx. 0.23 m) tons. F i g u r e 2 - 7 3 : D i s t r i c t l e ve l d i s t r i b u t i o n o f W h e a t p r o d u c t i o n f o r t h e b a s i n i n Bangladesh (2009/10) Final Report: Volume I Confidential Page 76 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 2.3 Estimation of exposure values 2.3.1 ESTIMATION OF BUILT-UP AREA Literature survey, experience from field surveys conducted in the study area from previous projects and feedback from local consultants in other projects, helped in the determination of different structural types, average built up areas and unit costs for estimation of exposure values in this report. Residential, commercial and industrial The source of housing data is the Census of India (2011), which provides data based on the housing usage and construction materials. Further, field surveys conducted by RMSI and information gathered from local builders/real-estate websites was used for filling the data gaps and validation of information. The following Table 2-2 illustrates projects wherein field surveys were undertaken by RMSI in the study area. Table 2-2: Projects involving field work in Stud y Area Sr. Area Name of project Client Duration Remarks/Details No. visited Operational Research Villages in House hold survey, Focus to Support Asia Budi Two group discussion with 1. Mainstreaming of Development Gandak months communities and stake Integrated Flood Bank basin holders consultations Management Villages in Flood Modeling of The World Seven Building Exposure survey 2. Kosi river Ganges basin in India Bank USA days and data collection basin The field visits exemplify the various usages (occupancy types) and construction materials used in housing types. The average areas observed during these visits are given in Table 2-3. As mentioned previously, residential buildings can be of different classes, grass, mud, light metal, burnt brick, masonry and RCC. Usually the commercial and industrial buildings are Reinforces Concrete Structures (RCC). Therefore, the commercial, industrial, and other houses have been assigned RCC (Structural type 6) classification. T a b l e 2 - 3 : A r e a s o f va r i o u s h o u s i n g i n I n d i a Type of Type of Housing residence Avg Area in Sq Ft Kachha Grass/Thatch/Bamboo/Wood/Plastic etc house/hut 125 Mud/Unburnt Bricks Semi-kachha 200 Light Metal Hut 150 Burnt Brick Semi-kachha 250 Masonry Pakka 600 Reinforced Concrete Frame (RCF) Villa 1000 RCF Apartment 850 Reinforced Concrete Cement (RCC) Villa 1000 Final Report: Volume I Confidential Page 77 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Type of Type of Housing residence Avg Area in Sq Ft RCC Apartment 850 Other 200 The number of houses was multiplied with the average per unit built-up floor area of each structure type to estimate the total built-up area of houses. Average rates and price inputs were taken from property/real-estate websites relevant to the study area along with the local consultant/field expert’s feedback from previous projects/studies. The average unit replacement costs for India are given in Table 2-4. Several reports were referenced such as Catalogue of building typologies in India (NDMA, 2013), Industrial district profiles from Ministry of Micro and Small Enterprises (MSME) (2012), and State level Gazette information. For average unit costs pertinent to Nepal, the following were referenced: UN Habitat, Urban Housing Sector Profile (2010), The Federation of Nepalese Chambers of Commerce & Industry (FNCCI), and Nepal Property Market websites. The average costs have been represented in Table 2-5. Average unit costs for Bangladesh are provided in Table 2-6 and are supported by UN Habitat documentation on urban revitalization of mass housing in Dhaka, Bangladesh. Detailed references have been provided at the end of this report in addition to appropriate places in write-up. Final Report: Volume I Confidential Page 78 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 2 - 4 : U n i t r e p l a c e m e n t c o s t s f o r d i f f e r e n t t yp e s o f s t r u c t u r e s i n I n d i a Educational Residential Commercial Industrial Health Facility Other Inst. Avg. Area, Avg. Area, Avg. Area, Avg. Area, Avg. Area, Avg. Area, Unit Cost, Unit Cost, Unit Cost, Unit Cost, Unit Cost, Unit Cost, INR/Sq m INR/Sq m INR/Sq m INR/Sq m INR/Sq m INR/Sq m Description sq m sq m sq m sq m sq m sq m Grass, Thatch, Bamboo, Wood, Mud, Plastic, etc. 11.61 1,076 14.09 299 5.83 2,153 Mud/Unburnt Brick/Stone without mortar 18.58 2,153 22.55 598 39.91 916 Light Metal 13.94 1,615 16.91 448 29.93 687 Burnt Brick/ stone with mortar having temporary 23.23 5,382 28.19 1,495 49.90 2,290 Roof (Tiles, wood, GI, slate, etc.) Masonry building /Reinforced concrete frame with 75.87 25,295 114.71 7,026 162.97 10,764 brick infill RCC 85.94 38,750 70.73 16,146 59.81 12,917 278.71 10,764 184.59 16,489 107.35 10764 Any Other 18.58 10,764 22.55 2,990 39.91 4,580 2 Sources 2 BDA (2008), The Orissa Gazette, Bhubaneswar Development Authority (Planning and Building Standards), Regulation 2008, MSME (2012), Brief Industrial Profile of Khordha District, MSME-Development Institute, Ministry of MSME, Govt. of India, www.msmedicuttack.gov.in, NDMA (2013). Compilation of Catalogue of Building Typologies in India, Seismic Vulnerability Assessment of Building Types in India, Seismic Vulnerability Assessment Project Group, http://www.ndma.gov.in/images/disaster/earthquake/Catalogue%20of%20Building%20Types%20in%20India.pdf, NDMA (2013). Technical Document (Tech-Doc) on Typology of Buildings in India, Seismic Vulnerability Assessment of Building Types in India, Seismic Vulnerability Assessment Project Group, http://www.ndma.gov.in/images/disaster/earthquake/Building%20Typology%20Report.pdf Final Report: Volume I Confidential Page 79 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Very limited information was available for Nepal. Therefore, unit costs for Nepal were modified based on structure types and unit replacement costs for India along with the information available on mentioned websites. T a b l e 2 - 5 : U n i t r e p l a c e m e n t c o s t s f o r d i f f e r e n t t yp e s o f s t r u c t u r e s i n N e p a l All Housing Types Description Avg. Area, Unit Cost, sq m INR/Sq m Grass, Thatch, Bamboo, Wood, Mud, Plastic, etc. 11.6 404 Mud/Unburnt Brick/Stone without mortar 18.6 808 Burnt Brick/ stone with mortar having temporary Roof (Tiles, 23.2 2,020 wood, GI, slate, etc.) RCC 85.9 14,544 Any Other 18.6 4,040 3 Sources Very limited information was available for Bangladesh. Therefore, unit costs for Bangladesh were modified based on structure types and unit replacement costs for India along with the information available on mentioned websites. T a b l e 2 - 6 : U n i t r e p l a c e m e n t c o s t s f o r d i f f e r e n t t yp e s o f s t r u c t u r e s i n B a n g l a d e s h All Housing Types Description Avg. Area, Unit Cost, sq m INR/Sq m Grass, Thatch, Bamboo, Wood, Mud, Plastic, etc. 11.6 444 Mud/Unburnt Brick/Stone without mortar 18.6 889 Burnt Brick/ stone with mortar having temporary Roof (Tiles, 23.2 2,221 wood, GI, slate, etc.) RCC 85.9 15,995 4 Source Other types of structures Using a methodology similar to that explained in the previous sub-section, the built-up areas for other types of infrastructure have also been estimated. For linear features, such as roads and railways, the total length of the features were calculated, while for area features, such as educational institutes and health facilities, the area corresponding to the coverage was estimated. 2.3.2 CALCULATION OF TOTAL HOUSING EXPOSURE VALUES The estimated built-up floor area for each housing type was multiplied with the respective per unit replacement costs. Based on this, the total estimated exposure value for the houses was calculated. Based on similar kinds of material used in the house, the housings were 3 http://www.fncci.org/text/ind_fact.pdf http://www.nepalpropertymarket.com/index.php?linkID=178 http://unhabitat.org/?wpdmact=process&did=MzM5LmhvdGxpbms= 4 http://www.masshousingcompetition.org/sites/default/files/bd-dha-536summary_report.pdf Final Report: Volume I Confidential Page 80 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia grouped into seven structural classes (as mentioned previously) for the analysis and loss calculation. India The total exposure value of housing for the Ganges basin in India is estimated at INR 6,194,200 Crores. The state level distribution of total exposure value of various housing types for the basin in India is shown in Figure 2-74. F i g u r e 2 - 7 4 : S t a t e l e ve l d i s t r i b u t i o n o f t o t a l e x p o s u r e va l u e o f h o u s e s f o r t h e b a s i n i n India Nepal The total exposure value for housing in Nepal is estimated at INR 153,580 Crores. The province level distribution of total exposure value is shown in Figure 2-75. F i g u r e 2 - 7 5 : P r o v i n c e l e ve l d i s t r i b u t i o n o f t o t a l e x p o s u r e va l u e o f h o u s e s i n N e p a l Bangladesh The total exposure value of housing for the basin in Bangladesh is estimated at INR 55,800 Crores. The district level distribution of total exposure value is shown in Figure 2-76. Final Report: Volume I Confidential Page 81 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 2 - 7 6 : D i s t r i c t l e ve l d i s t r i b u t i o n o f t o t a l e x p o s u r e va l u e o f h o u s e s i n Bangladesh 2.3.3 TRANSPORTATION During disasters, the transportation systems of a region play an important role in rescue operations. Roads and railways have been considered in this part of the study. The length component has been estimated (in-house) using GIS software for both road and railway lines. Average unit replacement cost for both rail and road has been taken from literature review, RMSI’s experience/feedback from consultants and local field surveys fr om various previous projects conducted in the study area. 2.3.3.1 Roads The road data provides information regarding the length of roads for each sub-district level for India, district wise for Nepal and Bangladesh. Total length of the roads have been calculated to compute the exposure value for each district/sub-district. In order to estimate the replacement cost of roads, average unit cost of the roads have been determined by taking inputs from relevant literature. The average unit cost of roads in India has been determined from available resources in the National Highway Development Projects (NHDP) of the National Highways Authority of India, Ministry of Road, Transport and Highways, and Planning Commission documentation. For Nepal, the website of Department of Roads, Government of Nepal, and research papers such as the average unit cost of road infrastructure in developing countries (Collier. P et al.,2013) were used as inputs. For Bangladesh, the same research paper was referred to (Collier.P et al.,2013). Full references have been provided at the end of this report in addition to appropriate places in the write-up. India The total estimated exposure value for the road network has been estimated at INR 47,070 Crores. The average unit cost is INR 0.58 Crore/km5. The following Table 2-7 illustrates the types of roads and associated costs in India. 5 http://www.nhai.org/index.asp Final Report: Volume I Confidential Page 82 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 2 - 7 : T yp e s o f r o a d s a n d a s s o c i a t e d c o s t s i n I n d i a Sr. Type of Cost (INR/ Road Type No. Construction km) Sources Concrete, Double NHDP, National highway (Two Lane) 1 Seal 12,000,000 NHAI National highway (Four/ Six Concrete, Double NHDP, 2 Lane) Seal 60,000,000 NHAI 3 Main Roads (Four Lane) Double Seal 5,000,000 4 Other Roads (Two Lanes) Double/ Single Seal 2,500,000 5 Link Roads Single Seal 1,250,000 The road shape file collected from open source was not of much help in terms of attribute information. Data represented in Table 2-7 together with the length of each type of road and its respective replacement cost helped calculate the weighted average unit cost of roads as INR 0.58 Crores/km throughout the portion of Ganges basin in India. The following Table 2-8 illustrates the process used to reach the final weighted average unit cost. Table 2-8: Weighted average unit cost calculation for roads Avg Unit Cost in Total Value Final Weighted Total Length Road Type INR (based on (Col2 *Col3) Average Cost of roads, km Previous table) (1*106) in INR National highway (Sr No. 1 & 2) 36,000,000 29,214 1,051,704 Rest of the roads 5,798,849 (Sr No. 3, 4 & 5) 2,916,667 306,121 892,852 Total 38,916,667 335,335 1,944,556 Figure 2-77 shows the values at state level. F i g u r e 2 - 7 7 : S t a t e l e ve l d i s t r i b u t i o n o f e x p o s u r e va l u e o f r o a d s i n I n d i a Final Report: Volume I Confidential Page 83 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Nepal The total estimated exposure value for the road network has been estimated at INR 4,460 Crores. The average unit rate is INR 0.4 Crore/km6. Figure 2-78 shows the values at province level. F i g u r e 2 - 7 8 : P r o v i n c e l e ve l d i s t r i b u t i o n o f e x p o s u r e va l u e f o r r o a d s i n N e p a l Bangladesh The total estimated exposure value for the road network has been estimated at INR 388 Crore. The average unit rate is INR 0.5 Crore/km7. The district level exposure values have been shown in the Figure 2-79. F i g u r e 2 - 7 9 : D i s t r i c t l e ve l d i s t r i b u t i o n o f r o a d e x p o s u r e va l u e s f o r B a n g l a d e s h 2.3.3.2 Railways The data regarding the railway network of India has been obtained from literature survey and previous experience in the study area. Total length of the railway network has been calculated to compute the exposure value. In order to estimate the average replacement cost of the railway network, unit cost of the railway network has been determined by taking inputs from relevant literature. The estimated length of the railway network was multiplied with per average unit replacement costs of the railway. 6 http://www.sv.uio.no/esop/english/research/news-and-events/events/guest-lectures-seminars/esop- seminar/dokumenter/soderbom.pdf 7 http://www.sv.uio.no/esop/english/research/news-and-events/events/guest-lectures-seminars/esop- seminar/dokumenter/soderbom.pdf Final Report: Volume I Confidential Page 84 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia For India, relevant average unit cost was obtained from authentic sources such as Planning Commission, Total Transport System Study and Indian and Northern Railways’ websites of the Ministry of Railways, Government of India. Full references are provided at the end of this report. Since less information was obtained regarding Nepal and Bangladesh, a 10% increase of average unit cost was considered for Nepal given the terrain conditions. Bangladesh was assigned an average unit cost similar to India. India The total estimated exposure value for the railway network has been estimated as INR 122,425 Crores. The average unit cost is INR 6.3 Crore/km8. This Figure is based on: Average replacement cost = INR 5.5 crore per km (single broad gauge line) and additional electrification costs = INR 0.8 crore per km (single line). Figure 2-80 shows the state level values across India. F i g u r e 2 - 8 0 : S t a t e l e ve l e x p o s u r e va l u e s o f r a i l w a y t r a c k s i n I n d i a Nepal The total estimated exposure value for the railway network was estimated at INR 1,065 Crore. The average unit cost was INR 7.0 Crore/km. Figure 2-81 shows the province level values. 8 http://www.core.indianrailways.gov.in/view_section.jsp?lang=0&id=0,294,302,536 http://www.ireeindia.org/RE%20Booklet.pdf http://www.nr.indianrailways.gov.in/view_detail.jsp?lang=0&dcd=3265&id=0,4,268 http://indiarailinfo.com/news/post/construction-of-a-new-broad-gauge-line-between-pirpainti-jasidih- mohanpur-indian-railways-news/167878 http://www.projectstoday.com/News/CCEA-approves-new-broad-gauge-lines-in-India http://pib.nic.in/newsite/erelease.aspx?relid=103623 Final Report: Volume I Confidential Page 85 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 2 - 8 1 : P r o v i n c e l e ve l d i s t r i b u t i o n o f e x p o s u r e va l u e s o f r a i l w a y t r a c k s i n N e p a l . Bangladesh The total estimated exposure value for the railway network has been estimated at INR 1,418 Crore. The unit average cost was INR 6.3 Crore/km. The district level railway exposure values are shown in the Figure 2-82. F i g u r e 2 - 8 2 : D i s t r i c t l e ve l r a i l w a y e x p o s u r e v a l u e s o f B a n g l a d e s h 2.3.4 AGRICULTURE The exposure values for Rice, Wheat and Maize of the three countries, India, Nepal and Bangladesh for the portion occupied by the Ganges basin are described below. The unit replacement costs for agriculture exposure are based on the wholesale market prices, which have been taken from various agriculture/food department websites of the concerned country. An average single replacement cost has been assigned for each crop per country. For data related to India, the Directorate of Economics and Statistics, Department of Agriculture and Cooperation, Ministry of Agriculture was referenced. Final Report: Volume I Confidential Page 86 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia For Nepal, agricultural data was referenced from websites dealing with commodity pricing in Nepalese rupees per metric ton. Exposure values for agriculture in Bangladesh were referenced from the Ministry of Food in Bangladesh Takas. Detailed references have also been provided at the end of this report in addition to appropriate places in text. Table 2-9, Table 2-10 and Table 2-11 display the average exposure values in INR per metric ton for India, Nepal, and Bangladesh respectively. For India, the price-index file was downloaded from the website of Directorate of Economics and Statistics which is governed by Department of Agriculture and Corporation (http://eands.dacnet.nic.in/Publication/pub-2013/Index.xls). The wholesale prices of each state were studied and an average value of all the states was calculated. The average value of all these values throughout India was taken per crop and is given in Table 2-9. T a b l e 2 - 9 : E x p o s u r e va l u e s f o r c r o p s i n I n d i a India Crop Average Unit Cost (INR/MT) Rice 24,341 Wheat 17,355 Maize 13,956 9 Sources The values for each crop in Nepal was taken from the literature review and is mentioned along with Table 2-10. Table 2-10: Exposure values for crops in Nepal Nepal 10 Average Unit Cost Crop (INR/MT) Rice 41,592 Wheat 20,902 Maize 10,879 11 Sources 9 http://eands.dacnet.nic.in/publications.htm http://eands.dacnet.nic.in/Publication/pub-2013/Index.xls 10 Conversion Rate : 1 Nepalese Rupee = 0.56 Indian Rupees 11 http://www.numbeo.com/food-prices/country_result.jsp?country=Nepal http://www.indexmundi.com/commodities/?commodity=wheat&months=12¤cy=npr Final Report: Volume I Confidential Page 87 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia The exposure value for Rice and Wheat in Bangladesh was taken from literature review available at Ministry of Food as illustrated in Table 2-11. Table 2-11: Exposure values for crops in Bangladesh Bangladesh 12 Average Unit Cost Crop (INR/MT) Rice 28,090 Wheat 17,780 13 Source India The exposure value of Rice sown in India was INR 40,130 Crores (2010/11) and is illustrated at the state level below in Figure 2-83. Uttar Pradesh has the highest exposure value in India. F i g u r e 2 - 8 3 : E x p o s u r e va l u e o f R i c e s o w n i n I n d i a d u r i n g 2 0 1 0 / 1 1 The exposure value of Wheat was INR 88,236 Crores (2010/11) and is illustrated at the state level in Figure 2-84. Uttar Pradesh has the highest exposure value in India. 12 Conversion Rate : 1 Bangladeshi Taka = 0.58 Indian Rupees 13 www.mofood.gov.bd http://www.mofood.gov.bd/site/page/5eaf13d6-834e-44e6-bb43-bd9fb9204af9/Market-Price Final Report: Volume I Confidential Page 88 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 2 - 8 4 : E x p o s u r e va l u e o f W h e a t s o w n i n I n d i a d u r i n g 2 0 1 0 / 1 1 The exposure value of Maize was INR 6,890 Crores (2010/11) and is illustrated at the state level below in Figure 2-85. Bihar has the highest exposure value in India. F i g u r e 2 - 8 5 : E x p o s u r e va l u e o f M a i z e s o w n i n I n d i a d u r i n g 2 0 1 0 / 1 1 Nepal The exposure value of Rice was INR 18,735 Crores (2012/13) and is illustrated at the province level in Figure 2-86. Central province has the highest exposure value in Nepal. Final Report: Volume I Confidential Page 89 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 2 - 8 6 : E x p o s u r e va l u e o f R i c e s o w n i n N e p a l d u r i n g 2 0 1 2 / 1 3 The exposure value of Wheat was INR 3,934 Crores (2012/13) and is illustrated at the province level below in Figure 2-87. Central province has the highest exposure value in Nepal. F i g u r e 2 - 8 7 : E x p o s u r e va l u e o f W h e a t s o w n i n N e p a l d u r i n g 2 0 1 2 / 1 3 The exposure value of Maize was INR 2,174 Crores (2012/13) and is illustrated at the province level below in Figure 2-88. Western province has the highest exposure value in Nepal. Final Report: Volume I Confidential Page 90 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 2 - 8 8 : E x p o s u r e va l u e o f M a i z e s o w n i n N e p a l d u r i n g 2 0 1 2 / 1 3 Bangladesh The exposure value of Rice was INR 5,607 Crores (2009/10) and is illustrated at the district level below in Figure 2-89. Thakurgaon district has the highest exposure value in Bangladesh. F i g u r e 2 - 8 9 : E x p o s u r e va l u e o f R i c e s o w n i n B a n g l a d e s h d u r i n g 2 0 0 9 / 1 0 The exposure value of Wheat was INR 404 Crores (2009/10) and is illustrated at the district level below in Figure 2-90. Thakurgaon district has the highest exposure value in Bangladesh. Final Report: Volume I Confidential Page 91 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 2 - 9 0 : E x p o s u r e va l u e o f W h e a t s o w n i n B a n g l a d e s h d u r i n g 2 0 0 9 / 1 0 Final Report: Volume I Confidential Page 92 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 3 Flood Hazard Assessment 3.1 Methodology overview Flood hazard assessment identifies and demarcates those parts of the study area, which are exposed to floods. It provides information on the extent and depth of flooding throughout flood prone areas for a range of flood magnitudes. The flood hazard model development framework adopted for this study is given in Figure 3-1, which comprises of the following: ï‚· Collection and compilation of relevant hydro meteorological and biophysical data. These data include terrain, soil, land use land cover, runoff/river discharge, and flood protection measures to form the input for the model. ï‚· Hydraulic modeling to estimate flood levels throughout the basins for historical flows ï‚· Calibration and validation of hydraulic model using the historical flood gauge-discharge data ï‚· Probabilistic analysis of runoff to simulate various return period events (2, 5, 10, 25, 50, and 100 years) ï‚· Flood hazard mapping to show flood extent and flood depth for different return periods Figure 3-1: Flood hazard assessment framework 3.2 Data availability This section describes the availability status of meteorological, hydrological, topographical, soil class, Land Use Land Cover (LULC) and other data required as inputs at the various steps of modeling. 3.2.1 METEOROLOGICAL DATA Daily rainfall data were obtained from IMD and DHM Nepal. The following were available: ï‚· Half degree gridded rainfall data from 1971 to 2005 from IMD, India. Final Report: Volume I Confidential Page 93 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia ï‚· Daily station rainfall from 1960 to 2014 from DHM Nepal Half-degree gridded rainfall data from IMD and daily rainfall data from DHM Nepal have been used for the hydrological modeling of the basin. The locations and the spatial distribution of the rain gauges are shown in Figure 3-2. Table 3-1 shows the list of the rainfall stations along with the duration of the data for the respective stations used in the study. Table 3-1: Station wise rainfall data records Sr. Station ID/Name Country Duration Source No. India Half Degree Gridded Data (283 India and 1 1971-2005 IMD, India Stations) Bangladesh Nepal 2 206- Asara Ghat Nepal 1963-2013 DHM, Nepal 3 218- Dipayal (Doti) Nepal 1982-2013 DHM, Nepal 4 311- Simikot Nepal 1978-2006 DHM, Nepal 5 403- Jamu (Tikuwa Kuna) Nepal 1963-2013 DHM, Nepal 6 606- Tatopani Nepal 1969-2013 DHM, Nepal 7 927- Bharatpur Nepal 2001-2014 DHM, Nepal 8 1002- Aru Ghat D. Bazar Nepal 1960-2013 DHM, Nepal 9 1004- Nuwakot Nepal 1960-2013 DHM, Nepal 10 1115- Nepalthok Nepal 1960-2013 DHM, Nepal 11 1210- Kurule Ghat Nepal 1960-2013 DHM, Nepal 12 1316- Chatara Nepal 1960-2013 DHM, Nepal 13 1317- Chepuwa Nepal 1960-2013 DHM, Nepal Final Report: Volume I Confidential Page 94 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Figure 3-2: Location of rainfall stations used in the stud y 3.2.2 HYDROLOGICAL DATA The flow data have been collected from various sources like Central Water Commission, India, Department of Hydrology and Meteorology (DHM), Nepal, and Dartmouth Flood Observatory (DFO). These flow data were available for limited duration/dates only and most of the data was satellite based/re-analyzed flow data, which had some quality issues. The discharge and runoffs data collected from Dartmouth Flood Observatory website were analyzed and estimated based on satellite remote sensing (NASA AMSR-E data). Some additional flow data have also been received from the World Bank during the course of the project. The average duration of flow data for these stations is from 1975 to 2005. All these flow data have been used in calibration and validation of the models as per their suitability. The location and the spatial distribution of the flow gauges are shown in Figure 3-3. Table 3-2 shows the details of the stations along with the duration of available flow data. As per Table 3-2, data of one station is from CWC, five stations from DHM, Nepal, and ten stations from DFO. The World Bank has provided the data for the rest of the 26 stations. T a b l e 3 - 2 : A va i l a b i l i t y o f d i s c h a r g e d a t a Sr. No. Gauge Name Latitude Longitude Duration Source 1 Agra 27.20 78.07 1950-2009 CWC, India 2 Baeraich 27.66 81.29 1998-2014 DFO 3 Begusarai 25.33 86.37 1998-2014 DFO 4 Bettiah 26.66 84.44 1998-2014 DFO 5 Dehri 24.71 84.04 1998-2014 DFO Final Report: Volume I Confidential Page 95 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Sr. No. Gauge Name Latitude Longitude Duration Source 6 English Bazar 24.98 87.90 1998-2014 DFO 7 Gopiganj 25.26 82.39 1998-2014 DFO 8 Gosaniganj 26.55 82.68 1998-2014 DFO 9 Jangipur 24.45 88.24 1998-2014 DFO 10 Rudauli-Faizabad 26.81 82.14 1998-2014 DFO 11 Saharsa 26.25 87.00 1998-2014 DFO 12 Chatara-Kothu 26.90 87.17 2002-2007 DHM, Nepal 13 Chispani 28.64 81.29 2002-2006 DHM, Nepal 14 Kali Khola 28.98 82.59 2002-2006 DHM, Nepal 15 Kota Gaon Shringe 27.75 84.35 2002-2006 DHM, Nepal 16 Narayan Ghat 27.71 84.43 2002-2006 DHM, Nepal 17 Ankinghat 26.74 80.13 1975-2005 The World Bank 18 Ayodhya 26.81 82.23 1995-2005 The World Bank 19 Barah 26.85 87.15 2009-2013 The World Bank 20 Baranwada 25.87 76.62 1975-2005 The World Bank 21 Bigod 25.20 74.95 1975-2005 The World Bank 22 Borlangpul 27.97 83.57 1975-2005 The World Bank 23 Chillaghat 25.76 80.51 1975-2005 The World Bank 24 Chisapani 28.64 81.29 1975-2005 The World Bank 25 Delhi Rly Bridge 28.51 77.42 1975-2005 The World Bank 26 Etawah 26.61 79.09 1975-2005 The World Bank 27 Hathidah 25.32 86.06 1975-2005 The World Bank 28 Jalkundi 27.95 82.23 1975-2005 The World Bank 29 Jamu 28.76 81.35 1975-2005 The World Bank 30 Japla 24.52 83.87 1975-2005 The World Bank 31 Kampughat 26.87 86.82 1975-2005 The World Bank 32 Koelwar 25.70 84.91 1975-2005 The World Bank 33 Kosi Barrage (Birpur) 26.53 86.93 2009-2013 The World Bank 34 Kotagaun 27.75 84.35 1975-2005 The World Bank 35 Lalbegia ghat 26.50 85.03 1975-2005 The World Bank 36 Mejja Road 25.27 82.09 1975-2005 The World Bank 37 Paliakalan 28.25 80.71 1995-2005 The World Bank 38 Rudraprayag 30.30 79.02 1975-2005 The World Bank 39 Sahibganj 25.10 87.82 1975-2005 The World Bank 40 Samaijighat 28.52 81.66 1975-2005 The World Bank 41 Tonk 26.18 75.73 1975-2005 The World Bank 42 Varanasi 25.49 83.16 1975-2005 The World Bank *DFO-Dartmouth Flood Observatory, *DHM -Department of Hydrology and Meteorology, Nepal Final Report: Volume I Confidential Page 96 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Figure 3-3: Flow gauge station locations For readers’ information, a list and map of all the hydrological gauges lying in Ganges Basin have been provided in Appendix A of the report. In addition to the above hydrological data, a separate set of gauge-discharge data was made available for RMSI modeling team to calibrate and validate Ganges Hydraulic model which is discussed in hydraulic modeling section. 3.2.3 TOPOGRAPHICAL INFORMATION Topographical data is required for the delineation of catchment areas and for generating the river network. The same topographical data is required to estimate the elevation information for cross-sections in hydraulic modeling. SRTM DEM having 90-m resolution was used for this purpose. Figure 3-4 shows the elevation map of the study area. Final Report: Volume I Confidential Page 97 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 3 - 4 : E l e va t i o n m a p o f t h e s t u d y a r e a The Ganges basin lies between elevations of 0 to 8,800 m. Most of the Gangetic plains lie at an elevation of less than 400 m with some southern and southwestern parts having an elevation range of 400 m to 800 m. Uttarakhand in India and Nepal lie at an elevation of more than 1,000 m. 3.2.4 LAND USE LAND COVER (LULC) Land-use land-cover (LULC) classes have been identified from the U.S. Geological Survey's (USGS) site. According to this data, approximately 7% of the study area is forest, 24% area is open/grassland, and the remaining 62% area is used as cropland. Figure 3-5 shows the LULC map of Ganges River basin. Final Report: Volume I Confidential Page 98 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 3 - 5 : L U L C m a p o f G a n g e s R i ve r b a s i n 3.2.5 SOIL MAP Soil data is an essential parameter used to estimate the hydrological response characteristics of a river basin. The project team reviewed available data sets from project documents and in-house datasets. Since a single soil database is not available for the entire basin, soil data included in the Food and Agriculture Organization’s (FAO) Harmonized World Soil Database was used. This soil data is provided as a set of land units, each with a unique ID number. This unique ID number is used to match the textural properties and other parameters of soils. Based on the soil textural class, a hydrological soil group was assigned to each land unit within the basin. The soil map for the study area is shown in Figure 3-6. The soils are mainly clay (53%) and loamy sand (23%). Final Report: Volume I Confidential Page 99 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Figure 3-6: Soil map for the stud y area 3.2.6 FLOOD HISTORY The RMSI team tried to gather damage and loss data from government websites and/or publicly available sources. Flood history is one of the important aspects, which needs to be compiled to get an overall picture of frequently flood-impacted parts of the study area. In this study, the collected flood maps were processed and have been used in calibration of flood extent. In addition, the damage details give an idea of the flood prone districts, damage details and other valuable information, which can be used to compare the kind of damages being estimated using the modeled hazard. The collected damage details mostly cover the states of Bihar and UP. This is firstly because the maximum area of the Ganges Basin lies in these two states. Secondly, these two states get frequently affected due to floods in the basin. Detailed information for the whole basin for a single event is sparse. The present report presents this collected information only on an as-is-where-is-basis for end user reference. Historical loss information for major flood events that occurred during 1979-2006 for the state of Bihar is provided in Table 3-3. Similarly, detailed damage information caused by the flood event of 1973 for the state of Uttar Pradesh has been presented in Table 3-4. Final Report: Volume I Confidential Page 100 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 3 - 3 : F l o o d l o s s s u m m a r y: B i h a r , 1 9 7 9 t o 2 0 0 6 Affected Affected Public No. of No. of No. of No. of Affected Crop No. of Area (in Crop Area Property Year Affected Affected Affected Affected Population Damage (in Houses Lakh (in Lakh Damage (in Districts Blocks Panchayats Villages (in Lakh) Lakh INR) Affected ha) ha) Lakh INR) 2006 14 63 375 959 10.89 1.81 0.87 706.63 18,637 8,456.17 2005 12 81 562 1,464 21.04 4.60 1.35 1,164.50 5,538 305.00 2004 20 211 2,788 9,346 212.99 27.00 13.99 52,205.64 9,29,773 1,03,049.60 2003 24 172 1,496 5,077 76.02 15.08 6.10 6,266.13 45,262 1,035.16 2002 25 6 2,504 8,318 160.18 19.69 9.40 51,149.61 419,014 40,892.19 2001 22 194 1,992 6,405 90.91 11.95 6.50 26,721.79 222,074 18,353.78 2000 33 213 2,327 12,351 90.18 8.05 4.43 8,303.70 343,091 3,780.66 1999 24 150 1,604 5,057 65.66 8.45 3.04 24,203.88 91,813 5,409.99 1998 28 260 2,739 8,347 134.70 25.12 12.84 36,696.68 199,611 9,284.04 1997 26 169 1,902 7,043 69.65 14.71 6.55 5,737.66 174,379 2,038.09 1996 29 195 2,049 6,417 67.33 11.89 7.34 7,169.29 116,194 1,035.70 1995 26 177 1,901 8,233 66.29 9.26 4.24 19,514.32 297,765 2,183.57 1994 21 112 1,045 2,755 40.12 6.32 3.50 5,616.33 33,876 151.66 1993 18 124 1,263 3,422 53.52 15.64 11.35 13,950.17 219,826 3,040.86 1992 8 19 170 414 5.56 0.76 0.25 58.09 1,281 0.75 1991 24 137 1,336 4,096 48.23 9.80 4.05 2,361.03 27,324 139.93 1990 24 162 1,259 4,178 39.57 8.73 3.21 1,818.88 11,009 182.27 1989 16 74 652 1,821 18.79 4.71 1.65 704.88 7,746 83.7 1988 23 181 1,616 5,687 62.34 10.52 3.95 4,986.32 14,759 150.64 1987 30 382 6,112 24,518 286.62 47.50 25.70 67,881.00 1,704,999 680.86 1986 23 189 1,828 6,509 75.80 19.18 7.97 10,513.51 136,774 3,201.99 1985 20 162 1,245 5,315 53.09 7.94 4.38 3,129.52 103,279 204.64 1984 23 239 3,209 11,154 135.00 30.50 15.87 18,543.85 310,405 2,717.72 1983 22 138 1,224 4,060 42.41 18.13 5.78 2,629.25 38,679 258.14 1982 15 110 1,112 3,708 46.81 9.32 3.23 9,700.00 68,242 955.33 1981 21 201 2,138 7,367 69.47 12.61 7.71 7,213.19 75,776 1980 21 193 1,869 7,010 74.45 17.86 9.43 7,608.43 118,507 1979 13 110 37.38 8.06 2.74 1,901.52 27,816 Final Report: Volume I Confidential Page 101 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 3 - 4 : F l o o d l o s s s u m m a r y: U t t a r P r a d e s h , 1 9 7 3 f l o o d No. of Number of Affected Sr. Affected Affected Affected Crop Reported Area Name Affected Houses Area (in Lakh No. Population Area (ha.) Area (ha.) Casualty Villages Affected ha) 1 Muzaffarnagar - 8,913 54,543 92,113 2,784 - 10.76 2 Meerut 158 45,039 99,764 42,443 1,053 2 14.87 3 Buland Shahar 64 22,857 21,628 8,298 40 - 12.08 4 Aligarh 8 140 170 166 25 - - 5 Mathura 110 39,093 23,883 14,332 668 - 9.38 6 Agra 35 2,714 6,000 741 13 - 11.90 7 Etah 127 96,191 91,207 38,445 523 - 10.97 8 Farrukhabad 407 116,392 206,338 71,205 3,625 4 11.52 9 Etawah 116 9,992 26,728 12,057 - - 10.07 10 Bijnor 444 119,641 151,461 61,509 7,463 13 11.54 11 Bareilly 934 231,103 175,880 84,999 909 - 10.17 12 Pilibhit 126 6,065 128,000 34,408 39 - 8.64 13 Shahjahanpur 672 270,106 122,864 78,666 2,164 2 11.28 14 Muradabad 735 201,758 334,032 174,444 19,069 1 14.62 15 Badaun 685 274,892 361,646 147,981 3,681 2 12.78 16 Rampur 561 238,165 162,414 144,836 7,350 - 5.73 17 Varanasi 358 48,147 26,593 14,218 16 - 12.57 18 Mirjapur 344 138,225 121,800 38,540 73 - 27.96 19 Jaunpur 495 194,952 125,000 41,155 713 4 8.87 20 Gazipur 534 312,733 125,620 71,971 15 2 8.35 21 Balia 1,564 695,577 304,232 182,463 17,226 12 7.56 22 Gorakhpur 1,478 750,483 392,171 222,256 29,910 17 15.56 23 Deoria 3,509 2,600,000 919,097 722,470 50,882 35 13.35 24 Basti 3,832 970,016 498,628 403,707 23,787 8 18.05 25 Azamgarh 1,043 675,753 228,565 409,530 2,103 14 14.21 26 Gonda - - 28,700 98,693 1,200 - 18.10 27 Behraich 242 50,000 63,000 54,208 489 - 16.76 28 Faizabad 721 148,400 23,360 64,122 825 15 10.90 29 Sultanpur 213 174,995 26,284 26,284 175 - 10.96 30 Prataphgarh 980 712,084 228,658 186,536 288 - 9.33 31 Allahabad 537 205,922 152,391 58,197 2,981 2 17.92 32 Kanpur 157 13,481 26,958 7,289 390 - 15.08 33 Fatehpur 124 25,351 33,858 82,224 15 - 10.39 Final Report: Volume I Confidential Page 102 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia No. of Number of Affected Sr. Affected Affected Affected Crop Reported Area Name Affected Houses Area (in Lakh No. Population Area (ha.) Area (ha.) Casualty Villages Affected ha) 34 Kheri 137 36,605 33,207 26,860 255 4 19.02 35 Sitapur 188 105,000 50,000 13,000 - -1 14.31 36 Hardoi 448 290,398 242,755 96,212 2,477 2 14.80 37 Lucknow 70 30,257 5,540 3,212 130 1 6.20 38 Unnao 476 350,000 73,612 73,612 2,595 1 11.35 39 Raibariely 1,283 874,099 831,593 270,988 7,198 6 11.25 40 Barabanki 242 92,823 101,291 52,242 1,600 1 11.96 41 Jalaun 83 23,691 44,892 20,123 - - 11.28 42 Hamirpur 112 23,537 71,053 38,340 100 - 17.76 43 Banda 265 118,230 190,103 77,110 465 - 18.86 Final Report: Volume I Confidential Page 103 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia RMSI has also obtained historical flood footprint maps for various flood events that have occurred at various places from year 2001 to 2008. These have been collected from various public sources such as DFO. These maps have been processed, geo-referenced, and digitized to validate the hydraulic model output. These flood footprints are also used in the Web Risk Atlas to identify the historically damaged areas of the basin. One such flood footprint of the 2007 flood event has been shown in Figure 3-7.The reported damages due to this event in Bihar have also been obtained and are shown in the Table 3-5. Figure 3-7: Flood Footprint: 2007 flood Final Report: Volume I Confidential Page 104 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 3 - 5 : F l o o d l o s s s u m m a r y: B i h a r , 2 0 0 7 f l o o d Estimated Public No. of No. of Affected Crop Number of Sr. Value of house Property Reported District Name Affected Affected Crop Area Damage (in Houses No. damage In Damage (in Casualty Blocks Villages (Lakh ha.) Lakh INR) Affected/Damaged (Lac INR) Lakh INR) 1 Muzaffarpur 15 1,704 1.24 12,663.00 65,550 11,073.00 24,951.00 104 2 Sitamarhi 17 806 0.51 7,803.94 103,193 16,084.85 63,618.25 33 3 Saharsa 6 184 0.26 985.82 16,412 935.25 140.38 35 4 E.Champaran 27 1,159 1.58 15,400.00 52,840 8,278.31 96 5 Supaul 6 94 0.25 574.84 15,000 300.00 17.75 1 6 Darbhanga 18 2,104 1.75 6,606.10 83,127 13,106.14 18,271.02 140 7 Madhubani 20 836 1.39 7,936.23 96,362 9,116.45 25,733.68 49 8 Samastipur 19 842 1.25 16,710.07 29,391 775.00 17,896.46 157 9 Sheohar 5 150 0.25 693.00 50,728 6,477.10 105.00 4 10 Khagaria 7 203 0.5 8,507.33 32,500 8,507.33 372.00 101 11 Begusarai 11 346 0.94 16,057.84 40,740 11,773.10 444.00 54 Total 151 8,428 9.92 93,938.17 585,843 86,426.53 151,549.54 774 Final Report: Volume I Confidential Page 105 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 3.3 Basin and River Delineation In order to do hydraulic modeling, detailed river network and catchment boundary is required. For this purpose, a number of hydrologic modeling inputs were developed using the Geospatial Hydrologic Modeling Extension (HEC- GeoHMS) tool (USACE 2009). HEC-GeoHMS works within a Geographic Information System (GIS) interface. HEC- GeoHMS transforms digital terrain information like drainage paths and watershed boundaries into a hydrologic data structure that represents the watershed response to precipitation. Using HEC Geo-HMS, the river network and sub basins were delineated using a systematic approach. The approach creates raster grids for catchment delineation using DEM as input. Activities to complete the model include filling sinks, creating flow directions and flow accumulation grids, processing catchment grid, and processing drainage line. Filling sinks is the process of numerically correcting the DEM, where large sinks (abnormal depressions) or voids are present. The delineated river network using the model has been overlaid and verified with the river network map available at WRIS14 system. Physical representation of the basin incorporates various hydrologic elements (sub basins, river reaches, junctions, and reservoirs), which are connected in a dendritic network to simulate the rainfall-runoff process. Figure 3-8 shows the map of delineated sub basins for the Ganges River basin. Various small catchments were merged together based on the major river they are contributing to. A major sub basin file was created to calibrate and validate the model and to aggregate exposure, hazard, and damage. The list of major sub basins is provided in Table 3-6. Table 3-6: Major sub basins considered in stud y Name of Basin Names of Major Sub basins Bagmati Betwa Chambal Gandak Ghagra Gomti Kamla-Balan Ken Ganges River Basin Kosi Lower Ganges Mahananda Middle Ganges Ramganga Sind Sone Tons Upper Ganges 14 Water Resource Information System of India WRIS is governed by Central Water Commission, Government of India. (http://india-wris.nrsc.gov.in/wrpinfo/index.php?title=Ganga) Final Report: Volume I Confidential Page 106 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Name of Basin Names of Major Sub basins Yamuna F i g u r e 3 - 8 : D e l i n e a t e d c a t c h m e n t s a n d m a j o r s u b b a s i n s o f G a n g e s R i ve r B a s i n 3.4 Hydraulic modeling Flood flows are provided as the input to the hydraulic model. A detailed hydraulic analysis of the major streams of the study area was conducted to develop water surface elevations, floodplain boundaries, and depth of flooding in streams for selected frequency storm events. The hydraulic analysis was performed using HEC-RAS and HEC-GeoRAS. The steps included in the hydraulic analysis are: a) Model set up b) Model calibration and validation c) Estimation of return period flows d) Flood hazard mapping 3.4.1 MODEL SET UP The hydraulic model calculates flood elevations along streams and rivers for flood flows of various historical and return periods ranging from the most frequent to rare events. Flood elevations are then used to delineate the aerial extent of flooding adjacent to the streams and rivers. This technical effort serves to identify areas of flood inundation within the floodplain that are at risk and subject to flood damage. Derivation of flood extent and flood depths is determined using 1D hydraulic modeling through the river system for all historical and probabilistic events. Final Report: Volume I Confidential Page 107 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Detailed hydraulic modeling requires discharge information, cross-sections of streams and rivers, and elevation information. HEC-RAS Model developed by the United States Army Corps of Engineer’s Hydrologic Engineering Centre (USACE 2010), was used for performing hydraulic calculations for the river stretches. HEC-RAS is an integrated system of software that contains one-dimensional hydraulic analysis components for both steady and unsteady, gradually varied flow simulation for a full network of natural and constructed channels. The basic computational procedure is based on the solution of the one-dimensional energy equation. Energy losses are evaluated by friction (Manning’s equation) as also expansion and contraction losses. The momentum equation is utilized in situations where the water surface profile is rapidly varied. The situations include a mixed flow regime (USACE 2010). Basin geometric data consist of the river system connecting all segments, cross-section data, reach lengths, energy loss coefficients, and stream junction information. The river system schematic defines how the various river reaches are connected as well as establishes the naming conventions for referencing all the other data. The connecting river reaches are important for the model to understand how the computations should proceed from one reach to the next. The river system schematic is performed using HEC-Geo-RAS (an Arcview extension for pre and post processing of RAS) in GIS environment using ESRI’s Arcview. HEC-GeoRAS was used to create a HEC-RAS import file containing geometric attribute data from a Digital Elevation Model (DEM). The HEC-RAS models were setup for all segments of the river basins under the study area. Steady flow simulation was adopted for this study. Using the rivers delineated in the basin delineation process of hydrological modeling, the cross sectional geometry was derived for all the rivers at approximate intervals of 100 m. The elevation information was extracted using the SRTM DEM. Hydraulic roughness was estimated initially using the land use map and visual representation from Google Earth. For this purpose, reference of published literature, on variations of roughness coefficient with different kinds of channels and flood plains, has been taken. [Chow, 1959]. Initial roughness values of 0.025 were adopted for the river channels and 0.035 for the floodplains. These were subsequently varied during the model calibration process. Estimated runoff is, thus, routed through the river system using the above one-dimensional hydraulic analysis to delineate flood extents and depth at all the river segments. Figure 3-9 shows the HEC-RAS set up with a plan view of the modeled reaches, longitudinal riverbed profile of a reach, and a typical cross-section for a reach for the river network in the study area. Final Report: Volume I Confidential Page 108 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Figure 3-9: HEC RAS model set up for the stud y area 3.4.2 MODEL CALIBRATION AND VALIDATION The development of hydraulic models across a large floodplain requires a rigorous calibration process to ensure the hydraulic model accurately reproduces the observed flooding behavior. The calibration process consists of systematically comparing observed flooding behavior within the study area against the hydraulic model’s reproduction of that behavior. This process generally incorporates comparisons between simulated flood levels and observed flood levels. (http://www.wcma.vic.gov.au/index2.php?option=com_docman&task=doc_view&gid=385&Ite mid=50).The approach requires detailed data about the flood water levels over time (temporal distribution) at discrete points of interest within and along the river at various locations to validate the simulated water level. Flood extent and flood depth maps have been generated by post-processing the simulated results of HEC-RAS in Arcview environment with HEC-GeoRAS extension. Flood plain boundaries and inundation depth data sets were generated from exported cross-sectional water surface elevations. The Central Water Commission (CWC) has facilitated calibration and validation of Ganges Hydraulic model by allowing RMSI team to work at CWC premises. The following is the summary of whole process: 1. Data Available- The Central Water Commission (CWC) made available daily gauge discharge data along with corresponding water surface elevations and reduced levels of gauges for 179 stations. Data was provided for seven events i.e. 1978, 1981, 1988, 1996, 1998, 2004, and 2008. Events selected were based on historical flood records. Final Report: Volume I Confidential Page 109 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Ganges basin is prone to severe recurring floods and events so chosen were in the descending order of severity of floods. 2. Quality Check and data filtering- Quality analysis of data received was carried out. First check was carried out to see if data received was consistent without showing any atypical deviation. Also compared were the values at upstream and downstream stations relative to each other to check if any upstream station discharge value was greater than that of value at its downstream station without any storage, structure or dam. Out of 179 stations, a few stations were duplicated. Some stations had data values which were abnormally high for a single event, signifying a flawed value. Rest of the stations showed consistency in values from upstream to downstream. The repeated stations and stations with erroneous data were abandoned and 170 stations were finally used to calibrate and validate the model. 3. Spatial Distribution of Gauge Discharge Stations- The location of gauge discharge stations was uniform across the basin in India and all the major sub basins had at least two or more gauge discharge stations. This was important as it could facilitate calibrating and validating the model throughout rather than concentrating at particular portions. The stations used in calibration and validation are shown in Figure 4. F i g u r e 3 - 1 0 : L o c a t i o n m a p o f t h e g a u g e s t a t i o n s u s e d i n H E C R AS m o d e l 4. The discharge data was used as input in the HEC-RAS model of Ganges basin. Resulting output of the model was in the form of water surface elevation and depth of water in the channel. 5. Out of the seven events, five (1978, 1981, 1988, 1996, and 1998) were used in calibration and rest two (2004 and 2008) were used in validation. Output of the model for Final Report: Volume I Confidential Page 110 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia these events was matched with the observed CWC water levels at 35 gauge stations spread across the basin. The locations of these stations where variations were noted for modeled and actual water levels during calibration and validation are shown in Figure 5. The result showed little variability and did not match very well initially. Alteration of model parameters (Manning’s roughness coefficient and boundary conditions of channel) was required to bring a close match of results to observed data. Several iterations were carried out with changed parameters to attune the model. F i g u r e 3 - 1 1 : L o c a t i o n m a p o f t h e g a u g e s t a t i o n s u s e d i n c a l i b r a t i o n - va l i d a t i o n o f H E C R AS m o d e l 6. Model Calibration- The gauge stations had the best match between simulated and observed water levels. Further tweak in parameters did not affect the model result. The following is the summary of the calibration events: ï‚· 1978 Event A devastating flood occurred in River Yamuna in 1978. Many rural embankments were breached due to heavy discharge from its upstream Tajewala Headworks in Haryana. Rise in water levels also caused back flows in the connecting drains and had effect on the city drain network causing overflow resulting in damage to property, public utilities and agriculture. HFL at Delhi Railway Bridge was recorded as 207.15 m, 2.66 m above danger level of 204.49 m. Figure 6 shows the deviation plotted between simulated and observed water levels at various gauge stations for the calibration event. Final Report: Volume I Confidential Page 111 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 3 - 1 2 : D e vi a t i o n b e t w e e n s i m u l a t e d a n d o b s e r ve d w a t e r l e ve l s a t va r i o u s g a u g e s t a t i o n s f o r t h e c a l i b r a t i o n e ve n t o f 1 9 7 8 ï‚· 1981 Event A continuous heavy rainstorm in a short period caused exceptionally high rainfall and flood in arid Rajasthan in 1981, breaking a past eighty year record. The Morel dam was washed away causing floods in Bangagna, Bhadrawati and Borkheda rivers. Figure 7 shows the deviation plotted between simulated and observed water levels at various gauge stations for the calibration event. F i g u r e 3 - 1 3 : D e vi a t i o n b e t w e e n s i m u l a t e d a n d o b s e r ve d w a t e r l e ve l s a t va r i o u s g a u g e s t a t i o n s f o r t h e c a l i b r a t i o n e ve n t o f 1 9 8 1 ï‚· 1988 Event 1988 flood in Delhi recorded the second highest peak flow of 1,547.36 cumec. Continuous torrential rains in Northwest part of country caused heavy discharge in Yamuna. The flow in the river Yamuna at Delhi Railway Bridge was recorded above warning level i.e. 204 m. It was a high magnitude flood causing misery and loss of life and property to the residents of the city and damage to transport network. Figure 8 shows the deviation plotted between simulated and observed water levels at various gauge stations for the calibration event. Final Report: Volume I Confidential Page 112 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 3 - 1 4 : D e vi a t i o n b e t w e e n s i m u l a t e d a n d o b s e r ve d w a t e r l e ve l s a t va r i o u s g a u g e s t a t i o n s f o r t h e c a l i b r a t i o n e ve n t o f 1 9 8 8 ï‚· 1996 Event Intrusion of flash flood in North Bihar in 1996 was largely due to man-made structures like dams, embankments and roads. Gandak, Kosi, Sone and Bagmati rivers carried high discharges. The 1996 flood proved to be disastrous for Bihar as it crossed the bar of estimated Danger level (i.e. 48.60 m) and the flood discharge reached 57,600 cumec at Gandhighat, Patna. Figure 9 shows the deviation plotted between simulated and observed water levels at various gauge stations for the calibration event. F i g u r e 3 - 1 5 : D e vi a t i o n b e t w e e n s i m u l a t e d a n d o b s e r ve d w a t e r l e ve l s a t va r i o u s g a u g e s t a t i o n s f o r t h e c a l i b r a t i o n e ve n t o f 1 9 9 6 ï‚· 1998 Event Eastern Uttar Pradesh faced unprecedented flood disaster in 1998. Rivers Ghagra, Rapti, Gangdak, Narayani etc. caused as what was quoted the most devastating flood of the century. It was a quick flood and people were far from equipped to cope with it. In 1998, an embankment constructed to check the flooding at Gaighat, breached and the village experienced the worst flood in history. The resultant flash flood destroyed many structures and agriculture fields. Figure 10 shows the deviation plotted between simulated and observed water levels at various gauge stations for the calibration event. Final Report: Volume I Confidential Page 113 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 3 - 1 6 : D e vi a t i o n b e t w e e n s i m u l a t e d a n d o b s e r ve d w a t e r l e ve l s a t va r i o u s g a u g e s t a t i o n s f o r t h e c a l i b r a t i o n e ve n t o f 1 9 9 8 The stations with matched values were uniformly distributed across the basin and on all major rivers with an average variation of -0.03 m, which demonstrated acceptable behavior of the model. Hence, it could be concluded that model was adequately calibrated. 7. Model Validation- The two events of 2004 and 2008 were then used to validate the model. The following is the summary of the calibration events: ï‚· 2004 Event Floods in 2004 manifested the severity of the flood challenge when a vast area of Bihar state was badly affected by the floods of Bagmati, Kamla Balan and Adhwara group of rivers. Many embankments in North Bihar were breached which resulted in flood inundation in many areas, causing large scale devastation to life and property. For the first time, Ganges crossed the danger mark at Farakka Barrage. Figure 11 shows the deviation plotted between simulated and observed water levels at various gauge stations for the validation event. F i g u r e 3 - 1 7 : D e vi a t i o n b e t w e e n s i m u l a t e d a n d o b s e r ve d w a t e r l e ve l s a t va r i o u s g a u g e s t a t i o n s f o r t h e va l i d a t i o n e ve n t o f 2 0 0 4 Final Report: Volume I Confidential Page 114 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia ï‚· 2008 Event An appreciable amount of rainfall at the commencement of monsoon season followed by heavy rains in July and August led to the breach of Eastern Kosi Afflux embankment in Nepal. This turned into a catastrophe causing misery to lakhs of people in Nepal and India, particularly Bihar. This calamity also led Kosi river to change its course, with the spread of new course as wide as 15 km. Lack of preparedness for this flood disaster resulted in high casualties. The damage caused by 2008 Kosi (Ganges Basin) flood is the highest in five decades of flood history in Bihar. Figure 12 shows the deviation plotted between simulated and observed water levels at various gauge stations for the validation event. F i g u r e 3 - 1 8 : D e vi a t i o n b e t w e e n s i m u l a t e d a n d o b s e r ve d w a t e r l e ve l s a t va r i o u s g a u g e s t a t i o n s f o r t h e va l i d a t i o n e ve n t o f 2 0 0 8 The validation results were in line with the observed values. Thus this substantiated the model. 3.4.2.1 Historical Flood Extent RMSI has tried to compare the results of hydraulic model wherever historical flood information is available. Two such historical flood footprints were available and have been used for this purpose. The flood extent maps of the July 2004 and September 2008 events were available as shown in Figure 3-19 and Figure 3-20 respectively. During the July 2004 flood event, flood plains of Bagmati, Lower Ganges, Kamla Balan and Gandak sub basins were severely affected as shown in Figure 3-19. The dark blue and red areas in the map represent the flood-inundated areas. The flood maps represent the event between June 30, 2004 to July 27, 2004. Similarly, during the September 2008 flood event, flood plains of Kosi, Lower Ganges and Mahananda sub basins were severely affected as shown in Figure 3-20. The blue and red areas in the map represent the flood-inundated areas. This reported event is of September 02, 2008. The reported flooded areas in these imageries are not very accurate, as it has not included flood extents of all the nearby tributaries. These areas were compared with the corresponding historical year’s simulated flood extent. Final Report: Volume I Confidential Page 115 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 3 - 1 9 : F l o o d E x t e n t M a p o f J u l y, 2 0 0 4 f l o o d e ve n t ( S o u r c e : D F O ) Figure 3-20: Flood Extent Map of September, 2008 flood event (Source: DFO) Flood extent and flood depth maps have been generated by post-processing the simulated results of HEC-RAS in Arcview environment with HEC-GeoRAS extension. Flood plain boundaries and inundation depth data sets were generated from exported cross-sectional water surface elevations. For generating the flood extent map of this event, the HEC RAS Final Report: Volume I Confidential Page 116 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia model set up was used and peak flow values during the period of these events (July,2004 and September, 2008) were given as input. The Manning’s roughness coefficients were varied during the trials to match the simulated flood extents with observed flood extents. A comparison of simulated and observed flood extent for the July 2004 event has been made in Figure 3-21. Similarly, comparison of simulated and observed flood extent for the September 2008 event has been made in Figure 3-22. Both indicate that the simulated flood extent is in good agreement with the observed flood extents. F i g u r e 3 - 2 1 : C o m p a r i s o n b e t w e e n o b s e r ve d ( l e f t ) a n d m o d e l e d ( r i g h t ) f l o o d e x t e n t s o f J u l y, 2 0 0 4 e ve n t F i g u r e 3 - 2 2 : C o m p a r i s o n b e t w e e n o b s e r ve d ( l e f t ) a n d m o d e l e d ( r i g h t ) f l o o d e x t e n t s o f September, 2008 event The above calibration and validation results for various historical events indicate that the hydraulic model has been adequately validated. Therefore, this set up was then used to derive the flood extent maps for design events of 2, 5, 10, 25, 50, and 100 years. Final Report: Volume I Confidential Page 117 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 3.4.3 ESTIMATION OF RETURN PERIOD FLOWS 3.4.3.1 Annual peak discharge data: In addition to daily gauge-discharge data used in calibration and validation, the annual peak discharge data of 220 stations were also provided by CWC for the Indian part of the basin from 1959 onwards. Since Ganges basin encompasses entire Nepal and a significant fraction of rivers originate from there, peak daily discharge data of 98 gauge discharge stations starting from year 1962 in Nepal portion of Ganges basin from DHM, Nepal was also used. Data was suitably checked and analyzed to compute the probabilistic flows. Similar quality checks to identify outliers in data was carried out by comparing the value of a station with its long term mean. Also, only those stations, which had annual peak discharge data of 20 years or more were considered, since any data less than that might not ensure correctness in the statistical analysis of flow. Finally, data from 250 stations was used to carry out the analysis to simulate the probabilistic flows of 2, 5, 10, 25, 50 and 100-year return periods. 3.4.3.2 Return period flow estimation: Availability of flow data both for India as well as Nepal gave extra soundness to the model since actual data was available for most parts of the basin. Statistical analysis was performed on all these stations to estimate return period flows. The same return period flows were given as an input in the already validated hydraulic model without modifying any parameters. As stated already, there should be at least 20 years of continuous historical data to perform return period flow estimation. The stations having at least 20 years of data were finally used in the analysis to derive return period flows for 2, 5, 10, 25, 50, and 100-years. These station locations have already been shown in the previous section of the report. Return period flows have been estimated using two different distributions, namely, the Gumbel distribution and the Generalized Extreme Value (GEV) distribution. After comparing the results of the model using both the flows in the model with the available literature for Ganges basin, Gumbel distribution was further used for the estimation of the return period event generation. The L moments method was used for the estimation of the parameters for GEV and Gumbel distributions. L-Moments are based on probability-weighted moments (PWMs) and provide a greater degree of accuracy and ease. L-Moments are a modification of the PWMs, as they use the PWMs to calculate parameters that are easier to interpret and that can be used in the calculation of parameters for statistical distributions. L-Moments are based on linear combinations of data. They provide an advantage, as they are easy to work with, and more reliable as they are less sensitive to outliers. The method of L-Moments calculates more accurate parameters than the Method of Moments (MoM) technique (Kochanek, 2010). The MoM techniques only apply to a limited range of parameters, whereas L-Moments can be more widely used, and are nearly unbiased (Rowinski, 2001). The four L-Moments (λ1, λ2, λ3, λ4) are derived using the four PWMs. λ1 = L1 = M100 λ2 = L2 = 2M110 − M100 λ3 = L3 = 6M120 − 6M110 + M100 λ4 = L4 = 20M130 − 30M120 + 12M110 − M100 Final Report: Volume I Confidential Page 118 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Where M100, M110, M120, and M130 are the four probability-weighted moments, which are defined as 1 M100 = N ∑N i=1 Q i 1 (i−1) M110 = N ∑N i=1 (N−1) Q i 1 (i−1)(i−2) M120 = ∑N i=1 (N−1)(N−2) Q i N 1 (i−1)(i−2)(i−3) M130 = N ∑N i=1 (N−1)(N−2)(N−3) Q i In which N is the sample size, Q is the data value, and i is the rank of the value in ascending order. After Estimating the 4 L- Moments ((λ1, λ2, λ3, λ4), the Gumbel parameters α (scale parameter) and u (location parameter) can be obtained as 2 λ α = log 2 𝑢 = 𝜆1 − (í µí»¼í µí±?) In which c=0.58 (Euler’s Constant). In addition to above four moments, L-CV, L-Skewness and L-Kurtosis are used for the distribution fitting. L-CV is similar to the normal coefficient of variation (CV). The larger the CV value, the larger the variation of the data set from the mean. For example, in arid regions that receive few events, the variation will be large, as one storm will deviate greatly from the low mean. L2 Ï„2 = L1 (L − CV) L-Skewness is a measure of the lack of symmetry in a distribution. If the value is negative, the left tail is long compared with the right tail, and if the value is positive, the right tail is longer. For GEV frequency analysis, a positive L-Skewness value is desired, as we are interested in the extreme events that occur in the right side tail of the distribution. L3 Ï„3 = L2 (L − Skewness) L-Kurtosis is difficult to interpret, however, and is often described as the measure of “peakednessâ€? of the distribution (Hosking, 1997). L-kurtosis is much less biased than ordinary kurtosis. L4 Ï„4 = L2 (L − Kurtosis) Figure 3-23 shows the L-moment ratio diagram for stations used in the return period values estimation. Final Report: Volume I Confidential Page 119 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Figure 3-23: L moment ratio diagram The return period flow values were then used in the calibrated-validated HECRAS model without modifying any parameters to get flood extents and flood depths. 3.4.4 FLOOD HAZARD MAPPING FOR RETURN PERIOD FLOWS Flood hazard maps depicting flood extents and flood depths were derived by performing one-dimensional hydraulic routing through the river system for 2, 5, 10, 25, 50, and 100-year return period discharges. The derived flood extents can be used to determine the various exposures at risk. The flood hazard maps for 2, 5, 10, 25, 50, and 100-year return periods for the whole basin are shown in Figure 3-24 to Figure 3-29. Final Report: Volume I Confidential Page 120 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 3 - 2 4 : F l o o d h a z a r d m a p f o r 2 - ye a r r e t u r n p e r i o d f o r G a n g e s b a s i n F i g u r e 3 - 2 5 : F l o o d h a z a r d m a p f o r 5 - ye a r r e t u r n p e r i o d f o r G a n g e s b a s i n Final Report: Volume I Confidential Page 121 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 3 - 2 6 : F l o o d h a z a r d m a p f o r 1 0 - ye a r r e t u r n p e r i o d f o r G a n g e s b a s i n F i g u r e 3 - 2 7 : F l o o d h a z a r d m a p f o r 2 5 - ye a r r e t u r n p e r i o d f o r G a n g e s b a s i n Final Report: Volume I Confidential Page 122 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 3 - 2 8 : F l o o d h a z a r d m a p f o r 5 0 - ye a r r e t u r n p e r i o d f o r G a n g e s b a s i n F i g u r e 3 - 2 9 : F l o o d h a z a r d m a p f o r 1 0 0 - ye a r r e t u r n p e r i o d f o r G a n g e s b a s i n These flood hazard grids for each return period are used to estimate the damage/loss for each type of exposure described in the exposure section. Final Report: Volume I Confidential Page 123 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 4 Risk Assessment Risk assessment undertakes the assessment of losses of exposure (buildings, infrastructure and agriculture) and impact on demography due to various probabilistic flood events (e.g. 2, 5, 10, 25, 50, and 100-Year return period floods) using the vulnerability/damage functions. The methodology is divided into four modules namely i) hazard ii) exposure iii) damage/vulnerability, and iv) Loss module as shown in Figure 4-1. Figure 4-1: Basic building blocks of risk assessment methodolog y The hazard module constitutes development of probabilistic flood hazard maps, considering various flood hazard events based on their frequency and severity. In the exposure module (discussed in the Chapter 2), an inventory of the various exposure element such as residential, commercial and industrial buildings, agriculture, essential facilities, and infrastructural at risk are created. The photographs (Figure 4-2) show various types of buildings affected due to flood in the Ganges Basin. Final Report: Volume I Confidential Page 124 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 15 F i g u r e 4 - 2 : V a r i o u s t yp e s o f b u i l d i n g s a f f e c t e d d u e t o f l o o d i n G a n g e s B a s i n The building exposure is classified into seven classes of buildings, which has been discussed in detail in Exposure section. These classes are given in Table 4-1. Table 4-1: Classification of buildings based on the construction material Structural Sub Composition of housing material class Structural 1 Grass, Thatch, Bamboo, Wood, Mud, Plastic, etc. Structural 2 Mud/Unburnt Brick/Stone without mortar Structural 3 Light Metal Structural 4 Burnt Brick/ stone with mortar having temporary Roof (Tiles, wood, GI, slate, etc.) Structural 5 Masonry building /Reinforced concrete frame with brick infill Structural 6 RCC Structural 7 Any Other These building classes are defined in such a way that each type reflects a distinct vulnerability against flood. These categories are based on the construction material, roof- type, and the wall-types of buildings. The vulnerability module or damage functions give the relationship between behavior of exposure types with respect to the severity of the flood hazard. 15 http://www.itbhuglobal.org/chronicle/archives/2008/09/helping_victims.php http://archives.deccanchronicle.com/130816/news-current-affairs/gallery/monsoon-attack-india http://www.tehelka.com/theuttarakhandfloods/ Press Trust of India. Final Report: Volume I Confidential Page 125 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia The loss module computes the monitory losses using the information from the hazard, exposure, and vulnerability modules. The following section describes the vulnerability functions (flood depth vs damage relationship), used in this study to assess the risks. 4.1 Vulnerability functions Damage susceptibility associated with a given level of hazard is measured in terms of a mean damage ratio (MDR) defined as the expected proportion of the monetary value of repair needed to bring back the facility to pre-event condition, over the replacement value of the facility, as a consequence of the hazard. The curve that relates the MDR to the hazard is called a vulnerability function. Vulnerability functions are developed for various assets for flood, using analytical/synthetic and statistical methods complemented with expert engineering or heuristic judgment based on local and/or international experiences. Physical vulnerability refers to the degree to which an asset would undergo damage or be destroyed in a hazardous environment caused by catastrophic events. The vulnerability assessment involves quantifying the damage susceptibility of each asset class with respect to hazard parameters of the peril. 4.1.1 METHODOLOGY The physical vulnerability functions have been developed using an analytical approach complemented by engineering analyses along with expert judgment based on international experience. Figure 4-3 presents a flow chart indicating the interface between different components of vulnerability functions development. Proposed methodology for developing the vulnerability functions uses three different approaches i.e., damage statistics of past events, analytical/synthetic and/or engineering studies, and the international experience. F i g u r e 4 - 3 : D e ve l o p m e n t m e t h o d o l o g y f o r v u l n e r a b i l i t y f u n c t i o n s Final Report: Volume I Confidential Page 126 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia The following steps were undertaken as part of the methodology shown above: ï‚· Locality-specific damage data in terms of quantity and extent of damage is collected for historical events wherever available. This data, in general, is collected in a variety of text and descriptive formats. They are digitized by way of engineering interpretation of damage states and repair strategies. ï‚· Building stock from census in the region is reviewed and categorized into meaningful classes of typical buildings. ï‚· Building specific data is collected with emphasis on vulnerability parameters such as material of construction, age (to include practice and code enforcement), height (in form of number of stories), shape (in matters of cross-section), design characteristics (both architectural and structural), maintenance levels (including upgrades to codes), and reconstruction, wherever possible. ï‚· Special emphasis is given for collection of mortality rates, hospitalization by gender, age, physical condition, treatment, need for hospitalization, etc. during historical events. In addition to this, efforts are made to capture historical disaster information about number of people looking for shelter by age, gender, income, ethnicity, and duration, etc. In addition, efforts are made to understand the infrastructure of each of the three countries. ï‚· Literature survey is undertaken to extract and compile useful information on buildings, contents, infrastructure, and damage and vulnerability in the region. This includes resources such as technical papers, historical event damage reports, building code provisions (and degree of its enforcement), studies on social impact of disasters, gender behaviors and needs, etc. Note that data is also adopted into the study for regions with similar infrastructure and building performance, for purpose of enhancing the statistical convergence. ï‚· Information on damage mechanisms and behavior of different coverage classes in the region during past historical events; construction and design practices; code provisions and compliance levels; maintenance levels and regional trends, is also studied. ï‚· Physical vulnerability models implemented as part of open risk modeling platforms like HAZUS, CAPRA are also studied for suitability for the basin, for comparison and relativity reasons. 4.1.2 REVIEW OF VULNERABILITY (DEPTH – DAMAGE) FUNCTIONS The following sections detail the vulnerability (damage) functions studied to develop damage functions for this study. 4.1.2.1 Asian Institute of Technology Depth-Damage Functions The Geoinformatics Center at the Asian Institute of Technology (AIT) has developed depth damage functions for adobe, brick, and reinforced concrete structures in Nepal (Hazarika, 2006). These depth-damage functions were developed using post flood damage assessments of 1,514 adobe structures, 3,532 brick structures, and 339 reinforced concrete structures. The damage assessments did not include a separate content damage analysis. After reviewing these damage functions, it was determined that the occupancies as well as the building construction classes were similar to that of Ganges basin as Nepal falls in the Ganges basin. Moreover, the type of flooding included flash floods as well as longer duration floods. Thus, these functions were analyzed and found to be applicable for the present work. Final Report: Volume I Confidential Page 127 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 4 - 2 : S u i t a b i l i t y o f AI T d a m a g e - f u n c t i o n s When to use When not to use Numerous damage assessment were Only a couple of hundred damage completed for adobe and brick assessments were used for the RCC structures which makes those curves structures making that curve a little less more accurate known. This curve would be more accurate in This curve would not be as accurate in normal flood conditions. flash flood conditions Should be used for fresh water Should not be used for salt water flooding. flooding. 4.1.2.2 U.S. Army Corps of Engineers (USACE) Depth-Damage Functions A number of prolific USACE District depth damage functions have been collected from across the U.S. Functions (Figure 4-4) have been compiled for the Chicago and New Orleans districts. These functions were analyzed to determine their applicability to the study region. F i g u r e 4 - 4 : E x a m p l e U S AC E d e p t h – d a m a g e f u n c t i o n - o n e s t o r e y n o b a s e m e n t 4.1.2.2.1 Chicago District The Chicago District developed seven sets of generic structure and content damage functions to represent commercial, industrial, and public occupancies in conjunction with the 1996 Feasibility Study on the Upper Des Plaines River in northeast Illinois (IWR 85-R-5). These damage functions, based on models developed by the Baltimore and Galveston districts, classify structures as low, mid and high structure vulnerability, and low, mid and high contents vulnerability, resulting in seven curves representing the various ranking combinations. In addition, seven residential damage functions (1-story, 2-story, split-level with and without basement, and mobile home) were also provided. After reviewing these damage functions, it was determined that although the occupancies were similar to Ganges basin, the building construction types were different. Thus, these functions were not included in present study. 4.1.2.2.2 New Orleans District The New Orleans District has developed expert opinion damage functions (GEC, 1996), for a variety of depth-damage functions including residential and non-residential structures and contents damage for four types of flooding: Final Report: Volume I Confidential Page 128 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 1. Hurricane flooding, long duration (one week), salt water 2. Hurricane flooding, short duration (one day), salt water 3. Riverine or rainfall flooding, slow rising and slow receding, freshwater 4. Riverine or rainfall flooding, flash flood, freshwater In addition, non-residential contents damage functions are provided for a variety of occupancies: Eating/recreation, groceries/gas stations, multi-family residences, professional businesses, government facilities, retail stores, and warehouses. Structures are assumed to be “no basementâ€? structures and the reference point for water depth appears to be the top of the finished floor, based on review of detailed component loss tables. After reviewing these damage functions, it was determined that the occupancies were the similar to the study region and the building construction types were also similar (damage functions existed for adobe brick, stone, and reinforced concrete building types), the type of flooding included flash flood which is prominent in some sub basins of the study region. Table 4-3: Suitabilit y of the New Orleans District damage functions When to use When not to use Numerous damage assessment were completed for adobe, brick, stone, and RCC structures which makes those curves more accurate There are curves available for Modeling structures with a basement. flash flood and normal flood conditions. Curves are available for fresh and salt water flooding. 4.1.3 FINAL VULNERABILITY FUNCTIONS Damage functions for using the consensus from the above studies are adjusted for the present study. The studies are selected so that they have close relationships with building and construction practices in the Ganges basin. If these were not met, factors (assigned subjectively and based on engineering judgment) were introduced to update these damage functions. Following vulnerability curves (Figure 4-5) have been used to assess the building losses due to flood for all the seven structure types of various occupancy classes i.e. Residential, Commercial, Industrial, Educational Institutes, Health Facilities, and other houses. These seven classes of buildings have already been described in Table 4-1. Similarly, Figure 4-6 shows the depth-damage function used for infrastructure (railway and road). Final Report: Volume I Confidential Page 129 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Figure 4-5: Vulnerability function for buildings Figure 4-6: Vulnerability function for Infrastructure Final Report: Volume I Confidential Page 130 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia The vulnerability of agricultural crops due to flood hazard involves developing a relationship between crop production and hazard parameter (flood depth). Generally, with regard to crops, the flooding can be classified into two types: (a) Water-logging, in which root and some portion of the shoot goes under water, and (b) Complete submergence, where the whole plant goes under water. Water logging is a major problem for Wheat cultivation around the world and in USA where around 12% of cultivated soil is affected by excess water16. About 39-40% yield loss is recorded under water logged condition1718. Maize is also susceptible to water logging, which causes loss of yield in tropical and subtropical region. Fifteen percent of all maize growing areas of South-East Asia face water logging problem, which may lead to yield loss of a range about 25–30% annually19. Coefficient of damage functions at different flood height (Table 4-4), developed by Herath (2003)20, was used to generate vulnerability curve for Rice crop. Further, in order to generate vulnerability curve for Maize and Wheat, we applied the logic of sensitivity of Maize and Wheat to flood as compared to Rice in the absence of any concrete coefficient of damage functions for Maize and Wheat crops. As per the literature and working experiences, Maize and Wheat are almost twice and thrice more sensitive to flood as compared to Rice21. Hence, we multiplied the damage function of Rice by a factor of two and three to generate the vulnerability curves for Maize and Wheat crops respectively. Table 4-4: Coefficient of damage functions for Rice crop at different flood heights Coefficient of damage functions at different flood height Crop Name Depth: 0.2 m - 0.5 m Depth: 0.5 m - 1.0 m Depth: 1.0 m and above c1 c2 c1 c2 c1 c2 Rice crop 12.72 0.2197 13.77 0.2357 34.471 0.1251 2 D=c1h + c2h (where D is the damage percentage and h is the water level in meter) In the context of the above, the team developed flood vulnerability curves for maize, rice, and wheat crops (Figure 4-7). 16 Boyer, J. S. (1982). Plant productivity and environment (crop genetic improvement). Science, vol. 218, no. 4571, pp. 443–448. 17 Collaku, A. and Harrison, S. A. (2002). Losses in wheat due to water logging. Crop Science, vol. 42, no. 2, pp. 444–450. 18 Musgrave, M. E. and Ding, N. (1998). Evaluating wheat cultivars for water logging tolerance,â€? Crop Science, vol. 38, no. 1, pp. 90–97. 19 Rathore, T. R. and Warsi, M. Z. K. (1998). Production of maize under excess soil moisture (water nd logging) conditions in Proceedings of the 2 Asian Regional Maize Workshop PACARD, Laos Banos, Philippines. 20 Herath, S. (2003). Flood Damage Estimation of an Urban Catchment Using Remote Sensing and GIS. International Training Program on Total Risk Management. Pp. 55. http://www.adrc.asia/publications/TDRM2003June/10.pdf 21 Ahmed, F., Rafii, M.Y., Ismail, M.R., Juraimi, A.S., Rahim, H.A., Asfaliza, R. and Latif, M. A. (2013). Water logging Tolerance of Crops: Breeding, Mechanism of Tolerance, Molecular Approaches, and Future Prospects, BioMed Research International, vol. 2013, Article ID 963525, 10 pages, doi:10.1155/2013/963525. Final Report: Volume I Confidential Page 131 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Figure 4-7: Vulnerability function for Agriculture 4.2 Risk analysis - loss module Economic Loss estimation with respect to direct losses has been undertaken in this part for the whole study region. 4.2.1 METHODOLOGY Direct Loss is a function of the damage ratio, which is derived through the damage function (vulnerability curve), translated into currency loss by multiplying the damage ratio by the value at risk. L = MDR( j, h) * Value_At_Risk ( j ) Equation 1 where: MDR(j,h) = Mean Damage Ration for a exposure type ‘j’ at a specific hazard intensity ‘h’ Value_At_Risk(j) = Replacement cost of the exposure type ‘j’ For crop loss, the MDR and crop production in terms of economic value or crop yield, acreage and recent year prices for each crop, are used to compute direct loss. As discussed in the hazard section, a complete probabilistic flood event set of 2, 5, 10, 25, 50, 100-Years has been created, and a specific rate of occurrence is assigned to these events. Direct loss has been calculated for every event in the probabilistic set and for all types of exposure at risk like residential, commercial, industrial buildings, essential facilities, infrastructure, and agriculture. This is done for each asset class at each location where the treatment of location differs from asset class to asset class. Losses are then aggregated at administrative level of resolutions (district/sub-district) as required. Once the losses have been computed for every event, an Event Loss Table (ELT) is generated for every exposure type that has the total loss for that event and the associated rate of occurrence. Two most common types of outputs generated using an ELT are Average Annual Loss (AAL) and Loss Exceedance Probability Curves (LEC). A typical AAL curve is shown in Figure 4-8. AAL is calculated using the following equation: AAL(j) = âˆ‘í µí²? 𝒊=𝟎 𝑳(𝒊, 𝒋) ∗ 𝑹(𝒊) Equation 2 Final Report: Volume I Confidential Page 132 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia where: L(i,j) = Loss for event ‘i’ and exposure type ‘j’ R(i) = Rate of occurrence of event ‘i’ Figure 4-8: Sample of AAL LEC curve is the second output for risk analysis. LEC is a graphical representation of the probability that a certain level of loss is exceeded in a given time period (Figure 4-9). This is expressed in terms of either loss in production (mt) in case of agricultural losses or in monetary values for all other types of exposure. The abscissa of the LEC is loss, while the ordinate is the frequency or probability of loss (for most losses, probability and frequency are equivalent). Small losses occur frequently, and large losses rarely, so the curve slopes downward to the right. The probability of loss is obtained from the rate of occurrence and is calculated as shown below: EP( I ) = e power(∑𝒊 𝟎 𝑹𝒊) Equation 3 Where R is the rate of occurrence of an event Using the LEC losses are estimated for key return periods for all the hazards. Figure 4-9: Sample of LEC Next, GIS based risk maps showing AAL, and probable maximum losses (PML) for various key return periods have been generated showing the areas likely to get affected at the administrative level. Final Report: Volume I Confidential Page 133 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia In the present context, the direct loss is the hazard-induced loss in terms of financial losses for various structures, infrastructure, and agricultural crops based on current available valuations. The spatial distribution of the modeled risk outputs are portrayed in the form of maps showing the hazard, exposure, and risk characteristics. The temporal characteristics of the modeled risk outputs are depicted in the form of LEC. Based on the above approach, losses have been computed for all the exposure elements. As discussed in the exposure section, various exposure elements have been categorized into two broad categories: Aggregated Exposure – where the area and replacement cost of buildings representing the exposure type are summed at province/district level. Site Specific Exposure – where every asset in the exposure category is represented by a separate location (Longitude, Latitude) on surface of earth with replacement cost. The following sub sections describe, how the above-described approach has been applied for these broad exposure categories. 4.2.2 AGGREGATED EXPOSURE Aggregated exposure is treated differently for different hazards. The reason for differentiation is that hazards like earthquake are a regional hazard whereas flood is a site- specific hazard. Therefore, for flood hazard it is important to know in which part of the district the exposure lies. For this purpose, the aggregated exposure has already been further segregate on district/sub-district level for various types of exposure elements. For flood hazard, the losses are estimated at the district/sub-district level for various exposure types based on the area overlap of the district/sub-district polygon with the hazard grid. Figure 4-10 shows the area overlap of the district/sub-district with flood inundation area. The polygons with black outline represent the sub-district polygon and the filled polygons represent the flood depth grid. The losses are estimated only on the area of the polygon that is under the depth grid. F i g u r e 4 - 1 0 : D i s t r i c t / s u b - d i s t r i c t o ve r l a p w i t h f l o o d h a z a r d Final Report: Volume I Confidential Page 134 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 4.2.3 SITE SPECIFIC EXPOSURE Site-specific exposure includes line type exposure like roads and railway lines for the whole Ganges Basin. Line Type Exposure The treatment for line type exposure is different from aggregate exposure. Since line type exposure elements are spread over a long area, a single hazard value cannot be used to estimate the losses to them. Since a line type exposure element is made of a set of smaller segments, the loss is estimated at the centroid of every segment. The losses of all the segments are summed up to estimate the loss to the line exposure element. Figure 4-11 shows an example of the segments for Railways lines, overlaid with flood depth grid, at which the losses are estimated. Figure 4-11: Railwa y lines affected with flood hazard The generic equations for loss and AAL computation for line type exposure elements takes the following form: L(i, j, k) = ∑𝒎 í µí²?=𝟎 𝑴𝑫𝑹(𝒋, 𝒊, 𝒉) ∗ í µí±½í µí²‚í µí²?𝒖𝒆_𝑨𝒕_𝑹𝒊𝒔𝒌(𝒋, 𝒌, í µí²?) Equation 4 where: L(i, j, k) = The loss from event ‘I’ for exposure type ‘j’ and line element ‘k’ MDR(j,i,h) = Mean Damage Ration for exposure type ‘j’ at a hazard intensity ‘h’ for event i Value_At_Risk(j,k,l) = Replacement cost of the exposure type ‘j’ and line element ‘k’ and segment ‘l’ AAL at any location is calculated using the following equation: AAL(j, k) = âˆ‘í µí²? 𝒊=𝟎 𝑳(𝒊, 𝒋, 𝒌) ∗ 𝑹(𝒊) Equation 5 where: Final Report: Volume I Confidential Page 135 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia AAL(j, k) = AAL for exposure type ‘j’ and line element ‘k’ L(i,j,k) = Loss for event ‘i’ and exposure type ‘j’ and line element ‘k’ R(i) = Rate of occurrence of event ‘i’ 4.2.4 HISTORICAL AND MODELED FLOOD DAMAGE 4.2.4.1 Introduction Flood damage assessment and estimation of losses is an essential aspect of flood risk management. This assessment immensely supports policy analysis and forms basis for insurance against flood losses. Such kind of assessment also becomes an essential tool for the policy makers to understand the level of damages and losses which could be incurred from floods. Hence, the estimated flood losses need to be validated with the help of various authentic information sources to ascertain near accurate estimation of losses. From the available literature it has been observed, each selected flood model can give different results when applied to the same event (De Moeland Aerts, 2011; Jongman et al., 2012; Chatterton et al., 2014)22. This aspect quite common, because the models are often tailored to suit region-specific characteristics of flooding and considered assets of the region. These differences in results indicate that flood damage models are prone to large uncertainties. In the present study, HEC-RAS has been used for flood depth and flood extent mapping (Hydraulic modeling). These models have selected on the basis of their applicability for the basin and RMSI's previous experience. Flooding is a regular feature in many parts of Ganges basin. However, data on flooding in the basin is often found to be incomplete or inconsistent. Besides, floods in the basin have been assessed by a number of institutions from their own point of view using different data sets from a variety of sources. Both these issues make comparing of the model results with reported sources an uphill task. Nevertheless, the comparisons made in the following sub sections suggest that the modeled losses from the present study are in fair agreement with reported losses taken from various authentic sources. 4.2.4.2 Factors Influencing Flood Losses The flood damage and loss estimations from the present study and those from other studies are likely to be influenced by the following factors: Differences in input exposure data sets It is worth noting that the flood damages reported by various sources and used for validating the losses from this study may not necessarily have used identical or even remotely identical data sets and methods to estimate the distribution of houses and population, and in arriving at exposure values for agriculture, infrastructure etc. Therefore, any comparison should take into account these basic differences. The following paragraphs list some of such instances. In this study, the exposure data have been collected from various State and Central government organizations. The census data does not provide the details of the houses and population distribution at individual building level. It provides the aggregated values at sub- district/block level. Under this study, RMSI considered a homogeneous distribution of buildings and their replacement cost per sq. km area for all States of the basin. The distribution of population and other exposures have also been applied uniformly over all blocks/districts. The present study aimed to estimate flood losses due to major agricultural crops, buildings and infrastructure (roads and railway lines). RMSI considered three major crops i.e. rice, 22 www.nat-hazards-earth-syst-sci-discuss.net/3/.../nhessd-3-607-2015.pdf Final Report: Volume I Confidential Page 136 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia maize and wheat, while estimating the flood losses to the agriculture sector based on the most common cropping pattern. A homogeneous crop yield per hectare has been adopted for the entire block/district due to its similar agro-ecological situation. Complete infrastructure (roads and railways) data was not available for Nepal and Bangladesh. In addition, the entire minor road network of the study area has not been considered while estimating the total infrastructure losses for the basin. Besides, exposure like livestock, bridges, electric lines, airports, and pump stations have also not been considered in this study. Differences in level of aggregation The present study emphasizes on modeling losses on a basin-level. Therefore, sub-basin level losses that are extracted from the total basin level modeled losses may not match reported losses or losses modeled only at a sub basin level. For example, damages corresponding to Bagmati basin, extracted from the estimated damages for the entire Ganges basin and damages estimated only for the Bagmati basin may differ for the same flood event, due to different levels of estimation. Individual damage estimates for small floods or for local jurisdictions tend to be inaccurate. Damage estimates become more accurate at higher levels of aggregation. When estimates from many events are added together, the errors become proportionately smaller. Differences in types of losses included (direct/indirect losses) As mentioned above, the flood losses estimated in this study are related to losses due to major agricultural crops, buildings and infrastructure. Content losses and indirect losses, including losses due to business interruption, are not considered in this study. Hence, a variation may be observed when these losses are compared with the reported losses for the same flood event from any other information source. Differences in unit cost estimation Similarly, RMSI considered a homogeneous number of houses and replacement costs per sq. km area for all states. On ground, the cost of houses varies between cities and states. The housing costs could even differ from one area to another area in the same city/municipality. Similarly, average homogeneous wholesale prices for agriculture crops have been adopted for the entire Ganges basin for each crop. However, on ground the prices of crops change based on locations and seasons, which have not been accounted for in this study. At the time of flooding, the prices of vegetables and fruits may increase up to twice or thrice the market prices in the flood-affected cities. Such type of localized variations are difficult to factor in on a larger scale exercise, because of a paucity of micro-level, accurate data, which may finally influence the estimated from other studies. Uncertainties in people’s response Different flood events in the same area may lead to vary dissimilar level of damages due to diverse human responses and existing flood hazard warning systems as well as proactive actions taken by the concerned agencies. Differences in selected dynamic flooding parameters Potentially relevant flood event characteristics are maximum water depth, flood duration, flow velocity, pollution, warning time, and other possible aspects of the flood. In general, only the water depth is used in flood damage modeling, occasionally supplemented by one or two other parameters. Other possibly significant parameters, for example: flow velocity, flood duration, warning time and preparation, can influence the final outcome and could influence loss numbers. In this study, mean flood depth corresponding to various return period events have been used in damage assessment. Hence, a variation may be observed when these losses are compared with reported losses for the same flood event. Final Report: Volume I Confidential Page 137 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Differences in damage functions RMSI used homogeneous damage functions for various exposure types over the entire basin. However, these damage functions are likely to vary from one place to another place or in different topographic conditions, micro level damage data have not been accounted due to unavailability of damage functions at that level. These types of differences can lead to a variations in modeled and estimated losses when compared with the reported losses published by various authors at various levels on different occasions. 4.2.4.3 Validation of flood losses Keeping in view of the above factors, RMSI compared the reported losses with the estimated losses at various levels, namely country level, state level etc., The modeled losses in terms of numbers and values have been matched with those from reliable sources and research papers to provide a strong validation to the risk assessment analysis in this study. During this analysis it was observed some the losses are at lower side while others are at higher side. These differences are mainly attributed to adopted methodology, scale of exposure data, price variations over the time to name a few. A comparison of modeled losses with reported losses at various levels is provided in the following subsections. Affected Population The Ganges river basin is one of the most fertile and densely populated basins in the world. The river flows through 29 cities (with populations over 100,000), 23 cities (with populations between 50,000 and 100,000), and about 48 towns. It is, therefore, not surprising that millions of people suffer from severe floods almost every year in the basin. A comparison of modeled affected population with reported affected population for a set of events is given in the Table 4-5. Table 4-5: Validation for affected population Loss Validation: Affected Population Equivalent Sr. Event State/ Reported Modeled Deviation Reported Affected Return No Year Basin (Million) (Million) (%) Population Source Period 23 Eklavya Prasad, Nandan 1 1987 Bihar 28.20 28.13 -0.2% 100 Mukherjee, 2014 24 2 2002 Bihar 15.80 19.20 22% 10 BBA University, Lucknow 25 Flood Forecasting 3 2004 Bihar 29.99 28.12 -6% 50 Monitoring Directorate, CWC, GOI, 2012 26 National Institute of Disaster 4 2007 Bihar 22.00 25.60 16% 25 Management (NIDM), 2007 27 Ganga Flood Control 5 2010 Haryana 1.70 1.00 -41% 50 Commission (GFCC), 2010 Himachal Ganga Flood Control 6 2010 0.010 0.011 10% 50 Pradesh Commission (GFCC), 2010 28 Ganges Giriraj and Amarnath, IWMI, 7 2010 67.00 61.08 -8.8% 10 Basin 2010 23 https://cmsdata.iucn.org/downloads/situation_analysis_on_floods.pdf 24 https://ejournals.unm.edu/index.php/nsc/article/download/633/782 25 http://www.indiawaterportal.org/articles/state-wise-data-damage-caused-due-floods-during-1953- 2011-compilation-central-water 26 https://nidm.gov.in/PDF/pubs/Bihar%20Flood%202007.pdf 27 https://gfcc.bih.nic.in/Docs/GFCC-AR-2010-2011.pdf Final Report: Volume I Confidential Page 138 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia As given in Table 4-5, about 76 million (Giriraj and Amarnath, IWMI, 2010) people were affected in the entire Ganges basin, including Damodar basin. Out of this, 9 million people were affected in the Damodar basin. RMSI’s modeled affected population excluding Damodar basin comes to around 61 million. Also the comparison of affected population for flood years 1987, 2002, 2004 and 2007 for Bihar and flood year 2010 for Haryana and Himachal Pradesh have been given in Table 4-5. Figure 4-12 shows the comparison of thematic distribution of reported and modeled affected populations. The reported distribution for a 12-year recurrent flood closely matches RMSI’s modeled distribution for a 10-year return period event. Figure 4-12: Affected population reported b y Girir aj and Amarnath, IWMI (left) and RMSI’s modeled affected population (right) A separate validation of affected population for various flood events for 29Supaul district of Bihar and 30Burhi Gandak Basin has been shown in Figure 4-13 and Figure 4-14 respectively. To arrive at the modeled loss numbers of Burhi Gandak Basin, the losses of affected blocks in the Burhi Gandak Basin have been culled out. 28 http://www.un-spider.org/sites/default/files/7_GirirajAmarnath_IWMI.pdf 29 http://www.academia.edu/14897302/Kosi_Flood_Hazard_and_Disaster_Management_A_case_Stud y_of_Supaul_District 30 http://wrmin.nic.in/writereaddata/NWM_OR-FM-CC_2015_Vol-3.pdf Final Report: Volume I Confidential Page 139 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Figure 4-13: Validation of affected population for Supaul district Figure 4-14: Valida tion of affected population for Burhi Gandak sub -basin Affected Area Percentage distribution of flood-affected areas for 11 states lying in the Ganges basin have been reported in the GFCC annual report and shown in Figure 4-15. The state level distribution of maximum affected area estimated in RMSI’s modeled result have been shown in Figure 4-16, which indicate a close agreement with the distribution shown in Figure 4-15. Final Report: Volume I Confidential Page 140 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 4 - 1 5 : S t a t e w i s e a f f e c t e d a r e a r e p o r t e d b y G F C C a n n u a l r e p o r t o f ye a r 2 0 1 0 - 1 1 Figure 4-16: State wise maximum affected area from modeled result The validation summary for the modeled affected area is given Table 4-6. Affected areas due to flood events of 2002, 2004, 2007, and 2010 have been compared with available affected areas from various sources/organizations. Table 4-6: Validation for affected area Loss Validation: Affected Area Due to Flood Reported Modeled Equivalent Sr. Event Country/ Deviation Reported Area Category (Million (Million Return No Year State (%) Source Ha) Ha) Period 31 Bihar Kosi Flood Area Assessment Report 1 2002 affected due Bihar 1.97 1.72 -12.7% 10 by World Bank to flood (2010) 31 https://www.gfdrr.org/sites/gfdrr/files/publication/GFDRR_India_PDNA_2010_EN.pdf Final Report: Volume I Confidential Page 141 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Loss Validation: Affected Area Due to Flood Reported Modeled Equivalent Sr. Event Country/ Deviation Reported Area Category (Million (Million Return No Year State (%) Source Ha) Ha) Period 32 Area Disaster 2 2004 affected due Bihar 2.70 2.09 -22.6% 50 Management to flood Department, Bihar 33 Flood Forecasting Area Monitoring 3 2007 affected due Bihar 1.88 1.90 1.1% 25 Directorate, CWC, to flood GOI, 2012 Bihar Kosi Flood Area Assessment Report 4 2010 affected due Bihar 2.3 2.09 -9.1% 50 by World Bank to flood (2010) Percentage 34 Giriraj and area 5 2010 Nepal 0.98% 1.17% 19.4% 10 Amarnath, IWMI, affected due 2010 to flood Building Losses Lower Ganges sub-basin shows the maximum average annual losses in the Ganges River Basin. A summary for validation of the units of damaged residential buildings and corresponding modeled losses for Bihar state is are given in Table 4-7 and Table 4-8 respectively. Table 4-7: Validation for residential housing units Loss Validation: Damaged Residential Buildings Sr. Event Reported Modeled Deviation Equivalent State Reported Loss Source No Year (Millions) (Million) (%) Return Period 35 1 2002 Bihar 0.40 0.42 5% 10 BBA University, Lucknow Table 4-8: Validation for residential housing losses Loss Validation: Residential Building Losses Sr. Event Reported Modeled Deviation Equivalent State Reported Loss Source No Year (Million INR) (Million INR) (%) Return Period 36 Disaster Management 1 2002 Bihar 5,262 7,552 43% 10 Department Report , Bihar, 2004 Disaster Management 2 2004 Bihar 7,580 10,267 35.4% 50 Department Report , Bihar, 2004 37 National Institute of Disaster 3 2007 Bihar 8,422 8,708 3.4% 25 Management (NIDM), 2007 32 http://disastermgmt.bih.nic.in/Statitics/Statistics.htm 33 http://www.indiawaterportal.org/articles/state-wise-data-damage-caused-due-floods-during-1953- 2011-compilation-central-water 34 http://www.un-spider.org/sites/default/files/7_GirirajAmarnath_IWMI.pdf 35 https://ejournals.unm.edu/index.php/nsc/article/download/633/782 36 http://www.disastermgmt.bih.nic.in/Statitics/Final%20Report%202004.pdf Final Report: Volume I Confidential Page 142 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Loss Validation: Residential Building Losses Sr. Event Reported Modeled Deviation Equivalent State Reported Loss Source No Year (Million INR) (Million INR) (%) Return Period 38 Flood Forecasting 4 2007 Bihar 8,314 8,708 4.7% 25 Monitoring Directorate, CWC, GOI, 2012 As given in Table 4-8, modeled results for current exposure data, are 30-40% higher than the losses reported in the 2002, 2004, and 2007 flood events in Bihar state. These loss numbers, therefore, appear to be in fair agreement considering current exposure values. A separate validation of residential buildings losses for various flood events for 39Supaul district of Bihar and 40Burhi Gandak Basin has been shown in Figure 4-17 and Figure 4-18 respectively. Figure 4-17: Validation of residential buildings losses for Supaul district 37 https://nidm.gov.in/PDF/pubs/Bihar%20Flood%202007.pdf 38 http://www.indiawaterportal.org/articles/state-wise-data-damage-caused-due-floods-during-1953- 2011-compilation-central-water 39 http://www.academia.edu/14897302/Kosi_Flood_Hazard_and_Disaster_Management_A_case_Stud y_of_Supaul_District 40 http://wrmin.nic.in/writereaddata/NWM_OR-FM-CC_2015_Vol-3.pdf Final Report: Volume I Confidential Page 143 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Figure 4-18: Validation of residential buildings losses for Burhi Gandak Basin Average Annual Losses (AAL) A comparison of modeled AAL with reported AAL for Bihar and Uttar Pradesh states has been given in Table 4-9. The modeled AAL for Bihar is almost double as compared to the published AAL. The data up to year 2011 has been used in reported losses. The modeled AAL is higher due to corresponding increase in the exposure values for Bihar. Table 4-9: Validation for AAL Loss Validation: Average Annual Losses Reported Modeled Deviation Sr. No. State Reported Loss Source (Million INR) (Million INR) (%) Flood Forecasting Monitoring Directorate, 1 West Bengal 352 491 39.5% CWC, GOI, 2012 Flood Forecasting Monitoring Directorate, 2 Madhya Pradesh 77 117 53.0% CWC, GOI, 2012 Flood Forecasting Monitoring Directorate, 3 Bihar 2,427 3,083 27.0% CWC, GOI, 2012 Disaster Management Department, Uttar 4 Uttar Pradesh 4,320 4,390 1.6% Pradesh (http://rahat.up.nic.in/) Flood Forecasting Monitoring Directorate, 5 Uttar Pradesh 3,250 4,390 35.1% CWC, GOI, 2012 Agriculture Losses Districts in Bihar like Muzaffarpur, West and East Champaran, Katihar, and Begusarai are most prone to flooding. These districts have higher vegetable production compared to food grains production. Modeled Agriculture losses have been compared with reported losses for the 2007 flood event in Bihar state. The modeled agriculture losses for an equivalent return period of 25-years are estimated at about 1.07 times the reported losses for the same event. The summary of validation of agriculture losses for Bihar for the event is given in Table 4-10. Final Report: Volume I Confidential Page 144 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 4 - 1 0 : V a l i d a t i o n f o r Ag r i c u l t u r e l o s s e s Loss Validation: Agriculture Losses Sr. Event Reported Modeled Deviation Equivalent State Reported Loss Source No Year (Million INR) (Million INR) (%) Return Period 41 National Institute of 1 2007 Bihar 300 320 6.7% 25 Disaster Management (NIDM), 2007 A separate validation of agriculture losses for various flood events for 42Supaul district of Bihar and 43Burhi Gandak Basin has been shown in Figure 4-19 and Figure 4-20 respectively. Figure 4-19: Validation of agriculture losses for Supaul district Figure 4-20: Validation of agriculture losses for Burhi Gandak Basin 41 https://nidm.gov.in/PDF/pubs/Bihar%20Flood%202007.pdf 42 http://www.academia.edu/14897302/Kosi_Flood_Hazard_and_Disaster_Management_A_case_Stud y_of_Supaul_District 43 http://wrmin.nic.in/writereaddata/NWM_OR-FM-CC_2015_Vol-3.pdf Final Report: Volume I Confidential Page 145 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Public Property Losses The modeled public property losses have been compared with reported losses for the 2004 flood event in Bihar state. A variation of 0.4% has been observed in public property losses. The validation summary of public property losses is given in Table 4-11. Table 4-11: Validation for public property losses Loss Validation: Public Property Losses Sr. Event Reported Modeled Deviation Equivalent Reported Loss State No Year (Million INR) (Million INR) (%) Return Period Source 44 Flood Management 1 2004 Bihar 10,304 10,267 -0.4% 50 Information System, Bihar To summarize, the following factors can be attributed for the deviation in the estimated loss numbers with the reported ones: ï‚· The reported losses includes direct and indirect (all sectors including primary, secondary etc.) whereas RMSI losses are direct and pin pointed to sectors and specific areas mentioned in the report ï‚· Various types of exposure and crops have not been considered ï‚· Homogenous distribution of buildings and exposure over district/sub-district ï‚· Same unit replacement cost for all the states ï‚· Same wholesale price of the crop for all the sates ï‚· Accuracy and authenticity of the source of reported losses ï‚· Inflation and time value of money The following chapter illustrates the detailed estimated losses for various return period events. The losses are presented in the form of tables and maps depicting the losses in different ways to make the underlying risk easy to understand. 44 http://fmis.bih.nic.in/history.html Final Report: Volume I Confidential Page 146 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 5 Findings – Basin Level This section presents estimated losses and damages due to various probabilistic flood hazard events. These findings can be used by the concerned authorities and local district officials in understanding the potential impacts of floods and will be helpful in planning flood mitigation options. The findings here are summarized at a country level for India, Nepal, and Bangladesh and at sub-basin level for the entire Ganges Basin. The basin area lying in China could not be included in the risk estimation due to data limitations. 5.1 Demography The flood hazard grids for all return periods of 2, 5, 10, 25, 50, and 100-Years were overlaid on the sub-district/district boundary to calculate the percent affected area of each polygon due to all return periods. The same percentage has been applied on the demography to estimate the affected population due to various return period events. The process has been done for all sub-districts/districts of the three countries viz. India, Nepal, and Bangladesh and then summed up to give the affected population for the entire basin. Table 5-1 shows the number of affected males, females and total persons due to various return period flood events for the entire Ganges basin as well as the country-wise distribution. The analysis shows that the total number of persons affected due to flood in the basin varies from 47 million to 73 million for 2-year and 100-year return period floods respectively. These numbers are 10% and 16% of the total population of the Ganges Basin respectively. The same is depicted in the return period wise bar chart (Figure 5-1) for the entire basin. Figure 5-1 shows the total affected persons (%) in the Ganges basin. Table 5-1: Total number (in millions) of persons affected due to flood in Ganges B a s i n f o r va r i o u s r e t u r n p e r i o d f l o o d e ve n t s Demography Male Female Total Male Female Total Male Female Total Return Period 2-Year 5-Year 10-Year Entire Ganges Basin 24.70 22.56 47.26 29.53 26.98 56.50 31.96 29.19 61.15 Bangladesh 0.71 0.72 1.42 0.80 0.81 1.61 0.83 0.85 1.68 India 23.78 21.62 45.40 28.50 25.92 54.42 30.88 28.08 58.96 Nepal 0.21 0.22 0.43 0.23 0.25 0.48 0.24 0.26 0.50 Return Period 25-Year 50-Year 100-Year Entire Ganges Basin 34.98 31.97 66.95 37.71 34.48 72.19 38.33 35.04 73.37 Bangladesh 0.88 0.89 1.77 0.95 0.97 1.92 1.00 1.01 2.02 India 33.84 30.80 64.64 36.48 33.22 69.70 37.03 33.71 70.75 Nepal 0.26 0.28 0.54 0.28 0.30 0.58 0.29 0.31 0.61 Final Report: Volume I Confidential Page 147 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 5 - 1 : T o t a l n u m b e r o f p e r c e n t a g e a f f e c t e d p e r s o n s i n G a n g e s B a s i n f o r va r i o u s return period flood events The Lower Ganges Sub-basin has the maximum share of affected population (21%) in the basin whereas Ken sub-basin has a mere contribution of 0.3% due to a100-year return period flood event. Figure 5-2 depicts the percentage distribution of total affected persons for various sub-basins due to a 100-year return period flood. F i g u r e 5 - 2 : S u b - b a s i n l e ve l d i s t r i b u t i o n o f p e r c e n t a g e a f f e c t e d p e r s o n s i n G a n g e s B a s i n f o r 1 0 0 - ye a r r e t u r n p e r i o d f l o o d e ve n t The aggregated number of total persons affected for the three most severely affected sub- basins of Lower Ganges, Bagmati, and Kosi are given in Table 5-2. Sub-basin wise details of affected persons are provided in the next chapter. The district/sub-district level affected populations due to 100-year return period flood are shown thematically in Figure 5-3. Similar Final Report: Volume I Confidential Page 148 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia thematic maps showing affected persons for all return period flood events are provided in the Ganges Basin Risk Atlas Report. Table 5-2: Total number (in thousands) of persons affected due to flood in Ganges B a s i n f o r va r i o u s r e t u r n p e r i o d s f l o o d e ve n t s – L o w e r G a n g e s , B a g m a t i , a n d K o s i subbasins Sub-basin: Lower Ganges Sub-basin: Bagmati Sub-basin: Kosi Return Period Male Female Total Male Female Total Male Female Total 2-Year 5,395 4,903 10,298 1,181 1,065 2,246 644 585 1,229 5-Year 6,247 5,676 11,924 1,382 1,245 2,627 718 652 1,371 10-Year 6,643 6,034 12,677 1,556 1,402 2,958 762 692 1,454 25-Year 7,129 6,476 13,605 1,688 1,521 3,208 829 752 1,582 50-Year 7,781 7,070 14,851 1,877 1,690 3,567 913 829 1,742 100-Year 7,807 7,096 14,903 1,879 1,693 3,572 915 831 1,747 Figure 5-3: Spatial distribution of total affected persons (in thousands) for a 100 - Year Return Period Flood Demography data for India has additional details of children (age 0-6 years), and Scheduled Caste (SC) and Scheduled Tribe (ST) residing in the Indian part of the Ganges Basin. Similar analysis has been done to calculate the total number of affected children, SCs and STs due to various return periods events. Table 5-3 shows the total number of affected children, SCs and STs in the Indian part due to various return period flood events. Final Report: Volume I Confidential Page 149 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 5 - 3 : T o t a l n u m b e r ( i n m i l l i o n s ) o f c h i l d r e n ( a g e 0 - 6 ye a r s ) , S c h e d u l e d C a s t e s (SCs) and Scheduled Tribes (STs) affected due to floods in the Indian part of Ganges B a s i n f o r va r i o u s r e t u r n p e r i o d s Demography Male Female Total Male Female Total Male Female Total Return Period 2-Year 5-Year 10-Year Children (Age<6 year) 3.86 3.53 7.39 4.62 4.23 8.85 5.01 4.58 9.59 SC 3.98 3.62 7.60 4.76 4.33 9.10 5.16 4.70 9.86 ST 0.32 0.30 0.62 0.38 0.36 0.74 0.40 0.38 0.78 Return Period 25-Year 50-Year 100-Year Children (Age<6 year) 5.50 5.03 10.53 5.94 5.44 11.38 6.03 5.51 11.54 SC 5.66 5.15 10.81 6.12 5.57 11.69 6.22 5.67 11.89 ST 0.45 0.42 0.87 0.48 0.45 0.93 0.48 0.46 0.94 5.2 Economic Losses The direct losses (in INR) to various exposure classes, namely, buildings (residential, commercial, and industrial), essential facilities (education institute and health facility), Infrastructure (road and rail networks), and agriculture (rice, wheat, and maize) have been summarized in this and following sections. The total probable maximum loss (PML), combining losses for all the above exposure classes for the entire Ganges Basin, varies from INR 71,962 million to INR 1,42,486 million due to 2-year and 100-year return period flood events respectively. Similarly, the Average Annual Loss (AAL) of all the exposure classes for the entire Ganges basin is estimated around INR 42,342 million. Estimated PML and AAL for various exposure classes for the entire Ganges Basin due to various return period flood events are shown in Table 5-4. T a b l e 5 - 4 : P M L a n d A A L d u e t o f l o o d e ve n t s o f va r i o u s r e t u r n p e r i o d s f o r va r i o u s exposure classes Losses (Million INR): Ganges Basin Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Entire Ganges Basin 71,962 92,343 104,949 121,096 137,496 142,486 42,342 Residential 46,537 60,134 68,526 79,443 90,489 93,842 27,518 Commercial 17,558 22,371 25,422 29,202 33,113 34,325 10,288 Industrial 423 550 629 727 814 858 251 Building Education 109 140 159 184 210 218 64 Health 86 112 128 148 168 175 51 Others 1,153 1,469 1,666 1,910 2,161 2,243 675 Road 353 469 552 664 795 840 216 Infrastructure Railway 256 349 416 506 608 632 159 Rice 2,009 2,440 2,682 3,005 3,375 3,447 1,136 Agriculture Wheat 3,147 3,905 4,329 4,816 5,221 5,360 1,797 Maize 330 403 441 490 541 547 186 Damage to buildings due to floods is one of the main contributors to flood loss. The risk assessment analysis for various types of buildings shows that residential buildings bear the maximum loss followed by commercial buildings. Among all the exposure classes, residential building have the maximum AAL of INR 27,518 million that is 65% of total AAL on a basin level. The main reason for high flood losses in residential buildings is their large Final Report: Volume I Confidential Page 150 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia share of around 59% in total building exposure. Commercial buildings and wheat crop are the second and third biggest contributors in total AAL with 24% and 4% contributions respectively. The exposure type wise percentage distribution of total AAL is shown in Figure 5-4. F i g u r e 5 - 4 : P e r c e n t a g e d i s t r i b u t i o n o f A A L f o r va r i o u s a s s e s t s c l a s s A loss exceedance curve (LEC) has been generated using PML and their respective exceedance probabilities for all return period flood events for the entire basin. Figure 5-5 shows the LEC for the entire basin considering flood losses to all the exposure types. F i g u r e 5 - 5 : E x c e e d a n c e p r o b a b i l i t y c u r ve s h o w i n g t h e t o t a l l o s s e s d u e t o f l o o d f o r a l l e x p o s u r e t yp e s – G a n g e s B a s i n Final Report: Volume I Confidential Page 151 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 5.2.1 BUILDINGS As discussed in the preceding section, damage to buildings due to floods is one of the main contributors to flood loss. The risk assessment analysis for various types of buildings shows that residential buildings bear the maximum loss followed by commercial buildings. The average annual loss (AAL) of buildings for the entire Ganges basin is estimated around INR 38,848 million, out of which residential and commercial buildings contribute about 71% and 26% loss values respectively. Table 5-5 shows the AAL for buildings due to flood for various occupancy types and the percentage distribution is shown in Figure 5-6. Details of each occupancy type are provided in the following sections. T a b l e 5 - 5 : A A L d u e t o f l o o d f o r va r i o u s o c c u p a n c y t yp e s Occupancy Type AAL (Million INR) Residential 27,518 Commercial 10,288 Industrial 251 Educational Institutes 64 Health Facilities 51 Others 675 Total 38,848 F i g u r e 5 - 6 : P e r c e n t a g e d i s t r i b u t i o n o f A A L f o r va r i o u s o c c u p a n c y t yp e s 5.2.1.1 Residential As stated in the previous section, residential buildings have the maximum share of 70.8% of total building AAL on a basin level. The main reason for high flood losses in residential buildings is their large share of around 59% in total building exposure. On a basin level, residential buildings incur a total PML of around INR 93,842 million due to a 100-year return period flood event and AAL of INR 27,518 million. Estimated PML and AAL for residential buildings for the entire Ganges Basin and across all three countries due to various return period flood events are shown in Table 5-6. Final Report: Volume I Confidential Page 152 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 5 - 6 : P M L a n d A A L d u e t o f l o o d e v e n t s o f va r i o u s r e t u r n p e r i o d s f o r b u i l d i n g s - Residential Buildings Loss (Million INR): Residential Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Entire Ganges Basin 46,537 60,134 68,526 79,443 90,489 93,842 27,518 Bangladesh 682 785 847 931 1,047 1,099 374 India 45,802 59,290 67,614 78,443 89,364 92,661 27,116 Nepal 52 59 64 70 78 82 28 A loss exceedance curve (LEC) has been generated using PML and their respective exceedance probabilities for all return period flood events for residential buildings. Figure 5-7 shows the LEC for the entire basin considering flood losses to residential buildings. F i g u r e 5 - 7 : E x c e e d a n c e p r o b a b i l i t y c u r ve s h o w i n g t h e t o t a l l o s s e s d u e t o f l o o d f o r Buildings – Residential Similarly, losses are aggregated at each of the 18 sub-basins of the Ganges basin. Sub- basin wise AAL are shown in Figure 5-8. Yamuna sub-basin accounts for a maximum of 30% of residential building AAL followed by the Lower Ganges sub-basin with 27%. The Yamuna and Lower Ganges sub-basins records the highest losses because they have the maximum share in base exposure and also experiences flooding more frequently. The same is portrayed spatially by the AAL map (Figure 5-9) for the entire basin. Final Report: Volume I Confidential Page 153 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 5 - 8 : S u b - b a s i n l e ve l A A L ( m i l l i o n I N R ) f o r B u i l d i n g s – R e s i d e n t i a l The district/sub-district level AAL is joined with the administrative boundary and shown thematically in Figure 5-9. Similar thematic maps showing PML for all return period flood events are provided in the Ganges Basin Risk Atlas. Sub-basin wise details are provided in the next chapter. Figure 5-9: AAL (million INR) due to flood for Buildings – Residential Final Report: Volume I Confidential Page 154 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 5.2.1.2 Commercial As discussed earlier, commercial buildings have the second highest contribution of 26.5% to total building AAL on a basin-level. Commercial buildings incur high flood losses because they constitute the second highest share (37%) of the total building exposure. Commercial buildings for the entire Ganges Basin incur a PML of around INR 34,325 million due to a 100-year return period flood event and an AAL of INR 10,288 million. Estimated PML and AAL for commercial buildings for the entire Ganges Basin and across all three countries due to various return period flood events are shown in Table 5-7. T a b l e 5 - 7 : P M L a n d A A L d u e t o f l o o d e ve n t s o f va r i o u s r e t u r n p e r i o d s f o r b u i l d i n g s - Commercial Buildings Loss (Million INR): Commercial Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Entire Ganges Basin 17,558 22,371 25,422 29,202 33,113 34,325 10,288 Bangladesh 1,093 1,266 1,368 1,504 1,699 1,784 601 India 16,260 20,878 23,812 27,439 31,132 32,244 9,577 Nepal 205 228 242 258 282 296 110 An LEC has been generated using PML and their respective exceedance probabilities for all return period flood events for commercial buildings. Figure 5-10 shows the LEC for the entire Ganges basin considering flood losses to commercial buildings. F i g u r e 5 - 1 0 : E x c e e d a n c e p r o b a b i l i t y c u r ve s h o w i n g t h e t o t a l l o s s e s d u e t o f l o o d f o r Buildings – Commercial Similarly, losses are aggregated at each of the 18 sub-basins of Ganges basin. Sub-basin wise AAL are shown in Figure 5-11. The Lower Ganges sub-basin accounts for a maximum of 19% of commercial building AAL followed by the Yamuna (18%) and Middle Ganges (15%) sub-basins. The Lower Ganges sub-basin records the highest losses because it experiences flooding more frequently due to its location in the most low-lying down-stream Final Report: Volume I Confidential Page 155 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia area of the Ganges basin even though it has a share of only 7% in total commercial exposure. The same is portrayed spatially by the AAL map (Figure 5-12) for the entire basin. Similar thematic maps showing PML for all return period flood events are provided in the Ganges Basin Risk Atlas. Sub-basin wise detailed discussion is provided in the next chapter. F i g u r e 5 - 1 1 : S u b - b a s i n l e ve l A A L ( m i l l i o n I N R ) f o r B u i l d i n g s – C o m m e r c i a l Figure 5-12: AAL (million INR) due to flood for Buildings – Commercial Final Report: Volume I Confidential Page 156 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 5.2.1.3 Industrial Though industrial buildings are an important part of building exposure, they have a share of only 0.6% of total building AAL for the entire Ganges Basin. The main reason for such low flood losses for Industrial buildings is their low share of around 0.74% in the total building exposure. Industrial buildings incur a PML of around INR 858 million due to a 100-year return period flood event and a total AAL of INR 251 million for the entire Ganges Basin. Estimated PML and AAL for Industrial buildings for the entire Ganges Basin and across all three countries due to various return period flood events are shown in Table 5-8. T a b l e 5 - 8 : P M L a n d A A L d u e t o f l o o d e ve n t s o f va r i o u s r e t u r n p e r i o d s f o r b u i l d i n g s - Industrial Buildings Loss (Million INR): Industrial Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Entire Ganges Basin 423.40 549.94 629.46 726.93 813.98 857.55 251.03 Bangladesh 27.23 31.56 34.09 37.50 42.36 44.48 14.99 India 391.05 512.70 589.33 683.00 764.59 805.69 233.31 Nepal 5.11 5.68 6.03 6.43 7.03 7.38 2.74 A loss exceedance curve (LEC) has been generated using PML and their respective exceedance probabilities for all return period flood events for Industrial buildings. Figure 5-13 shows the LEC for the entire Ganges basin considering flood losses for Industrial buildings. F i g u r e 5 - 1 3 : E x c e e d a n c e p r o b a b i l i t y c u r ve s h o w i n g t h e t o t a l l o s s e s d u e t o f l o o d f o r Buildings – Industrial Similarly, losses are aggregated at each of the 18 sub-basins of Ganges basin. Sub-basin wise AAL are shown in Figure 5-14. Yamuna sub-basin accounts for a maximum of 44% of Industrial building AAL followed by Lower Ganges (14%) and Middle Ganges (11%) sub- basins. The Yamuna sub-basin incurs highest losses because it has the maximum share of around 33% in total Industrial exposure. The same is portrayed spatially by the AAL map Final Report: Volume I Confidential Page 157 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia (Figure 5-15) for the whole basin. Similar thematic maps showing PML for all return period flood events are provided in the Ganges Basin Risk Atlas. Sub-basin wise detailed discussion is provided in the next chapter. F i g u r e 5 - 1 4 : S u b - b a s i n l e ve l A A L ( m i l l i o n I N R ) f o r B u i l d i n g s – I n d u s t r i a l Figure 5-15: AAL(thousand INR) due to flood for Buildings – Industrial Final Report: Volume I Confidential Page 158 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 5.2.2 EDUCATIONAL INSTITUTIONS Educational Institution buildings, such as schools, colleges, etc., play a crucial part at the time of any disaster as shelters or relief centers. Thus, they come under essential facilities. It has a share of only 0.2% of total building AAL for the entire Ganges Basin. The main reason for such a low contribution to flood loss is the lower share of around 0.21% in the total exposure of buildings. Educational institution buildings incur a PML of around INR 218 million due to a 100-year return period flood event and have a total AAL of around INR 64 million for the entire Ganges Basin. Estimated PML and AAL for Educational Institution buildings for the entire Ganges Basin and across all three countries due to various return period flood events are shown in Table 5-9. T a b l e 5 - 9 : P M L a n d A A L d u e t o f l o o d e ve n t s o f va r i o u s r e t u r n p e r i o d s f o r b u i l d i n g s - Educational Institutions Buildings Loss (Million INR): Educational Institute Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Entire Ganges Basin 109.27 140.16 159.47 184.36 210.14 217.83 64.33 Bangladesh 5.90 6.79 7.32 8.04 9.05 9.62 3.24 India 102.92 132.86 151.59 175.71 200.41 207.50 60.85 Nepal 0.45 0.52 0.56 0.61 0.68 0.71 0.25 A loss exceedance curve (LEC) has been generated using the PML losses and their respective exceedance probabilities for all return period flood events for Educational Institution buildings. Figure 5-16 shows the LEC for the whole Ganges basin considering flood losses to Educational Institution buildings. F i g u r e 5 - 1 6 : E x c e e d a n c e p r o b a b i l i t y c u r ve s h o w i n g t h e t o t a l l o s s e s d u e t o f l o o d f o r Buildings - Educational Institutions Similarly, losses are aggregated at each of the 18 sub-basins of Ganges basin. Sub-basin wise AAL are shown in Figure 5-17. Lower Ganges sub-basin accounts for a maximum of Final Report: Volume I Confidential Page 159 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 26% of Educational institutions AAL followed by the Yamuna (22%) and Middle Ganges (12%) sub-basins. Though the Lower Ganges sub-basin has a share of only 11% in total educational institution building exposure, it incurs the highest losses because it experiences more frequent floods due to its location at the most low-lying down-stream area of the Ganges basin. The same is portrayed spatially by the AAL map (Figure 5-18) for the whole basin. Similar thematic maps showing PML for all return period flood events are provided in the Ganges Basin Risk Atlas. Sub-basin wise detailed discussion is provided in the next chapter. F i g u r e 5 - 1 7 : S u b - b a s i n l e ve l A A L ( m i l l i o n I N R ) f o r B u i l d i n g s – E d u c a t i o n a l Institutions Final Report: Volume I Confidential Page 160 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Figure 5-18: AAL (thousand INR) due to flood for Buildings – Educational Institutions 5.2.3 HEALTH FACILITIES Health facility buildings such as hospitals, clinics, etc. also play a very crucial role at the time of any disaster for medical aid and as emergency shelters and come under the essential facilities category. Health facility buildings have a total share of 0.1% in total building AAL for the entire Ganges Basin. They contribute insignificantly to total flood losses since their share is only around 0.12% of total building exposure. Health Facility buildings incur a PML of around INR 175 million due to a 100-year return period flood event and has an AAL of around INR 51 million for the entire Ganges Basin. Estimated PML and AAL for Health Facility buildings for the whole basin and across all three countries due to various return period flood events are shown in Table 5-10. T a b l e 5 - 1 0 : P M L a n d A A L d u e t o f l o o d e ve n t s o f va r i o u s r e t u r n p e r i o d s f o r b u i l d i n g s - Health Facilities Buildings Loss (Million INR): Health Facility Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Entire Ganges Basin 86.28 111.94 128.01 148.50 168.34 175.49 51.17 Bangladesh 1.45 1.67 1.80 1.98 2.22 2.33 0.79 India 84.72 110.15 126.08 146.37 165.96 172.99 50.31 Nepal 0.11 0.13 0.14 0.15 0.17 0.18 0.06 Final Report: Volume I Confidential Page 161 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia A loss exceedance curve (LEC) has been generated using PML and their respective exceedance probabilities for all return period flood events for Health facility buildings. Figure 5-19 shows the LEC for the whole Ganges basin considering flood losses to health facility buildings. F i g u r e 5 - 1 9 : E x c e e d a n c e p r o b a b i l i t y c u r ve s h o w i n g t h e t o t a l l o s s e s d u e t o f l o o d f o r Buildings - Health Facilities Similarly, losses are aggregated at each of the 18 sub-basins of Ganges basin. Sub-basin wise AAL are shown in . Yamuna sub-basin accounts for 33% of health facility building AAL followed by Lower Ganges (23%) and Middle Ganges (12%) sub-basins. The Yamuna sub- basin has the highest losses because it has a maximum share of around 31% in total Health Facility building exposure followed by the Lower Ganges sub basin (10%). The same is portrayed spatially by the AAL map (Figure 5-21) for the whole basin. Similar thematic maps showing PML for all return period flood events are provided in the Ganges Basin Risk Atlas. Sub-basin wise detailed discussion is provided in the next chapter. Final Report: Volume I Confidential Page 162 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 5 - 2 0 : S u b - b a s i n l e ve l A A L ( m i l l i o n I N R ) f o r B u i l d i n g s – H e a l t h F a c i l i t i e s Figure 5-21: AAL (thousand INR) due to flood for Buildings – Health Facilities Final Report: Volume I Confidential Page 163 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 5.2.4 OTHER BUILDINGS Other buildings have a contribution of 1.7% in the total building AAL for the entire Ganges Basin. The main reason for low flood losses for these buildings is their lower share of around 2.5% in the total building exposure. Other buildings for the entire Ganges Basin incur a PML of around INR 2243 million due to a 100-year return period flood event and have an AAL of around INR 675 million. Estimated PML and AAL for other buildings for the entire Ganges Basin and across all three countries due to various return period flood events are shown in Table 5-11. T a b l e 5 - 1 1 : P M L a n d A A L d u e t o f l o o d e ve n t s o f va r i o u s r e t u r n p e r i o d s f o r b u i l d i n g s - Others Buildings Loss (Million INR): Others Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Entire Ganges Basin 1,153.13 1,468.79 1,665.76 1,910.43 2,161.48 2,242.58 675.01 Bangladesh 73.52 85.19 92.03 101.22 114.35 120.07 40.46 India 1,065.81 1,368.27 1,557.44 1,791.85 2,028.16 2,102.59 627.16 Nepal 13.81 15.33 16.29 17.36 18.97 19.92 7.39 A loss exceedance curve (LEC) for other buildings has been generated using the PML losses and their respective exceedance probabilities for all return period flood events. Figure 5-22 shows the LEC for the entire Ganges basin considering flood losses for these building. F i g u r e 5 - 2 2 : E x c e e d a n c e p r o b a b i l i t y c u r ve s h o w i n g t h e t o t a l l o s s e s d u e t o f l o o d f o r Buildings – Others Similarly, losses are aggregated at each of the 18 sub-basins of the Ganges basin. Sub- basin wise AAL are shown in Figure 5-23. Lower Ganges sub-basin accounts for 21% of Other buildings AAL followed by Yamuna (19%) and Middle Ganges (15%) sub-basins. Though the Lower Ganges sub-basin has a share of only 9% of total building exposure, it shows the highest losses because of frequent floods due to its location at the most low-lying Final Report: Volume I Confidential Page 164 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia down-stream area of the Ganges basin. The same is portrayed spatially by the AAL map (Figure 5-24) for the whole basin. The district/sub-district level AAL is joined with the admin boundary and shown thematically in Figure 5-24. Similar thematic maps showing PML for all return period flood events are provided in the Ganges Basin Risk Atlas. Sub-basin wise detailed discussion is provided in the next chapter. F i g u r e 5 - 2 3 : S u b - b a s i n l e ve l A A L ( m i l l i o n I N R ) f o r B u i l d i n g s – O t h e r s Figure 5-24: AAL (million INR) due to flood for Buildings – Others Final Report: Volume I Confidential Page 165 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 5.2.5 INFRASTRUCTURE This section describes the risk assessment for Infrastructure (road and rail networks) due to various return period flood events of 2, 5, 10, 25, 50, and 100-Years. 5.2.5.1 Road Network The road network is one of the important parts of infrastructure exposure. It plays a very crucial role at the time of any disaster for providing relief and medical aid to flood affected areas. Therefore, it becomes crucial to know the kind damages and losses incurred by them due to flood events of various return periods. They also come under essential facility. Total flood affected road length varies between 3,158 km to 5,594 km for 2-year and 100-year return period flood events respectively, which is around 3.5% and 6.2% of total road length in the Ganges Basin. Table 5-12 shows the affected road lengths due to various return period flood events. T a b l e 5 - 1 2 : Af f e c t e d l e n g t h ( k m ) d u e t o f l o o d e ve n t s o f v a r i o u s r e t u r n p e r i o d s f o r Infrastructure – Road network Infrastructure affected length (km): Road Network Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year Entire Ganges Basin 3,158 3,922 4,360 4,938 5,470 5,594 Bangladesh 39 46 52 56 65 68 India 2,973 3,717 4,141 4,708 5,214 5,325 Nepal 146 160 166 175 191 201 On a basin level, roads incur a PML of around INR 840 million due to a 100-year return period flood event and an AAL of around INR 216 million. Estimated PML and AAL for roads for the entire Ganges Basin and for all three countries due to various return period flood events are shown in Table 5-13. T a b l e 5 - 1 3 : P M L a n d A A L d u e t o f l o o d e ve n t s o f va r i o u s r e t u r n p e r i o d s f o r Infrastructure – Road network Infrastructure Loss (Million INR): Road Network Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Entire Ganges Basin 353.49 469.18 551.72 663.88 795.39 839.67 215.70 Bangladesh 4.33 5.09 5.90 6.69 8.55 8.98 2.47 India 320.33 433.28 513.84 623.75 751.13 793.21 198.18 Nepal 28.83 30.80 31.97 33.44 35.70 37.49 15.05 An LEC has been generated using the PML losses and their respective exceedance probabilities for all return period flood events for roads. Figure 5-25 shows the LEC for the entire Ganges basin considering flood losses to roads. Final Report: Volume I Confidential Page 166 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 5 - 2 5 : E x c e e d a n c e p r o b a b i l i t y c u r ve s h o w i n g t h e t o t a l l o s s e s d u e t o f l o o d f o r Infrastructure – Road network Similarly, losses are aggregated at each of the 18 sub-basins of Ganges basin. Sub-basin wise AAL are shown in Figure 5-26. Middle Ganges sub-basin shares 16% of total road network AAL followed by Lower Ganges (14%) and Yamuna (10%) sub-basins. Though it has a share of only 6.3% in total road exposure, the Lower Ganges sub-basin records the second highest losses because of frequent flooding due to its location at the most low-lying down-stream area of the Ganges basin. The same is portrayed spatially by the AAL map (Figure 5-27) for the whole basin. The district/sub-district level AAL is joined with the admin boundary and shown thematically in Figure 5-27. Similar thematic maps showing PML for all return period flood events are provided in the Ganges Basin Risk Atlas. Sub-basin wise detailed discussion is provided in the next chapter. F i g u r e 5 - 2 6 : S u b - b a s i n l e ve l A A L ( m i l l i o n I N R ) f o r I n f r a s t r u c t u r e – R o a d n e t w o r k Final Report: Volume I Confidential Page 167 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Figure 5-27: AAL (thousand INR) due to flood for Infrastructure – Road network 5.2.5.2 Rail Network Total flood affected railway length varies between 1,050 km to 1,910 km for 2-year and 100- year return period flood events respectively, which is around 5.3% and 9.6% of total rail network lengths in the Ganges Basin. Table 5-14 shows the affected length of railway lines due to various return period flood events. T a b l e 5 - 1 4 : Af f e c t e d l e n g t h ( k m ) d u e t o f l o o d e ve n t s o f v a r i o u s r e t u r n p e r i o d s f o r Infrastructure – Rail network Infrastructure affected length (km): Rail Network Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year Entire Ganges Basin 1,049.82 1,327.23 1,482.36 1,688.58 1,870.23 1,909.72 Bangladesh 16.49 22.10 24.26 24.60 27.61 28.99 India 1,030.88 1,302.17 1,454.79 1,660.29 1,838.54 1,876.44 Nepal 2.45 2.96 3.31 3.70 4.08 4.29 On a basin level, rail network incurs a PML of around INR 632 million due to a 100-year return period flood event and an AAL of around INR 159 million. Estimated PML and AAL for the rail network for the entire Ganges Basin and across all three countries due to various return period flood events are shown in Table 5-15. Final Report: Volume I Confidential Page 168 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 5 - 1 5 : P M L a n d A A L d u e t o f l o o d e ve n t s o f va r i o u s r e t u r n p e r i o d s f o r Infrastructure – Rail network Infrastructure Loss (Million INR): Rail Network Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Entire Ganges Basin 256.42 348.59 415.71 506.21 607.89 631.97 159.25 Bangladesh 3.66 4.77 5.60 6.29 7.92 8.32 2.20 India 252.53 343.55 409.80 499.57 599.54 623.20 156.92 Nepal 0.23 0.28 0.31 0.36 0.43 0.45 0.13 A loss exceedance curve (LEC) has been generated using PML and their respective exceedance probabilities for all return period flood events for the rail network. Figure 5-28 shows the LEC for the entire Ganges basin considering rail network flood losses. F i g u r e 5 - 2 8 : E x c e e d a n c e p r o b a b i l i t y c u r ve s h o w i n g t h e t o t a l l o s s e s d u e t o f l o o d f o r Infrastructure – Rail network Similarly, losses are aggregated at each of the 18 sub-basins of the Ganges basin. Sub- basin wise AAL are shown in Figure 5-29. The Lower Ganges sub-basin accounts for a maximum of 26% of rail network AAL followed by the Middle Ganges (17%) and Mahananda (8%) sub-basins. Though the Lower Ganges sub-basin has a share of only 9% in total rail network exposure, it has the highest losses because of frequent floods and due to its location in the most low-lying down-stream area of the Ganges basin. In contrast, the Yamuna sub basin has the highest share of total rail network exposure at 15% but only has 7% contribution in rail network AAL. The same is portrayed spatially by the AAL map (Figure 5-30) for the entire basin. The district/sub-district level AAL is joined with the admin boundary and shown thematically in Figure 5-30. Similar thematic maps showing PML for all return period flood events are provided in the Ganges Basin Risk Atlas. Sub-basin wise detailed discussion is provided in the next chapter. Final Report: Volume I Confidential Page 169 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 5 - 2 9 : S u b - b a s i n l e ve l A A L ( m i l l i o n I N R ) f o r I n f r a s t r u c t u r e – R a i l n e t w o r k Figure 5-30: AAL (thousand INR) due to flood for Infrastructure – Rail network Final Report: Volume I Confidential Page 170 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 5.2.6 AGRICULTURE Agriculture crops have a major share in total losses due to floods in a basin. Detailed risk assessment for the three major crops, viz., Rice, Wheat, and Maize are provided in the following sections. 5.2.6.1 Rice Rice is one of the major crops sown in the study area and usually termed as a water resistant crop. Total estimated flood affected rice cultivated area varies between 1,347 to 2,098 thousand hectare for 2-year and 100-year return period flood events respectively, which is around 10% and 15% of total cultivated area under rice in Ganges Basin. Table 5-16 shows the affected cultivated area of rice due to various return period flood events. T a b l e 5 - 1 6 : Af f e c t e d c u l t i va t e d a r e a ( t h o u s a n d h e c t a r e ) d u e t o f l o o d e ve n t s o f v a r i o u s return periods for Agriculture – Rice Agriculture Affected Area (Thousand hectare): Rice Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year Entire Ganges Basin 1,347 1,602 1,728 1,899 2,069 2,098 Bangladesh 126 142 149 156 168 177 India 1,195 1,431 1,548 1,710 1,865 1,884 Nepal 26 29 31 33 36 38 On a basin level, PML for rice crop is estimated to be around INR 3,447 million due to a 100- year return period flood event and an AAL of around INR 1,136 million. Estimated PML and AAL for rice crop for the entire Ganges Basin and for all three countries due to various return period flood events are shown in Table 5-17. T a b l e 5 - 1 7 : P M L a n d A A L d u e t o f l o o d e ve n t s o f va r i o u s r e t u r n p e r i o d s f o r Agriculture – Rice Agriculture Loss (Million INR): Rice 100- Return Period 2-Year 5-Year 10-Year 25-Year 50-Year Year AAL Entire Ganges Basin 2,008.59 2,440.08 2,682.02 3,004.77 3,375.31 3,447.21 1,135.83 Bangladesh 453.08 513.12 536.62 562.02 608.87 639.32 243.15 India 1,455.35 1,812.45 2,022.78 2,307.49 2,612.63 2,646.40 837.96 Nepal 100.16 114.52 122.62 135.25 153.80 161.49 54.71 A loss exceedance curve (LEC) has been generated using PML and their respective exceedance probabilities for all return period flood events for Rice crop. Figure 5-31 shows the LEC for the entire Ganges basin considering flood losses to rice crop. Final Report: Volume I Confidential Page 171 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 5 - 3 1 : E x c e e d a n c e p r o b a b i l i t y c u r ve s h o w i n g t h e t o t a l l o s s e s d u e t o f l o o d f o r Agriculture – Rice Similarly, losses are aggregated at each of the 18 sub-basins of Ganges basin. Sub-basin wise AAL are shown in Figure 5-32. The Mahananda sub-basin accounts for a maximum of 27% of rice crop AAL followed by Lower Ganges (26%) and Ghagra (14%) sub-basins. The Mahananda sub-basin has the highest losses because it has a high share of 18% in total rice crop exposure value that gets affected by frequent flooding. The same is portrayed spatially by the AAL map (Figure 5-33) for the whole basin. The district/sub-district level AAL is joined with the admin boundary and shown thematically in Figure 5-33. Similar thematic maps showing PML for all return period flood events are provided in the Ganges Basin Risk Atlas. Sub-basin wise detailed discussion is provided in the next chapter. F i g u r e 5 - 3 2 : S u b - b a s i n l e ve l A A L ( m i l l i o n I N R ) f o r A g r i c u l t u r e – R i c e Final Report: Volume I Confidential Page 172 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Figure 5-33: AAL (thousand INR) due to flood for Agriculture – Rice 5.2.6.2 Wheat Wheat is one of the major crops sown in the study area and usually termed as a highly water sensitive crop. Total estimated flood affected wheat cultivated area varies between 1,368 to 2,211 thousand hectares for 2-year and 100-year return period flood events respectively which is around 7% and 12% of total cultivated area of wheat in the Ganges Basin. Table 5-18 shows the affected cultivated area under wheat due to various return period flood events. T a b l e 5 - 1 8 : Af f e c t e d c u l t i va t e d a r e a ( t h o u s a n d h e c t a r e ) d u e t o f l o o d e ve n t s o f v a r i o u s return periods for Agriculture – Wheat Agriculture Affected Area (Thousand hectare): Wheat Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year Entire Ganges Basin 1,368 1,657 1,812 2,000 2,162 2,211 Bangladesh 12 14 14 15 17 17 India 1,344 1,629 1,782 1,969 2,128 2,176 Nepal 13 14 15 16 17 18 On a basin level, PML for wheat crop is estimated at around INR 5,360 million due to a 100- year return period flood event and an AAL of around INR 1,797 million. Estimated PML and AAL for wheat crop for the entire Ganges Basin and for all three countries due to various return period flood events are shown in Table 5-19. Final Report: Volume I Confidential Page 173 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 5 - 1 9 : P M L a n d A A L d u e t o f l o o d e ve n t s o f va r i o u s r e t u r n p e r i o d s f o r Agriculture – Wheat Agriculture Loss (Million INR): Wheat 100- Return Period 2-Year 5-Year 10-Year 25-Year 50-Year Year AAL Entire Ganges Basin 3,147.11 3,905.29 4,328.64 4,816.22 5,220.70 5,359.56 1,796.51 Bangladesh 28.23 32.34 34.07 35.96 39.01 40.96 15.27 India 3,087.23 3,837.21 4,256.86 4,738.83 5,135.23 5,269.82 1,764.13 Nepal 31.65 35.74 37.71 41.42 46.46 48.78 17.11 A loss exceedance curve (LEC) has been generated using PML and their respective exceedance probabilities for all return period flood events for wheat crop. Figure 5-34 shows the LEC for the entire Ganges basin considering flood losses to the wheat crop. Figure 5-34: Exceedance probability curve showing the total losses due to flood for Agriculture – Wheat Similarly, losses are aggregated at each of the 18 sub-basins of Ganges basin. Sub-basin wise AAL are shown in Figure 5-35. Ghagra sub-basin accounts for a maximum of 20% of wheat crop AAL followed by Yamuna (17%) and Upper Ganges (12%) sub-basins. The Ghagra sub-basin records high losses on account of its high share of 14% in total wheat crop exposure that is affected by frequent flooding. The same is portrayed spatially by the AAL map (Figure 5-36) for the whole basin. The district/sub-district level AAL is joined with the admin boundary and shown thematically in Figure 5-36. Similar thematic maps showing PML for all return period flood events are provided in the Ganges Basin Risk Atlas. Sub- basin wise detailed discussion is provided in the next chapter. Final Report: Volume I Confidential Page 174 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 5 - 3 5 : S u b - b a s i n l e ve l A A L ( m i l l i o n I N R ) f o r A g r i c u l t u r e – W h e a t Figure 5-36: AAL (thousand INR) due to flood for Agriculture – Wheat Final Report: Volume I Confidential Page 175 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 5.2.6.3 Maize Maize is one of the major crops sown in the study area and usually termed as a water sensitive crop. Total estimated flood affected cultivated area under maize varies between 265 to 408 thousand hectares for 2-year and 100-year return period flood events respectively which is around 7% and 11% of total cultivated area under Maize in the Ganges Basin. Table 5-20 shows the affected cultivated area under maize due to various return period flood events. T a b l e 5 - 2 0 : Af f e c t e d c u l t i va t e d a r e a ( t h o u s a n d h e c t a r e ) d u e t o f l o o d e ve n t s o f v a r i o u s return periods for Agriculture – Maize Agriculture Affected Area (Thousand hectare): Maize Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year Entire Ganges Basin 265 313 339 372 403 408 Bangladesh 0 0 0 0 0 0 India 257 304 329 362 393 397 Nepal 8 9 9 10 11 11 On a basin level, PML for maize is estimated at around INR 547 million due to a 100-year return period flood event and an AAL of around INR 186 million. Estimated PML and AAL for maize crop for the entire Ganges Basin and for all three countries due to various return period flood events are shown in Table 5-21. T a b l e 5 - 2 1 : P M L a n d A A L d u e t o f l o o d e ve n t s o f va r i o u s r e t u r n p e r i o d s f o r Agriculture – Maize Agriculture Loss (Million INR): Maize Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Entire Ganges Basin 329.87 403.35 440.99 490.28 540.70 546.60 186.44 Bangladesh - - - - - - - India 318.93 391.31 428.37 476.91 526.35 531.53 180.63 Nepal 10.94 12.04 12.62 13.37 14.35 15.07 5.81 A loss exceedance curve (LEC) has been generated using PML and their respective exceedance probabilities for all return period flood events for maize crop. Figure 5-37 shows the LEC for the entire Ganges basin considering flood losses to maize. Final Report: Volume I Confidential Page 176 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia F i g u r e 5 - 3 7 : E x c e e d a n c e p r o b a b i l i t y c u r ve s h o w i n g t h e t o t a l l o s s e s d u e t o f l o o d f o r Agriculture – Maize Similarly, losses are aggregated at each of the 18 sub-basins of Ganges basin. Sub-basin wise AAL are shown in Figure 5-38. Lower Ganges sub-basin accounts for a maximum of 32% of maize AAL followed by Kosi (11%) and Bagmati (10%) sub-basins. The Lower Ganges sub-basin records the highest losses on account of its high share of 8% in total maize crop exposure that is affected by frequent flooding due to its location at the most down-stream area of the basin. The same is portrayed spatially by the AAL map (Figure 5-39) for the whole basin. The district/sub-district level AAL is joined with the administrative boundary and shown thematically in Figure 5-39. Similar thematic maps showing PML for all return period flood events are provided in the Ganges Basin Risk Atlas. Sub-basin wise detailed discussion is provided in the next chapter. F i g u r e 5 - 3 8 : S u b - b a s i n l e ve l A A L ( m i l l i o n I N R ) f o r A g r i c u l t u r e – M a i z e Final Report: Volume I Confidential Page 177 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Figure 5-39: AAL (thousand INR) due to Flood for Agriculture – Maize Final Report: Volume I Confidential Page 178 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 6 Findings – Sub-basin Level This chapter summarizes all the losses at each of the 18 sub-basins of the Ganges Basin. These sub-basins are listed here in descending order of the severity of flood losses with respect to the base exposure. All the priority areas (hotspots) and main exposures at risk are also highlighted in these sections. It also provides State authorities with a breakup of the state-level losses at a sub-basin level so that they can take necessary action. The detailed tables and risk thematic maps at sub-basin level are provided in the Ganges Basin Risk Atlas Report. 6.1.1 LOWER GANGES SUB-BASIN Affected Population: The Lower Ganges sub-basin is the most severely affected sub-basin in the Ganges Basin. Most of the area of this sub-basin is low-lying, which makes the assets here more vulnerable to incur losses. The total number of persons affected due to a 2-year return period flood event is estimated at 717 thousand, 9,573 thousand, and 8 thousand for Bangladesh, India, and Nepal respectively. This is around 31%, 21%, and 9% of the total population of the area of the respective countries lying in the sub-basin. The total estimated number of persons affected in the sub-basin is given in Table 6-1. Table 6-1: Total number (in thousands) of persons affected due to various return p e r i o d f l o o d e ve n t s – L o w e r G a n g e s s u b - b a s i n Sub-basin: Lower Ganges Bangladesh India Nepal Return Period Male Female Total Male Female Total Male Female Total 2-Year 357 360 717 5,034 4,538 9,573 4 4 8 5-Year 408 412 820 5,835 5,260 11,095 4 5 9 10-Year 431 434 865 6,208 5,595 11,803 4 5 9 25-Year 461 465 926 6,663 6,007 12,670 5 5 10 50-Year 514 518 1,032 7,262 6,546 13,808 5 6 11 100-Year 540 544 1,084 7,262 6,546 13,808 5 6 11 Economic Losses: The total PML, combining losses for all the exposure classes under study for the Lower Ganges sub-basin, varies from INR 17,371 million to INR 32,820 million due to 2-year and 100-year return period flood events respectively. Similarly, AAL of all the exposure classes is estimated at around INR 10,028 million. The risk assessment analysis shows that residential buildings bear the maximum losses followed by commercial buildings. Among all the exposure classes, residential buildings have a maximum AAL of INR 7,347 million that is 73.3% of total AAL, while commercial buildings and rice crop are the second and third biggest contributors in total AAL with 19.3% and 2.9% contributions respectively. Health facilities have the least contribution to total AAL (0.1%). Estimated PML and AAL for various exposure classes for the sub-basin due to different return period flood events are shown in Table 6-2. Final Report: Volume I Confidential Page 179 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 6 - 2 : P M L a n d A A L d u e t o f l o o d e ve n t s o f d i f f e r e n t r e t u r n p e r i o d s f o r va r i o u s exposure classes – Lower Ganges sub-basin Losses (Million INR): Lower Ganges Sub-basin Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Lower Ganges Sub-basin 17,370.63 21,554.19 24,180.51 27,773.63 32,721.21 32,820.46 10,028.33 Residential 12,674.15 15,808.71 17,787.81 20,521.07 24,293.24 24,322.01 7,346.96 Commercial 3,362.52 4,149.10 4,641.45 5,289.89 6,192.52 6,241.63 1,932.29 Industrial 62.80 77.09 85.87 97.74 114.16 115.39 35.95 Building Education 28.39 35.28 39.60 45.35 53.10 53.45 16.39 Health 19.56 24.42 27.50 31.57 36.99 37.05 11.33 Others 243.01 300.39 336.02 383.70 448.98 452.28 139.79 Road 48.94 65.65 77.15 96.62 122.20 122.64 30.26 Infrastructure Railway 66.14 85.59 100.48 122.21 149.97 150.19 39.87 Rice 523.30 609.19 652.06 717.68 795.33 810.13 287.44 Agriculture Wheat 234.26 274.62 298.36 323.14 357.17 358.13 129.26 Maize 107.56 124.14 134.19 144.65 157.54 157.55 58.78 A loss exceedance curve (LEC) has been generated using PML and their respective exceedance probabilities for all return period flood events for the sub-basin. Figure 6-1 shows the LEC for the sub-basin considering flood losses to all the exposure types. F i g u r e 6 - 1 : E x c e e d a n c e p r o b a b i l i t y c u r ve s h o w i n g t h e t o t a l l o s s e s d u e t o f l o o d f o r a l l e x p o s u r e t yp e s – L o w er G a n g e s s u b - b a s i n The sub-basin level loss numbers can be further divided at state/province levels to provide a better understanding of the state-level losses. Bihar in India has the maximum AAL of INR 7,994 million, which is around 79.7% of sub-basin’s total AAL followed by West Bengal in India (10.8%) and Rajshahi in Bangladesh (3.5%). Rest of the states/provinces of these three countries have insignificant loss contributions in the sub-basin level losses. Table 6-3 shows the state/province level break-up of the PML and AAL for the sub-basin. Final Report: Volume I Confidential Page 180 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 6 - 3 : S t a t e / P r o vi n c e w i s e b r e a k u p o f P M L a n d A A L – L o w e r G a n g e s s u b - b a s i n State/Province wise Losses (Thousand INR): Lower Ganges Sub-basin Return Period Country State/Province 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Kushtia 526,831.88 625,725.69 677,767.79 751,436.12 877,246.69 921,109.02 294,300.48 Bangladesh Pabna 64,633.08 97,834.88 109,333.69 123,419.95 152,072.63 159,676.26 41,319.32 Rajshahi 660,211.89 733,634.52 784,665.08 857,032.61 941,658.48 988,841.38 354,952.57 Bihar 13,726,580.98 17,252,741.85 19,503,858.62 22,449,258.53 26,595,680.74 26,595,927.06 7,994,381.25 India Jharkhand 450,935.19 557,034.22 591,778.41 691,920.94 822,547.87 822,547.87 256,780.06 West Bengal 1,936,536.45 2,281,779.92 2,507,319.06 2,894,375.14 3,325,179.25 3,325,194.28 1,083,969.31 Nepal Eastern 4,900.24 5,434.39 5,786.77 6,190.95 6,824.35 7,165.57 2,624.44 Final Report: Volume I Confidential Page 181 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia The study further identifies hotspot areas based on the significance of their contribution to AAL. Patna rural, Munger, Dinapur-cum-khagaul, and Mokameh are the worst affected blocks (sub-districts) in India whereas Kustia is the worst affected district in Bangladesh. Table 6-4 shows the top five blocks/districts in Lower Ganges sub-basin based on AAL. As maximum losses are incurred by the residential building class, the residential AAL map is plotted to identify the areas with maximum losses. The district/sub-district level residential AAL map is shown in Figure 6-2. T a b l e 6 - 4 : B l o c k s / d i s t r i c t s h a vi n g m a x i m u m A A L – L o w e r G a n g e s s u b - b a s i n Sub-basin Country State District Block/sub-district AAL (million INR) India Bihar Patna Patna Rural 1101.51 India Bihar Patna Dinapur-cum-khagaul 475.51 India Bihar Munger Munger 390.06 Bangladesh Kushtia Kustia - 293.32 Lower Ganges India Bihar Patna Mokameh 267.76 Figure 6-2: AAL (million INR) due to floods for buildings: Residential – Lower Ganges sub-basin Final Report: Volume I Confidential Page 182 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 6.1.2 BAGMATI SUB-BASIN Affected Population: The Bagmati sub-basin is the second most severely affected sub-basin in the Ganges Basin. The downstream area of Bagmati river sub-basin is low-lying, which makes the assets at downstream more vulnerable to incur losses. The total number of persons affected due to a 2-year return period flood event is estimated at 2,168 thousand and 78 thousand for India and Nepal respectively. This is around 19% and 2% of the total population of the area of the respective countries lying in the sub-basin. The total estimated number of persons affected in the sub-basin is given in Table 6-5 Table 6-5: Total number (in thous ands) of persons affected due to various return p e r i o d f l o o d e ve n t s – B a g m a t i s u b - b a s i n Sub-basin: Bagmati India Nepal Return Period Male Female Total Male Female Total 2-Year 1,142 1,027 2,168 40 38 78 5-Year 1,338 1,203 2,540 44 43 87 10-Year 1,511 1,358 2,869 45 43 89 25-Year 1,641 1,475 3,116 47 45 92 50-Year 1,828 1,643 3,471 49 47 96 100-Year 1,828 1,643 3,471 51 50 101 Economic Losses: The total PML, combining losses for all the exposure classes under study for the Bagmati sub-basin, varies from INR 2,480 million to INR 4,733 million due to 2-year and 100-year return period flood events respectively. Similarly, AAL of all the exposure classes is estimated at around INR 1,435 million. The risk assessment analysis shows that residential buildings bear the maximum losses followed by commercial buildings. Among all the exposure classes, residential buildings have a maximum AAL of INR 1,041 million that is 72.6% of total AAL, while commercial buildings and wheat crop are the second and third biggest contributors in total AAL with 20.6% and 2.2% contributions respectively. Health facilities have the least contribution to total AAL (0.1%). Estimated PML and AAL for various exposure classes for the sub-basin due to different return period flood events are shown in Table 6-6. T a b l e 6 - 6 : P M L a n d A A L d u e t o f l o o d e ve n t s o f d i f f e r e n t r e t u r n p e r i o d s f o r va r i o u s exposure classes – Bagmati sub-basin Losses (Million INR): Bagmati Sub-basin 100- Return Period 2-Year 5-Year 10-Year 25-Year 50-Year Year AAL Bagmati Sub-basin 2,480.19 3,045.73 3,520.24 4,003.96 4,730.59 4,733.12 1,434.56 Residential 1,801.61 2,207.08 2,550.47 2,902.56 3,436.31 3,436.51 1,041.00 Building Commercial 509.85 627.18 726.77 823.20 967.75 968.93 295.11 Industrial 6.48 7.93 9.14 10.34 12.10 12.13 3.73 Final Report: Volume I Confidential Page 183 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Losses (Million INR): Bagmati Sub-basin 100- Return Period 2-Year 5-Year 10-Year 25-Year 50-Year Year AAL Education 2.89 3.56 4.10 4.66 5.51 5.51 1.67 Health 1.84 2.27 2.60 2.96 3.50 3.50 1.06 Others 22.29 27.47 31.50 35.78 42.08 42.16 12.88 Road 6.55 9.15 11.45 14.40 18.66 18.72 4.23 Infrastructure Railway 12.66 17.07 20.89 26.31 33.05 33.05 7.95 Rice 26.96 33.67 39.35 45.31 54.82 55.42 15.82 Agriculture Wheat 55.44 67.58 76.66 85.18 96.71 97.06 31.63 Maize 33.62 42.77 47.32 53.27 60.11 60.14 19.47 A loss exceedance curve (LEC) has been generated using PML and their respective exceedance probabilities for all return period flood events for the sub-basin. Figure 6-3 shows the LEC for the sub-basin considering flood losses to all the exposure types. F i g u r e 6 - 3 : E x c e e d a n c e p r o b a b i l i t y c u r ve s h o w i n g t h e t o t a l l o s s e s d u e t o f l o o d f o r a l l e x p o s u r e t yp e s – B a g m a t i s u b - b a s i n The sub-basin level loss numbers can be further divided at state/province levels to provide a better understanding of the state-level losses. The sub-basin’s maximum area lies in Bihar, which has the maximum exposure to loss. Bihar has the maximum AAL of INR 1,414 million which is around 98.6% of sub-basin’s total AAL. Central province in Nepal has insignificant loss contribution (1.4%) in the sub-basin level losses. Table 6-7 shows the state/province level break-up of the PML and AAL for the sub-basin. Final Report: Volume I Confidential Page 184 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 6 - 7 : S t a t e / P r o vi n c e w i s e b r e a k u p o f P M L a n d A A L – B a g m a t i s u b - b a s i n State/Province wise Losses (Thousand INR): Bagmati Sub-basin Return Period Country State/Province 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL India Bihar 2,442,543.69 3,002,599.07 3,475,714.18 3,957,081.72 4,679,978.16 4,679,978.16 1,414,307.06 Nepal Central 37,642.76 43,130.66 44,528.80 46,876.45 50,616.02 53,146.82 20,252.78 Final Report: Volume I Confidential Page 185 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia The study further identifies hotspot areas based on the significance of their contribution to AAL. Begusarai, Khagaria, Teghra, Barauni, and Musahri in India are the worst affected blocks, which incur the highest annualized losses every year in the sub-basin. Table 6-8 shows the top five blocks/districts in Bagmati sub-basin based on AAL. As maximum losses are incurred by the residential building class, the residential AAL map is plotted to identify the areas with maximum losses. The district/sub-district level residential AAL map is shown in Figure 6-4. T a b l e 6 - 8 : B l o c k s / d i s t r i c t s h a vi n g m a x i m u m A A L – B a g m a t i s u b - b a s i n Sub-basin Country State District Block/sub-district AAL (million INR) India Bihar Begusarai Begusarai 143.11 India Bihar Khagaria Khagaria 89.06 India Bihar Begusarai Teghra 65.72 India Bihar Begusarai Barauni 64.04 Bagmati India Bihar Muzaffarpur Musahri 56.22 Figure 6-4: AAL (million INR) due to floods for buildings: Residential – Bagmati sub- basin Final Report: Volume I Confidential Page 186 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 6.1.3 KOSI SUB-BASIN Affected Population: The Kosi sub-basin is the third most severely affected sub-basin in the Ganges Basin. The downstream area of Kosi river sub-basin is low-lying, which makes the assets at downstream more vulnerable to incur losses. The total number of persons affected due to a 2-year return period flood event is estimated at 1,158 thousand and 71 thousand for India and Nepal respectively. This is around 13.0%, and 2.1% of the total population of the area of the respective countries lying in the sub-basin. The total estimated number of persons affected in the sub-basin is given in Table 6-9. Table 6-9: Total number (in thousands) of persons affected due to various return p e r i o d f l o o d e ve n t s – K o s i s u b - b a s i n Sub-basin: Kosi India Nepal Return Period Male Female Total Male Female Total 2-Year 610 548 1,158 34 37 71 5-Year 682 613 1,294 37 40 77 10-Year 723 650 1,373 39 42 80 25-Year 788 709 1,497 41 44 85 50-Year 869 782 1,651 44 47 91 100-Year 869 782 1,651 46 50 96 Economic Losses: The total PML, combining losses for all the exposure classes under study for the Kosi sub- basin, varies from INR 1,408 million to INR 2,509 million due to 2-year and 100-year return period flood events respectively. Similarly, AAL of all the exposure classes is estimated at around INR 796 million. The risk assessment analysis shows that residential buildings bear the maximum losses followed by commercial buildings. Among all the exposure classes, residential buildings have a maximum AAL of INR 487 million that is 61.2% of total AAL, while commercial buildings and maize crop are the second and third biggest contributors in total AAL with 27.3% and 2.7% contribution respectively. Health facilities have the least contribution to total AAL (0.1%). Estimated PML and AAL for various exposure classes for the sub-basin due to different return period flood events are shown in Table 6-10. T a b l e 6 - 1 0 : P M L a n d A A L d u e t o f l o o d e ve n t s o f d i f f e r e n t r e t u r n p e r i o d s f o r va r i o u s exposure classes – Kosi sub-basin Losses (Million INR): Kosi Sub-basin 100- Return Period 2-Year 5-Year 10-Year 25-Year 50-Year Year AAL Kosi Sub-basin 1,407.51 1,688.20 1,866.14 2,137.46 2,504.93 2,508.81 795.93 Residential 857.97 1,034.42 1,148.31 1,321.63 1,561.82 1,562.10 487.40 Commercial 386.91 459.66 504.41 575.28 668.84 670.91 217.21 Industrial 5.23 6.18 6.75 7.68 8.84 8.89 2.92 Building Education 1.82 2.20 2.43 2.80 3.28 3.29 1.03 Health 0.96 1.16 1.29 1.49 1.75 1.75 0.55 Others 24.59 29.23 32.06 36.59 42.43 42.57 13.80 Final Report: Volume I Confidential Page 187 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Losses (Million INR): Kosi Sub-basin 100- Return Period 2-Year 5-Year 10-Year 25-Year 50-Year Year AAL Road 13.76 17.18 19.24 22.26 26.36 26.57 7.97 Infrastructure Railway 9.37 11.86 13.56 16.19 19.36 19.37 5.52 Rice 34.88 40.64 44.46 49.95 56.25 56.98 19.33 Agriculture Wheat 33.97 40.12 44.00 48.61 54.76 55.00 18.91 Maize 38.06 45.56 49.64 54.99 61.24 61.38 21.28 A loss exceedance curve (LEC) has been generated using PML and their respective exceedance probabilities for all return period flood events for the sub-basin. Figure 6-5 shows the LEC for the sub-basin considering flood losses to all the exposure types. F i g u r e 6 - 5 : E x c e e d a n c e p r o b a b i l i t y c u r ve s h o w i n g t h e t o t a l l o s s e s d u e t o f l o o d f o r a l l e x p o s u r e t yp e s – K o s i s u b - b a s i n The sub-basin level loss numbers can be further divided at state/province levels to provide a better understanding of the state-level losses. Bihar in India has the maximum AAL of INR 764 million which is around 96.0% of sub-basin’s total AAL followed by Eastern province in Nepal (3.2%). Rest of the states/provinces of these three countries have insignificant loss contributions in the sub-basin level losses. Table 6-11 shows the state/province level break- up of the PML and AAL for the sub-basin. Final Report: Volume I Confidential Page 188 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 6 - 1 1 : S t a t e / P r o vi n c e w i s e b r e a k u p o f P M L a n d A A L – K o s i s u b - b a s i n State/Province wise Losses (Thousand INR): Kosi Sub-basin Return Period Country State/Province 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL India Bihar 1,347,157.68 1,623,452.55 1,798,499.62 2,066,179.52 2,427,307.05 2,427,309.77 764,272.29 Central 11,357.92 11,779.64 12,002.74 12,274.16 12,696.10 13,330.91 5,811.26 Nepal Eastern 48,996.88 52,972.51 55,642.36 59,002.07 64,922.72 68,168.86 25,845.81 Final Report: Volume I Confidential Page 189 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia The study further identifies hotspot areas based on the significance of their contribution to AAL. Gogri, Kahara, Bihpur, Barari, and Narayanpur in India are the worst affected blocks, which incur the highest annualized losses every year in the sub-basin. As maximum losses are incurred by the residential building class, the residential AAL map is plotted to identify the areas with maximum losses. The district/sub-district level residential AAL map is shown in Figure 6-6. Figure 6-6: AAL (million INR) due to floods for buildings: Residential – Kosi sub- basin Final Report: Volume I Confidential Page 190 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 6.1.4 GANDAK SUB-BASIN Affected Population: The Gandak sub-basin is also one of the frequently affected sub-basins in the Ganges Basin. Most of the area of this sub-basin in India is low-lying whereas the area in Nepal is at high elevation, which makes the assets in Indian part more vulnerable to incur losses. The total number of persons affected due to a 2-year return period flood event is estimated at 2,794 thousand, and 44 thousand for India and Nepal respectively. This is around 29.2%, and 1.0% of the total population of the area of the respective countries lying in the sub-basin. The total estimated number of persons affected in the sub-basin is given in Table 6-12. Table 6-12: Total number (in thousands) of persons affected due to various return p e r i o d f l o o d e ve n t s – G a n d a k s u b - b a s i n Sub-basin: Gandak India Nepal Return Period Male Female Total Male Female Total 2-Year 1,441 1,353 2,794 20 24 44 5-Year 1,831 1,725 3,556 22 26 48 10-Year 1,938 1,825 3,763 24 28 51 25-Year 2,212 2,087 4,300 25 29 54 50-Year 2,332 2,200 4,531 28 33 61 100-Year 2,332 2,200 4,531 30 35 64 Economic Losses: The total PML, combining losses for all the exposure classes under study for the Gandak sub-basin, varies from INR 3,520 million to INR 7,230 million due to 2-year and 100-year return period flood events respectively. Similarly, AAL of all the exposure classes is estimated at around INR 2,109 million. The risk assessment analysis shows that residential buildings bear the maximum losses followed by commercial buildings. Among all the exposure classes, residential buildings have a maximum AAL of INR 1,344 million that is 63.7% of total AAL, while commercial buildings and wheat crops are the next biggest contributors in total AAL with 26.9% and 3.4% contributions respectively. Health facilities have the least contribution to total AAL (0.1%). Estimated PML and AAL for various exposure classes for the sub-basin due to different return period flood events are shown in Table 6-13. T a b l e 6 - 1 3 : P M L a n d A A L d u e t o f l o o d e ve n t s o f d i f f e r e n t r e t u r n p e r i o d s f o r va r i o u s exposure classes – Gandak sub-basin Losses (Million INR): Gandak Sub-basin 100- Return Period 2-Year 5-Year 10-Year 25-Year 50-Year Year AAL Gandak Sub-basin 3,519.61 4,674.36 5,298.74 6,175.03 7,223.42 7,229.78 2,109.28 Residential 2,234.64 2,983.51 3,390.19 3,950.73 4,632.88 4,633.75 1,343.83 Commercial 948.84 1,255.99 1,422.73 1,657.66 1,935.18 1,938.20 567.50 Building Industrial 11.38 14.98 16.95 19.67 22.96 23.04 6.78 Education 5.52 7.32 8.28 9.61 11.23 11.23 3.30 Health 3.23 4.28 4.85 5.63 6.57 6.58 1.93 Final Report: Volume I Confidential Page 191 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Losses (Million INR): Gandak Sub-basin 100- Return Period 2-Year 5-Year 10-Year 25-Year 50-Year Year AAL Others 49.63 65.86 74.56 86.79 101.64 101.85 29.72 Road 21.49 28.07 33.33 39.78 51.57 52.10 13.09 Infrastructure Railway 18.43 27.79 33.83 41.17 51.20 51.26 12.18 Rice 79.89 100.79 112.84 130.35 152.75 153.98 46.49 Agriculture Wheat 122.94 156.42 169.41 197.24 217.14 217.31 70.98 Maize 23.64 29.34 31.79 36.40 40.30 40.47 13.47 A loss exceedance curve (LEC) has been generated using PML and their respective exceedance probabilities for all return period flood events for the sub-basin. Figure 6-7 shows the LEC for the sub-basin considering flood losses to all the exposure types. F i g u r e 6 - 7 : E x c e e d a n c e p r o b a b i l i t y c u r ve s h o w i n g t h e t o t a l l o s s e s d u e t o f l o o d f o r a l l e x p o s u r e t yp e s – G a n d a k s u b - b a s i n The sub-basin level loss numbers can be further divided at state/province levels to provide a better understanding of the state-level losses. Bihar in India has the maximum AAL of INR 1,964 million which is around 93.1% of sub-basin’s total AAL followed by Uttar Pradesh in India (4.7%). Rest of the provinces of Nepal have insignificant loss contributions in the sub- basin level losses. Table 6-14 shows the state/province level break-up of the PML and AAL for the sub-basin. Final Report: Volume I Confidential Page 192 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 6 - 1 4 : S t a t e / P r o vi n c e w i s e b r e a k u p o f P M L a n d A A L – G a n d a k s u b - b a s i n State/Province wise Losses (Thousand INR): Gandak Sub-basin Return Period Country State/Province 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Bihar 3,266,807.88 4,362,440.78 4,952,436.32 5,782,056.89 6,773,413.56 6,773,489.99 1,964,542.79 India Uttar Pradesh 167,014.72 216,972.97 244,193.26 283,531.33 324,442.39 324,442.39 98,612.78 Central 21,239.47 24,378.66 26,393.30 28,775.61 33,417.46 35,088.34 11,653.88 Mid Western 11.23 11.69 11.84 12.00 12.36 12.98 5.74 Nepal Western 64,538.33 70,555.36 75,707.29 80,651.88 92,135.05 96,741.80 34,461.28 Final Report: Volume I Confidential Page 193 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia The study further identifies hotspot areas based on the significance of their contribution to AAL. Chapra, Sonepur, Dariapur, Garkha, and Dighwara in India are the worst affected blocks, which incur the highest annualized losses every year in the sub-basin. As maximum losses are incurred by the residential building class, the residential AAL map is plotted to identify the areas with maximum losses. The district/sub-district level residential AAL map is shown in Figure 6-8. Figure 6-8: AAL (million INR) due to floods for buildings: Residential – Gandak sub- basin Final Report: Volume I Confidential Page 194 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 6.1.5 MAHANANDA SUB-BASIN Affected Population: The Mahananda sub-basin is one of the severely affected sub-basin in the Ganges Basin. Most of the area of this sub-basin is low-lying, which makes the assets here more vulnerable to incur losses. This sub-basin covers the maximum area of Bangladesh lying in Ganges Basin. Besides, this area is one of the most densely populated areas of Ganges Basin. The total number of persons affected due to a 2-year return period flood event is estimated at 706 thousand, 2,290 thousand, and 52 thousand for Bangladesh, India and Nepal respectively. This is around 17.5%, 18.3%, and 2.1% of the total population of the area of the respective countries lying in the sub-basin. The total estimated number of persons affected in the sub-basin is given in Table 6-15. Table 6-15: Total number (in thousands) of persons affected due to various return p e r i o d f l o o d e ve n t s – M a h a n a n d a s u b - b a s i n Sub-basin: Mahananda Bangladesh India Nepal Return Period Male Female Total Male Female Total Male Female Total 2-Year 349 357 706 1,180 1,110 2,290 25 27 52 5-Year 389 397 786 1,394 1,312 2,706 27 30 57 10-Year 404 412 816 1,438 1,353 2,791 29 32 61 25-Year 416 425 841 1,590 1,494 3,084 31 34 65 50-Year 439 448 887 1,731 1,627 3,358 34 37 71 100-Year 461 470 931 1,731 1,627 3,358 36 39 75 Economic Losses: The total PML, combining losses for all the exposure classes under study for the Mahananda sub-basin, varies from INR 3,244 million to INR 5,327 million due to 2-year and 100-year return period flood events respectively. Similarly, AAL of all the exposure classes is estimated at around INR 1,806 million. The risk assessment analysis shows that residential buildings bear the maximum losses followed by commercial buildings. Among all the exposure classes, residential buildings have a maximum AAL of INR 749 million that is 41.5% of total AAL, while commercial buildings and rice crop are the second and third biggest contributors in total AAL with 34.3% and 16.3% contributions respectively. Health facilities have the least contribution to total AAL (0.1%). Estimated PML and AAL for various exposure classes for the sub-basin due to different return period flood events are shown in Table 6-16. T a b l e 6 - 1 6 : P M L a n d A A L d u e t o f l o o d e ve n t s o f d i f f e r e n t r e t u r n p e r i o d s f o r va r i o u s exposure classes – Mahananda sub -basin Losses (Million INR): Mahananda Sub-basin 100- Return Period 2-Year 5-Year 10-Year 25-Year 50-Year Year AAL Mahananda Sub-basin 3,244.17 3,850.39 4,157.17 4,607.48 5,244.29 5,327.21 1,805.58 Residential 1,335.33 1,599.59 1,735.32 1,949.76 2,249.60 2,273.39 748.90 Building Commercial 1,116.32 1,315.78 1,423.46 1,573.48 1,788.85 1,825.88 619.50 Final Report: Volume I Confidential Page 195 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Losses (Million INR): Mahananda Sub-basin 100- Return Period 2-Year 5-Year 10-Year 25-Year 50-Year Year AAL Industrial 25.84 30.30 32.76 36.06 40.91 41.83 14.30 Education 5.24 6.13 6.62 7.31 8.29 8.51 2.90 Health 2.62 3.10 3.36 3.72 4.26 4.31 1.46 Others 92.65 109.30 118.28 130.48 148.71 151.20 51.43 Road 16.43 20.59 23.26 27.69 35.10 35.34 9.64 Infrastructure Railway 21.53 28.26 31.77 37.09 44.41 44.62 12.82 Rice 536.57 628.60 668.70 717.22 788.06 804.71 294.22 Agriculture Wheat 72.95 85.64 89.14 96.42 104.58 105.78 39.83 Maize 18.69 23.11 24.51 28.27 31.54 31.64 10.59 A loss exceedance curve (LEC) has been generated using PML and their respective exceedance probabilities for all return period flood events for the sub-basin. Figure 6-9 shows the LEC for the sub-basin considering flood losses to all the exposure types. F i g u r e 6 - 9 : E x c e e d a n c e p r o b a b i l i t y c u r ve s h o w i n g t h e t o t a l l o s s e s d u e t o f l o o d f o r a l l e x p o s u r e t yp e s – M a h a n a n d a s u b - b a s i n The sub-basin level loss numbers can be further divided at state/province levels to provide a better understanding of the state-level losses. West Bengal in India has the maximum AAL of INR 955 million which is around 52.9% of sub-basin’s total AAL followed by Rajshahi in Bangladesh (31.7%) and Bihar in India (12.3%). Rest of the states/provinces of these three countries have insignificant loss contributions in the sub-basin level losses. Table 6-17 shows the state/province level break-up of the PML and AAL for the sub-basin. Final Report: Volume I Confidential Page 196 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 6 - 1 7 : S t a t e / P r o vi n c e w i s e b r e a k u p o f P M L a n d A A L – M a h a n a n d a s u b - b a s i n State/Province wise Losses (Thousand INR): Mahananda Sub-basin Return Period Country State/Province 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Dinajpur 61,594.90 80,275.84 88,684.57 98,199.05 110,139.50 115,646.47 36,048.97 Bangladesh Rajshahi 1,059,175.04 1,194,129.44 1,272,124.01 1,364,468.04 1,497,325.75 1,572,213.70 571,477.65 Bihar 377,725.29 487,385.78 528,745.00 644,593.58 769,119.91 769,119.91 222,055.14 Jharkhand 4,928.88 6,206.99 6,381.58 7,764.24 9,316.80 9,316.80 2,823.88 India West Bengal 1,707,977.43 2,043,927.95 2,219,659.66 2,446,910.31 2,807,929.84 2,807,929.86 955,062.41 Nepal Eastern 32,767.39 38,461.96 41,579.12 45,547.09 50,462.19 52,985.30 18,116.58 Final Report: Volume I Confidential Page 197 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia The study further identifies hotspot areas based on the significance of their contribution to AAL. Nawabganj and Naogaon are the worst affected districts in Bangladesh whereas Ratua - I, Harischandrapur – II, and Habibpur are the worst affected blocks in India. These blocks and districts have the highest annualized losses in the sub-basin. As maximum losses are incurred by the residential building class, the residential AAL map is plotted to identify the areas with maximum losses. The district/sub-district level residential AAL map is shown in Figure 6-10. Figure 6-10: AAL (million INR) due to floods for buildings: Residential – Mahananda sub-basin Final Report: Volume I Confidential Page 198 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 6.1.6 MIDDLE GANGES SUB-BASIN Affected Population: The Middle Ganges sub-basin is also highly flood prone. The risk analysis shows that the total number of persons affected due to a 2-year return period flood event is estimated at 4,967 thousand in India. This is around 10.7% of the total population of the area of India lying in the sub-basin. The total estimated number of persons affected in the sub-basin is given in Table 6-18. Table 6-18: Total number (in thousands) of persons affected due to various return p e r i o d f l o o d e ve n t s – M i d d l e G a n g e s s u b - b a s i n Sub-basin: Middle Ganges India Return Period Male Female Total 2-Year 2,595 2,372 4,967 5-Year 3,248 2,968 6,216 10-Year 3,622 3,312 6,935 25-Year 4,067 3,719 7,786 50-Year 4,376 4,005 8,380 100-Year 4,480 4,100 8,580 Economic Losses: The total PML, combining losses for all the exposure classes under study for the Middle Ganges sub-basin, varies from INR 8,727 million to INR 18,266 million due to 2-year and 100-year return period flood events respectively. Similarly, AAL of all the exposure classes is estimated at around INR 5,227 million. The risk assessment analysis shows that residential buildings bear the maximum losses followed by commercial buildings. Among all the exposure classes, residential buildings have a maximum AAL of INR 3,095 million that is 59.2% of total AAL, while commercial buildings and wheat crop are the second and third biggest contributors in total AAL with 30.2% and 4.1% contributions respectively. Health facilities and maize crop have the least contributions to total AAL (0.1%). Estimated PML and AAL for various exposure classes for the sub-basin due to different return period flood events are shown in Table 6-19. T a b l e 6 - 1 9 : P M L a n d A A L d u e t o f l o o d e ve n t s o f d i f f e r e n t r e t u r n p e r i o d s f o r va r i o u s exposure classes – Middle Ganges sub -basin Losses (Million INR): Middle Ganges Sub-basin Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Middle Ganges Sub-basin 8,726.72 11,451.36 13,281.55 15,476.99 17,451.99 18,266.12 5,226.76 Residential 5,159.41 6,804.63 7,849.16 9,180.30 10,335.03 10,829.66 3,094.49 Commercial 2,626.04 3,444.21 4,043.22 4,698.93 5,319.93 5,571.36 1,577.72 Industrial 45.38 60.35 71.55 83.65 94.40 99.46 27.55 Building Education 12.28 16.06 18.71 21.96 25.15 26.23 7.37 Health 10.39 13.70 15.99 18.77 21.38 22.40 6.26 Others 166.34 218.16 254.54 295.09 332.76 348.35 99.70 Infrastructure Road 53.00 74.19 90.05 109.68 131.59 142.25 33.65 Final Report: Volume I Confidential Page 199 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Losses (Million INR): Middle Ganges Sub-basin Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Railway 38.14 58.82 75.10 95.99 114.51 125.36 26.15 Rice 229.44 288.24 331.26 377.94 428.25 436.97 133.74 Agriculture Wheat 372.30 456.45 514.04 574.90 628.21 642.70 212.38 Maize 13.99 16.54 17.91 19.77 20.77 21.37 7.74 A loss exceedance curve (LEC) has been generated using PML and their respective exceedance probabilities for all return period flood events for the sub-basin. Figure 6-11 shows the LEC for the sub-basin considering flood losses to all the exposure types. F i g u r e 6 - 1 1 : E x c e e d a n c e p r o b a b i l i t y c u r ve s h o w i n g t h e t o t a l l o s s e s d u e t o f l o o d f o r a l l e x p o s u r e t yp e s – M i d d l e G a n g e s s u b - b a s i n The sub-basin level loss numbers can be further divided at state/province levels to provide a better understanding of the state-level losses. Uttar Pradesh in India has the maximum AAL of INR 4,226 million which is around 80.8% of sub-basin’s total AAL followed by Bihar in India (19.2%). Table 6-20 shows the state/province level break-up of the PML and AAL for the sub-basin. Final Report: Volume I Confidential Page 200 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 6 - 2 0 : S t a t e / P r o vi n c e w i s e b r e a k u p o f P M L a n d A A L – M i d d l e G a n g e s s u b - b a s i n State/Province wise Losses (Thousand INR): Middle Ganges Sub-basin Return Period Country State/Province 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Bihar 1,668,409.72 2,198,236.60 2,512,346.06 2,982,697.24 3,521,248.70 3,521,264.62 1,001,166.42 India Uttar Pradesh 7,058,306.28 9,253,120.29 10,769,202.26 12,494,289.02 13,930,741.13 14,744,853.74 4,225,597.78 Final Report: Volume I Confidential Page 201 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia The study further identifies hotspot areas based on the significance of their contribution to AAL. Varanasi, Mirzapur, Allahabad, Kanpur blocks of Bihar and Buxar block of Uttar Pradesh in India are the worst affected blocks, which incur the highest annualized losses every year in the sub-basin. As maximum losses are incurred by the residential building class, the residential AAL map is plotted to identify the areas with maximum losses. The district/sub-district level residential AAL map is shown in Figure 6-12. Figure 6-12: AAL (million INR) due to floods for buildings: Residential – Middle Ganges sub-basin Final Report: Volume I Confidential Page 202 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 6.1.7 GHAGRA SUB-BASIN Affected Population: The risk assessment for Ghagra sub-basin shows that the total number of persons affected due to a 2-year return period flood event is estimated at 5,502 thousand, and 129 thousand for India and Nepal respectively. This is around 12.1%, and 1.6% of the total population of the area of the respective countries lying in the sub-basin. The total estimated number of persons affected in the sub-basin is given in Table 6-21. Table 6-21: Total number (in thousands) of persons affected due to various return p e r i o d f l o o d e ve n t s – G h a g r a s u b - b a s i n Sub-basin: Ghagra India Nepal Return Period Male Female Total Male Female Total 2-Year 2,826 2,676 5,502 62 68 129 5-Year 3,471 3,289 6,760 68 75 143 10-Year 3,848 3,645 7,493 72 79 151 25-Year 4,365 4,136 8,501 78 85 163 50-Year 4,835 4,584 9,420 84 92 176 100-Year 4,836 4,585 9,420 88 97 185 Economic Losses: The total PML, combining losses for all the exposure classes under study for the Ghagra sub-basin, varies from INR 5,271 million to INR 10,987 million due to 2-year and 100-year return period flood events respectively. Similarly, AAL of all the exposure classes is estimated at around INR 3,131 million. The risk assessment analysis shows that residential buildings bear the maximum losses followed by commercial buildings. Among all the exposure classes, residential buildings have a maximum AAL of INR 1,474 million that is 47.1% of total AAL, while commercial buildings and wheat crop are the second and third biggest contributors in total AAL with 33.0% and 11.2% contributions respectively. Health facilities and education institute have the least contributions to total AAL (0.1% and 0.2% respectively). Estimated PML and AAL for various exposure classes for the sub-basin due to different return period flood events are shown in Table 6-22. T a b l e 6 - 2 2 : P M L a n d A A L d u e t o f l o o d e ve n t s o f d i f f e r e n t r e t u r n p e r i o d s f o r va r i o u s exposure classes – Ghagra sub-basin Losses (Million INR): Ghagra Sub-basin Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Ghagra Sub-basin 5,271.16 6,799.27 7,768.17 9,222.81 10,969.76 10,986.95 3,131.39 Residential 2,451.15 3,204.10 3,692.72 4,443.23 5,376.83 5,380.43 1,473.75 Commercial 1,761.57 2,235.83 2,536.32 2,978.49 3,525.15 3,531.84 1,034.37 Industrial 23.16 29.16 33.11 38.84 45.86 46.02 13.55 Building Education 8.50 11.07 12.73 15.37 18.66 18.68 5.10 Health 5.30 7.00 8.12 9.91 12.16 12.17 3.22 Others 92.29 116.82 132.67 155.93 184.55 185.01 54.14 Infrastructure Road 34.59 42.90 48.89 57.60 72.42 73.43 20.21 Final Report: Volume I Confidential Page 203 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Losses (Million INR): Ghagra Sub-basin Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Railway 16.67 23.95 29.59 35.95 44.20 44.22 10.77 Rice 248.72 323.47 370.10 445.18 527.11 530.90 148.65 Agriculture Wheat 603.56 770.38 864.72 995.99 1,110.80 1,111.99 352.14 Maize 25.64 34.58 39.20 46.33 52.03 52.26 15.47 A loss exceedance curve (LEC) has been generated using PML and their respective exceedance probabilities for all return period flood events for the sub-basin. Figure 6-13 shows the LEC for the sub-basin considering flood losses to all the exposure types. F i g u r e 6 - 1 3 : E x c e e d a n c e p r o b a b i l i t y c u r ve s h o w i n g t h e t o t a l l o s s e s d u e t o f l o o d f o r a l l e x p o s u r e t yp e s – G h a g r a s u b - b a s i n The sub-basin level loss numbers can be further divided at state/province levels to provide a better understanding of the state-level losses. Uttar Pradesh in India has the maximum AAL of INR 2,640 million which is around 84.3% of sub-basin’s total AAL followed by Bihar in India (10.3%). Rest of the states/provinces of India and Nepal have insignificant loss contributions in the sub-basin level losses. Table 6-23 shows the state/province level break- up of the PML and AAL for the sub-basin. Final Report: Volume I Confidential Page 204 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 6 - 2 3 : S t a t e / P r o vi n c e w i s e b r e a k u p o f P M L a n d A A L – G h a g r a s u b - b a s i n State/Province wise Losses (Thousand INR): Ghagra Sub-basin Return Period Country State/Province 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Bihar 555,851.26 701,888.80 780,928.14 916,212.35 1,072,286.53 1,072,286.53 323,569.92 Uttar Pradesh 4,397,024.27 5,751,225.17 6,623,135.30 7,918,189.78 9,479,446.20 9,482,072.37 2,639,596.90 India Uttarakhand 115,180.21 119,199.38 121,443.14 124,471.00 128,223.26 128,297.61 58,815.22 Far Western 62,551.78 71,442.56 77,283.32 83,323.05 90,795.56 95,335.34 34,074.56 Mid Western 121,282.90 133,057.33 140,660.73 148,319.82 158,500.63 166,425.66 64,345.91 Nepal Western 19,269.41 22,460.88 24,722.17 32,298.95 40,510.35 42,535.87 10,986.68 Final Report: Volume I Confidential Page 205 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia The study further identifies hotspot areas based on the significance of their contribution to AAL. Harraiya, Gorakhpur, Bairia, Tarabganj, and Gola in India are the worst affected blocks, which incur the highest annualized losses every year in the sub-basin. As maximum losses are incurred by the residential building class, the residential AAL map is plotted to identify the areas with maximum losses. The district/sub-district level residential AAL map is shown in Figure 6-14. Figure 6-14: AAL (million INR) due to floods for buildings: Residential – Ghagra sub- basin Final Report: Volume I Confidential Page 206 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 6.1.8 KAMLA-BALAN SUB-BASIN Affected Population: The risk assessment for Kamla-Balan sub-basin shows that the total number of persons affected due to a 2-year return period flood event is estimated at 1,941 thousand, and 50 thousand for India and Nepal respectively. This is around 12.7% and 1.6% of the total population of the area of the respective countries lying in the sub-basin. The total estimated number of persons affected in the sub-basin is given in Table 6-24. Table 6-24: Total number (in thousands) of persons affected due to various return p e r i o d f l o o d e ve n t s – K a m l a - B a l a n s u b - b a s i n Sub-basin: Kamla-Balan India Nepal Return Period Male Female Total Male Female Total 2-Year 1,012 929 1,941 25 25 50 5-Year 1,150 1,057 2,207 29 30 59 10-Year 1,248 1,146 2,393 31 31 62 25-Year 1,370 1,258 2,628 33 33 66 50-Year 1,567 1,437 3,004 35 35 70 100-Year 1,567 1,437 3,004 37 37 73 Economic Losses: The total PML, combining losses for all the exposure classes under study for the Kamla- Balan sub-basin, varies from INR 1,567 million to INR 2,715 million due to 2-year and 100- year return period flood events respectively. Similarly, AAL of all the exposure classes is estimated at around INR 878 million. The risk assessment analysis shows that residential buildings bear the maximum losses followed by commercial buildings. Among all the exposure classes, residential buildings have a maximum AAL of INR 577 million that is 65.8% of total AAL, while commercial buildings and wheat crop are the second and third biggest contributors in total AAL with 27.5% and 2.1% contributions respectively. Health facilities and education institute have the least contributions to total AAL (0.1% for each). Estimated PML and AAL for various exposure classes for the sub-basin due to different return period flood events are shown in Table 6-25. T a b l e 6 - 2 5 : P M L a n d A A L d u e t o f l o o d e ve n t s o f d i f f e r e n t r e t u r n p e r i o d s f o r va r i o u s exposure classes – Kamla-Balan sub-basin Losses (Million INR): Kamla-Balan Sub-basin 100- Return Period 2-Year 5-Year 10-Year 25-Year 50-Year Year AAL Kamla-Balan Sub-basin 1,566.55 1,844.07 2,046.94 2,306.27 2,713.05 2,714.91 877.59 Residential 1,033.96 1,211.07 1,342.13 1,506.35 1,765.41 1,765.54 577.26 Commercial 432.93 506.22 560.77 634.01 743.81 744.73 241.71 Industrial 6.31 7.40 8.16 9.17 10.69 10.72 3.52 Building Education 2.17 2.54 2.81 3.15 3.67 3.67 1.21 Health 1.38 1.62 1.80 2.03 2.39 2.39 0.77 Others 20.44 23.90 26.43 29.83 35.02 35.08 11.41 Final Report: Volume I Confidential Page 207 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Losses (Million INR): Kamla-Balan Sub-basin 100- Return Period 2-Year 5-Year 10-Year 25-Year 50-Year Year AAL Road 4.31 5.39 6.06 7.03 8.90 8.92 2.51 Infrastructure Railway 5.64 6.47 6.99 8.30 10.08 10.08 3.12 Rice 18.43 24.07 27.95 32.46 40.55 40.94 11.08 Agriculture Wheat 31.42 40.75 47.16 54.33 67.28 67.54 18.77 Maize 9.54 14.65 16.67 19.60 25.26 25.29 6.23 A loss exceedance curve (LEC) has been generated using PML and their respective exceedance probabilities for all return period flood events for the sub-basin. Figure 6-15 shows the LEC for the sub-basin considering flood losses to all the exposure types. F i g u r e 6 - 1 5 : E x c e e d a n c e p r o b a b i l i t y c u r ve s h o w i n g t h e t o t a l l o s s e s d u e t o f l o o d f o r a l l e x p o s u r e t yp e s – K a m l a - B a l a n s u b - b a s i n The sub-basin level loss numbers can be further divided at state/province levels to provide a better understanding of the state-level losses. Bihar in India has the maximum AAL of INR 864 million which is around 98.5% of sub-basin’s total AAL. Rest of the provinces of Nepal have insignificant loss (less than 1.0%) contributions in the sub-basin level losses. Table 6-26 shows the state/province level break-up of the PML and AAL for the sub-basin. Final Report: Volume I Confidential Page 208 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 6 - 2 6 : S t a t e / P r o vi n c e w i s e b r e a k u p o f P M L a n d A A L – K a m l a - B a l a n s u b - b a s i n State/Province wise Losses (Thousand INR): Kamla-Balan Sub-basin Return Period Country State/Province 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL India Bihar 1,542,325.71 1,815,298.91 2,016,205.74 2,273,141.09 2,675,790.53 2,675,790.53 864,178.58 Central 14,229.71 17,495.75 18,781.64 20,024.16 21,861.49 22,954.56 7,994.03 Nepal Eastern 9,989.82 11,270.85 11,951.96 13,102.69 15,395.54 16,165.32 5,418.81 Final Report: Volume I Confidential Page 209 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia The study further identifies hotspot areas based on the significance of their contribution to AAL. Ghoghardiha, Darbhanga, Kalyanpur, Lakhnaur, and Madhepur in India are the worst affected blocks, which incur the highest annualized losses every year in the sub-basin. As maximum losses are incurred by the residential building class, the residential AAL map is plotted to identify the areas with maximum losses. The district/sub-district level residential AAL map is shown in Figure 6-16. Figure 6-16: AAL (million INR) due to floods for buildings: Residential – Kamla-Balan sub-basin Final Report: Volume I Confidential Page 210 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 6.1.9 YAMUNA SUB-BASIN Affected Population: The risk assessment for Yamuna sub-basin shows that the total number of persons affected due to a 2-year return period flood event is estimated at 7,308 thousand for India. This is around 9.1% of the total population of the area of India lying in the sub-basin. The total estimated number of persons affected in the sub-basin is given in Table 6-27. Table 6-27: Total number (in thousands) of persons affected due to various return p e r i o d f l o o d e ve n t s – Y a m u n a s u b - b a s i n Sub-basin: Yamuna India Return Period Male Female Total 2-Year 3,900 3,408 7,308 5-Year 4,683 4,092 8,776 10-Year 5,064 4,426 9,490 25-Year 5,382 4,704 10,086 50-Year 5,595 4,890 10,485 100-Year 5,815 5,082 10,897 Economic Losses: The total PML, combining losses for all the exposure classes under study for the Yamuna sub-basin, varies from INR 17,379 million to INR 36,965 million due to 2-year and 100-year return period flood events respectively. Similarly, AAL of all the exposure classes is estimated at around INR 10,517 million. The risk assessment analysis shows that residential buildings bear the maximum losses followed by commercial buildings. Among all the exposure classes, residential buildings have a maximum AAL of INR 8,025 million that is 76.3% of total AAL, while commercial buildings and wheat crop are the second and third biggest contributors in total AAL with 17.4% and 2.9% contributions respectively. Education institute, maize crop and railway network have the least contribution to total AAL with less than 0.1% contribution. Estimated PML and AAL for various exposure classes for the sub-basin due to different return period flood events are shown in Table 6-28. T a b l e 6 - 2 8 : P M L a n d A A L d u e t o f l o o d e ve n t s o f d i f f e r e n t r e t u r n p e r i o d s f o r va r i o u s exposure classes – Yamuna sub-basin Losses (Million INR): Yamuna Sub-basin Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Yamuna Sub-basin 17,379.25 23,445.11 26,995.87 31,367.91 34,156.87 36,964.97 10,516.61 Residential 13,215.71 17,925.74 20,662.75 24,074.36 26,227.96 28,427.02 8,025.09 Commercial 3,032.85 4,053.19 4,668.65 5,407.91 5,889.65 6,357.24 1,825.92 Industrial 180.79 245.64 285.35 333.46 365.44 395.44 110.20 Building Education 22.93 31.08 35.90 41.84 45.65 49.42 13.93 Health 27.67 37.46 43.32 50.46 55.09 59.64 16.80 Others 205.47 274.28 315.85 365.43 397.65 428.93 123.60 Final Report: Volume I Confidential Page 211 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Losses (Million INR): Yamuna Sub-basin Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Road 31.42 45.35 55.97 69.42 79.60 89.84 20.40 Infrastructure Railway 17.74 23.53 26.94 31.37 34.82 38.49 10.65 Rice 107.00 139.73 157.88 178.42 191.97 205.55 63.09 Agriculture Wheat 531.56 661.23 734.42 805.50 858.71 902.35 303.38 Maize 6.10 7.88 8.83 9.75 10.32 11.05 3.56 A loss exceedance curve (LEC) has been generated using PML and their respective exceedance probabilities for all return period flood events for the sub-basin. Figure 6-17 shows the LEC for the sub-basin considering flood losses to all the exposure types. F i g u r e 6 - 1 7 : E x c e e d a n c e p r o b a b i l i t y c u r ve s h o w i n g t h e t o t a l l o s s e s d u e t o f l o o d f o r a l l e x p o s u r e t yp e s – Y a m u n a s u b - b a s i n The sub-basin level loss numbers can be further divided at state/province levels to provide a better understanding of the state-level losses. Delhi in India has the maximum AAL of INR 6,737 million, which is around 64.1% of sub-basin’s total AAL, followed by Uttar Pradesh (28.5%) and Haryana (5.2%) in India. Rest of the states of India have insignificant loss contributions in the sub-basin level losses. Table 6-29 shows the state/province level break- up of the PML and AAL for the sub-basin. Final Report: Volume I Confidential Page 212 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 6 - 2 9 : S t a t e / P r o vi n c e w i s e b r e a k u p o f P M L a n d A A L – Y a m u n a s u b - b a s i n State/Province wise Losses (Thousand INR): Yamuna Sub-basin Return Period Country State/Province 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Haryana 946,027.96 1,209,494.72 1,337,179.85 1,510,217.80 1,651,352.32 1,766,587.74 549,372.41 Himachal Pradesh 61,718.12 70,494.17 75,329.75 80,378.07 83,876.42 87,385.95 33,404.82 Madhya Pradesh 4,941.38 7,121.73 8,396.17 10,004.49 11,184.17 12,403.44 3,134.32 NCT of Delhi 10,994,589.15 15,110,488.86 17,515,230.82 20,502,675.83 22,390,890.31 24,284,495.95 6,737,146.86 Rajasthan 303,540.71 371,688.80 425,925.00 458,906.63 489,189.74 513,292.07 172,989.54 Uttar Pradesh 5,027,502.60 6,633,124.67 7,590,084.07 8,760,800.20 9,484,575.40 10,254,055.09 2,999,570.60 India Uttarakhand 40,925.46 42,698.09 43,721.30 44,926.29 45,801.99 46,750.98 20,994.78 Final Report: Volume I Confidential Page 213 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia The study further identifies hotspot areas based on the significance of their contribution to AAL. Saraswati Vihar, Seelam Pur, Punjabi Bagh, Preet Vihar, and Gandhi Nagar of Delhi in India are the worst affected blocks, which incur the highest annualized losses every year in the sub-basin. As maximum losses are incurred by the residential building class, the residential AAL map is plotted to identify the areas with maximum losses. The district/sub- district level residential AAL map is shown in Figure 6-18. Figure 6-18: AAL (million INR) due to floods for buildings: Residential – Yamuna sub-basin Final Report: Volume I Confidential Page 214 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 6.1.10 RAMGANGA SUB-BASIN Affected Population: The risk assessment for Ramganga sub-basin shows that the total number of persons affected due to a 2-year return period flood event is estimated at 1,552 thousand for India. This is around 8.6% of the total population of the area of India lying in the sub-basin. The total estimated number of persons affected in the sub-basin is given in Table 6-30. Table 6-30: Total number (in thousands) of persons affected due to various return p e r i o d f l o o d e ve n t s – R a m g a n g a s u b - b a s i n Sub-basin: Ramganga India Return Period Male Female Total 2-Year 824 728 1,552 5-Year 1,051 929 1,979 10-Year 1,162 1,028 2,190 25-Year 1,274 1,127 2,401 50-Year 1,340 1,186 2,526 100-Year 1,401 1,241 2,642 Economic Losses: The total PML, combining losses for all the exposure classes under study for the Ramganga sub-basin, varies from INR 1,282 million to INR 2,597 million due to 2-year and 100-year return period flood events respectively. Similarly, AAL of all the exposure classes is estimated at around INR 772 million. The risk assessment analysis shows that residential buildings bear the maximum losses followed by commercial buildings. Among all the exposure classes, residential buildings have a maximum AAL of INR 379 million that is 49.1% of total AAL, while commercial buildings and wheat crop are the second and third biggest contributors in total AAL with 29.7% and 17.6% contribution respectively. Rice crop and health facilities have the least contributions to total AAL (0.1% each). Estimated PML and AAL for various exposure classes for the sub-basin due to different return period flood events are shown in Table 6-31. T a b l e 6 - 3 1 : P M L a n d A A L d u e t o f l o o d e ve n t s o f d i f f e r e n t r e t u r n p e r i o d s f o r va r i o u s exposure classes – Ramganga sub-basin Losses (Million INR): Ramganga Sub-basin 100- Return Period 2-Year 5-Year 10-Year 25-Year 50-Year Year AAL Ramganga Sub-basin 1,282.39 1,736.54 1,979.23 2,248.31 2,426.37 2,597.18 772.33 Residential 631.84 847.99 966.63 1,101.63 1,194.22 1,284.23 379.17 Commercial 382.26 513.44 583.12 661.44 716.03 767.53 229.08 Industrial 6.12 8.35 9.53 10.85 11.77 12.63 3.71 Building Education 1.90 2.49 2.80 3.17 3.42 3.66 1.12 Health 1.24 1.66 1.88 2.13 2.31 2.47 0.74 Others 21.99 29.49 33.49 37.99 41.12 44.08 13.17 Final Report: Volume I Confidential Page 215 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Losses (Million INR): Ramganga Sub-basin 100- Return Period 2-Year 5-Year 10-Year 25-Year 50-Year Year AAL Road 5.09 7.09 8.49 10.35 11.81 13.49 3.21 Infrastructure Railway 4.03 6.53 7.87 9.78 11.64 13.10 2.78 Rice 0.93 1.22 1.39 1.58 1.71 1.84 0.55 Agriculture Wheat 222.11 312.14 357.41 402.28 424.97 446.46 136.05 Maize 4.88 6.12 6.63 7.12 7.37 7.70 2.77 A loss exceedance curve (LEC) has been generated using PML and their respective exceedance probabilities for all return period flood events for the sub-basin. Figure 6-19 shows the LEC for the sub-basin considering flood losses to all the exposure types. F i g u r e 6 - 1 9 : E x c e e d a n c e p r o b a b i l i t y c u r ve s h o w i n g t h e t o t a l l o s s e s d u e t o f l o o d f o r a l l e x p o s u r e t yp e s – R a m g a n g a s u b - b a s i n The sub-basin level loss numbers can be further divided at state/province levels to provide a better understanding of the state-level losses. Uttar Pradesh in India has the maximum AAL of INR 741 million, which is around 95.9% of sub-basin’s total AAL, followed by Uttarakhand in India (4.1%). Table 6-32 shows the state/province level break-up of the PML and AAL for the sub-basin. Final Report: Volume I Confidential Page 216 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 6 - 3 2 : S t a t e / P r o vi n c e w i s e b r e a k u p o f P M L a n d A A L – R a m g a n g a s u b - b a s i n State/Province wise Losses (Thousand INR): Ramganga Sub-basin Return Period Country State/Province AAL 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year Uttar Pradesh 1,223,876.90 1,670,175.39 1,908,655.70 2,172,903.49 2,347,234.03 2,515,466.90 740,785.03 India Uttarakhand 58,516.50 66,360.22 70,579.24 75,404.04 79,133.52 81,715.86 31,542.30 Final Report: Volume I Confidential Page 217 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia The study further identifies hotspot areas based on the significance of their contribution to AAL. Bareilly, Jalalabad, Sawayajpur, Moradabad, and Rampur blocks of Uttar Pradesh in India are the worst affected blocks, which incur the highest annualized losses every year in the sub-basin. As maximum losses are incurred by the residential building class, the residential AAL map is plotted to identify the areas with maximum losses. The district/sub- district level residential AAL map is shown in Figure 6-20. Figure 6-20: AAL (million INR) due to floods for buildings: Residential – Ramganga sub-basin Final Report: Volume I Confidential Page 218 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 6.1.11 GOMTI SUB-BASIN Affected Population: The risk assessment for Gomati sub-basin shows that the total number of persons affected due to a 2-year return period flood event is estimated at 1,364 thousand for India. This is around 5.2% of the total population of the area of India lying in the sub-basin. The total estimated number of persons affected in the sub-basin is given in Table 6-33. Table 6-33: Total number (in thousands) of persons affected due to various return p e r i o d f l o o d e ve n t s – G o m t i s u b - b a s i n Sub-basin: Gomti India Return Period Male Female Total 2-Year 701 662 1,364 5-Year 831 784 1,616 10-Year 891 840 1,731 25-Year 976 921 1,897 50-Year 1,019 962 1,981 100-Year 1,056 997 2,053 Economic Losses: The total PML, combining losses for all the exposure classes under study for the Gomti sub- basin, varies from INR 1,549 million to INR 3,043 million due to 2-year and 100-year return period flood events respectively. Similarly, AAL of all the exposure classes is estimated at around INR 912 million. The risk assessment analysis shows that residential buildings bear the maximum losses followed by commercial buildings. Among all the exposure classes, residential buildings have a maximum AAL of INR 432 million that is 47.4% of total AAL, while commercial buildings and wheat crop are the second and third biggest contributors in total AAL with 35.1% and 9.7% contributions respectively. Health facilities and education institute have the least contributions to total AAL (0.1% each). Estimated PML and AAL for various exposure classes for the sub-basin due to different return period flood events are shown in Table 6-34. T a b l e 6 - 3 4 : P M L a n d A A L d u e t o f l o o d e ve n t s o f d i f f e r e n t r e t u r n p e r i o d s f o r va r i o u s exposure classes – Gomti sub-basin Losses (Million INR): Gomti Sub-basin 100- Return Period 2-Year 5-Year 10-Year 25-Year 50-Year Year AAL Gomti Sub-basin 1,549.08 1,994.84 2,277.31 2,602.16 2,843.63 3,042.68 911.75 Residential 720.33 951.49 1,103.63 1,270.43 1,393.46 1,495.16 431.76 Commercial 546.73 697.86 792.65 907.46 995.26 1,069.63 320.16 Industrial 8.17 10.57 12.06 13.83 15.24 16.39 4.83 Building Education 1.75 2.28 2.62 3.01 3.31 3.57 1.04 Health 1.67 2.17 2.50 2.88 3.18 3.41 0.99 Others 34.64 44.79 51.10 58.55 64.42 69.30 20.45 Final Report: Volume I Confidential Page 219 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Losses (Million INR): Gomti Sub-basin 100- Return Period 2-Year 5-Year 10-Year 25-Year 50-Year Year AAL Road 8.68 11.80 13.84 16.46 18.59 20.40 5.34 Infrastructure Railway 6.32 8.25 9.72 12.05 13.77 15.11 3.83 Rice 55.70 68.52 75.78 84.35 90.39 94.37 31.64 Agriculture Wheat 159.82 190.81 206.61 225.64 238.11 247.10 88.79 Maize 5.26 6.29 6.78 7.50 7.90 8.24 2.93 A loss exceedance curve (LEC) has been generated using PML and their respective exceedance probabilities for all return period flood events for the sub-basin. Figure 6-21 shows the LEC for the sub-basin considering flood losses to all the exposure types. F i g u r e 6 - 2 1 : E x c e e d a n c e p r o b a b i l i t y c u r ve s h o w i n g t h e t o t a l l o s s e s d u e t o f l o o d f o r a l l e x p o s u r e t yp e s – G o m t i s u b - b a s i n The sub-basin level loss numbers can be further divided at state/province levels to provide a better understanding of the state-level losses. Only two states of India come in this sub- basin. Uttar Pradesh has the maximum AAL of INR 911 million, which is almost equal (99.9) to the sub-basin’s total AAL. Uttarakhand has an insignificant loss contribution in the sub- basin level losses. Table 6-35 shows the state/province level break-up of the PML and AAL for the sub-basin. Final Report: Volume I Confidential Page 220 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 6 - 3 5 : S t a t e / P r o vi n c e w i s e b r e a k u p o f P M L a n d A A L – G o m t i s u b - b a s i n State/Province wise Losses (Thousand INR): Gomti Sub-basin Return Period Country State/Province AAL 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year Uttar Pradesh 1,548,581.90 1,994,238.86 2,276,647.47 2,601,423.05 2,842,803.18 3,041,834.02 911,472.14 India Uttarakhand 499.05 598.36 660.75 740.54 829.66 847.95 280.78 Final Report: Volume I Confidential Page 221 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia The study further identifies hotspot areas based on the significance of their contribution to AAL. Lucknow, Jaunpur, Pindra, Musafirkhana, and Kerakat blocks of Uttar Pradesh in India are the worst affected blocks, which incur the highest annualized losses every year in the sub-basin. As maximum losses are incurred by the residential building class, the residential AAL map is plotted to identify the areas with maximum losses. The district/sub-district level residential AAL map is shown in Figure 6-22. Figure 6-22: AAL (million INR) due to floods for buildings: Residential – Gomti sub- basin Final Report: Volume I Confidential Page 222 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 6.1.12 SONE SUB-BASIN Affected Population: The risk assessment for Sone sub-basin shows that the total number of persons affected due to a 2-year return period flood event is estimated at 456 thousand for India. This is around 3.5% of the total population of the area of India lying in the sub-basin. The total estimated number of persons affected in the sub-basin is given in Table 6-36. Table 6-36: Total number (in thousands) of persons affected due to various return p e r i o d f l o o d e ve n t s – S o n e s u b - b a s i n Sub-basin: Sone India Return Period Male Female Total 2-Year 238 218 456 5-Year 294 270 564 10-Year 318 292 609 25-Year 355 325 680 50-Year 394 362 756 100-Year 394 362 756 Economic Losses: The total PML, combining losses for all the exposure classes under study for the Sone sub- basin, varies from INR 749 million to INR 1,667 million due to 2-year and 100-year return period flood events respectively. Similarly, AAL of all the exposure classes is estimated at around INR 457 million. The risk assessment analysis shows that residential buildings bear the maximum losses followed by commercial buildings. Among all the exposure classes, residential buildings have a maximum AAL of INR 303 million that is 66.2% of total AAL, while commercial buildings and rice crop are the second and third biggest contributors in total AAL with 22.5% and 3.6% contributions respectively. Health facilities have the least contribution to total AAL (0.1%). Estimated PML and AAL for various exposure classes for the sub-basin due to different return period flood events are shown in Table 6-37. T a b l e 6 - 3 7 : P M L a n d A A L d u e t o f l o o d e ve n t s o f d i f f e r e n t r e t u r n p e r i o d s f o r va r i o u s exposure classes – Sone sub-basin Losses (Million INR): Sone Sub-basin Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Sone Sub-basin 749.40 1,007.84 1,177.89 1,386.22 1,667.34 1,667.42 457.35 Residential 492.93 668.02 785.48 927.57 1,118.70 1,118.77 302.74 Commercial 169.28 225.69 263.26 309.88 374.38 374.40 102.83 Industrial 2.63 3.48 4.04 4.73 5.69 5.69 1.59 Building Education 1.66 2.23 2.61 3.07 3.69 3.69 1.01 Health 1.03 1.39 1.64 1.93 2.33 2.33 0.63 Others 13.83 18.28 21.17 24.77 29.78 29.79 8.34 Final Report: Volume I Confidential Page 223 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Losses (Million INR): Sone Sub-basin Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Road 18.68 24.44 27.78 32.19 38.91 38.91 11.14 Infrastructure Railway 2.74 3.89 5.25 6.42 8.59 8.59 1.83 Rice 27.39 36.17 40.46 46.20 52.66 52.66 16.24 Agriculture Wheat 17.58 22.32 24.13 27.19 30.12 30.12 10.10 Maize 1.64 1.93 2.07 2.26 2.48 2.48 0.90 A loss exceedance curve (LEC) has been generated using PML and their respective exceedance probabilities for all return period flood events for the sub-basin. Figure 6-23 shows the LEC for the sub-basin considering flood losses to all the exposure types. F i g u r e 6 - 2 3 : E x c e e d a n c e p r o b a b i l i t y c u r ve s h o w i n g t h e t o t a l l o s s e s d u e t o f l o o d f o r a l l e x p o s u r e t yp e s – S o n e s u b - b a s i n The sub-basin level loss numbers can be further divided at state/province levels to provide a better understanding of the state-level losses. Bihar in India has the maximum AAL of INR 262 million, which is around 57.4% of sub-basin’s total AAL, followed by Madhya Pradesh (16.0%), Uttar Pradesh (15.8%) and Jharkhand (8.5%) in India. The state of Chattisgarh in India has insignificant loss (2.3%) contributions in the sub-basin level losses. Table 6-38 shows the state/province level break-up of the PML and AAL for the sub-basin. Final Report: Volume I Confidential Page 224 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 6 - 3 8 : S t a t e / P r o vi n c e w i s e b r e a k u p o f P M L a n d A A L – S o n e s u b - b a s i n State/Province wise Losses (Thousand INR): Sone Sub-basin Return Period Country State/Province AAL 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year Bihar 426,458.16 582,829.25 678,981.71 804,690.20 974,427.25 974,427.35 262,541.62 Chattisgarh 18,319.64 22,116.63 24,620.28 27,596.55 31,766.59 31,766.59 10,372.03 India Jharkhand 62,108.52 87,221.81 103,391.20 123,128.61 151,252.35 151,252.35 39,045.83 Madhya Pradesh 126,012.34 157,830.23 180,113.38 205,281.92 240,651.12 240,724.87 73,312.93 Uttar Pradesh 116,501.51 157,843.77 190,778.73 225,525.04 269,247.04 269,251.42 72,077.04 Final Report: Volume I Confidential Page 225 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia The study further identifies hotspot areas based on the significance of their contribution to AAL. Maner, Dehri, Bihta, and Barhara blocks of Bihar and Dudhi block of Uttar Pradesh in India are the worst affected blocks, which incur the highest annualized losses every year in the sub-basin. As maximum losses are incurred by the residential building class, the residential AAL map is plotted to identify the areas with maximum losses. The district/sub- district level residential AAL map is shown in Figure 6-24. Figure 6-24: AAL (million INR) due to floods for buildings: Residential – Sone sub- basin Final Report: Volume I Confidential Page 226 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 6.1.13 UPPER GANGES SUB-BASIN Affected Population: The risk assessment for Upper Ganges sub-basin shows that the total number of persons affected due to a 2-year return period flood event is estimated at 2,118 thousand for India. This is around 7.2% of the total population of the area of India lying in the sub-basin. The total estimated number of persons affected in the sub-basin is given in Table 6-39. Table 6-39: Total number (in thousands) of persons affected due to various return p e r i o d f l o o d e ve n t s – U p p e r G a n g e s s u b - b a s i n Sub-basin: Upper Ganges India Return Period Male Female Total 2-Year 1,127 991 2,118 5-Year 1,286 1,130 2,416 10-Year 1,369 1,204 2,573 25-Year 1,451 1,275 2,727 50-Year 1,503 1,321 2,824 100-Year 1,551 1,363 2,914 Economic Losses: The total PML, combining losses for all the exposure classes under study for the Upper Ganges sub-basin, varies from INR 2,640 million to INR 4,212 million due to 2-year and 100- year return period flood events respectively. Similarly, AAL of all the exposure classes is estimated at around INR 1,466 million. The risk assessment analysis shows that residential buildings bear the maximum losses followed by commercial buildings. Among all the exposure classes, residential buildings have a maximum AAL of INR 694 million that is 47.3% of total AAL, while commercial buildings and wheat crop are the second and third biggest contributors in total AAL with 32.9% and 14.8% contributions respectively. Health facilities have the least contribution to total AAL (0.1%). Estimated PML and AAL for various exposure classes for the sub-basin due to different return period flood events are shown in Table 6-40. T a b l e 6 - 4 0 : P M L a n d A A L d u e t o f l o o d e ve n t s o f d i f f e r e n t r e t u r n p e r i o d s f o r va r i o u s exposure classes – Upper Ganges sub-basin Losses (Million INR): Upper Ganges Sub-basin 100- Return Period 2-Year 5-Year 10-Year 25-Year 50-Year Year AAL Upper Ganges Sub-basin 2,640.22 3,127.20 3,416.20 3,743.82 3,979.27 4,211.80 1,466.55 Residential 1,245.45 1,476.46 1,619.88 1,789.56 1,916.24 2,043.15 693.86 Commercial 863.44 1,032.50 1,127.98 1,237.99 1,315.57 1,390.93 481.79 Industrial 13.08 15.67 17.13 18.83 20.01 21.15 7.31 Building Education 4.88 5.72 6.23 6.83 7.28 7.73 2.70 Health 3.46 4.09 4.47 4.94 5.28 5.62 1.92 Others 44.30 53.06 58.00 63.71 67.74 71.67 24.75 Final Report: Volume I Confidential Page 227 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Losses (Million INR): Upper Ganges Sub-basin 100- Return Period 2-Year 5-Year 10-Year 25-Year 50-Year Year AAL Road 26.25 29.87 32.50 36.13 38.78 41.57 14.34 Infrastructure Railway 6.73 8.83 10.26 11.91 13.35 14.72 4.04 Rice 16.20 20.33 22.79 25.62 27.61 29.55 9.34 Agriculture Wheat 398.92 460.06 494.74 524.85 543.14 560.58 216.90 Maize 17.51 20.61 22.20 23.45 24.27 25.14 9.61 A loss exceedance curve (LEC) has been generated using PML and their respective exceedance probabilities for all return period flood events for the sub-basin. Figure 6-25 shows the LEC for the sub-basin considering flood losses to all the exposure types. F i g u r e 6 - 2 5 : E x c e e d a n c e p r o b a b i l i t y c u r ve s h o w i n g t h e t o t a l l o s s e s d u e t o f l o o d f o r a l l e x p o s u r e t yp e s – U p p e r G a n g e s s u b - b a s i n The sub-basin level loss numbers can be further divided at state/province levels to provide a better understanding of the state-level losses. Uttar Pradesh in India has the maximum AAL of INR 1,320 million, which is around 90.0% of sub-basin’s total AAL, followed by Uttarakhand in India (10.0%). The state of Himachal Pradesh in India has an insignificant loss contribution in the sub-basin level losses. Table 6-41 shows the state/province level break-up of the PML and AAL for the sub-basin. Final Report: Volume I Confidential Page 228 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 6 - 4 1 : S t a t e / P r o vi n c e w i s e b r e a k u p o f P M L a n d A A L – U p p e r G a n g e s s u b - b a s i n State/Province wise Losses (Thousand INR): Upper Ganges Sub-basin Return Period Country State/Province AAL 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year Himachal Pradesh 27.90 28.25 28.75 29.23 29.23 29.61 14.09 India Uttar Pradesh 2,361,145.16 2,823,705.35 3,097,008.84 3,407,117.24 3,631,314.13 3,853,265.12 1,319,522.75 Uttarakhand 279,049.83 303,466.03 319,158.44 336,676.89 347,928.11 358,507.25 147,008.95 Final Report: Volume I Confidential Page 229 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia The study further identifies hotspot areas based on the significance of their contribution to AAL. Kaimganj, Mawana, Budaun, Gunnaur, and Dataganj blocks of Uttar Pradesh in India are the worst affected blocks, which incur the highest annualized losses every year in the sub-basin. As maximum losses are incurred by the residential building class, the residential AAL map is plotted to identify the areas with maximum losses. The district/sub-district level residential AAL map is shown in Figure 6-26. Figure 6-26: AAL (milli on INR) due to floods for buildings: Residential – Upper Ganges sub-basin Final Report: Volume I Confidential Page 230 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 6.1.14 TONS SUB-BASIN Affected Population: The risk assessment for Tons sub-basin shows that the total number of persons affected due to a 2-year return period flood event is estimated at 211 thousand for India. This is around 4.1% of the total population of the area of India lying in the sub-basin. The total estimated number of persons affected in the sub-basin is given in Table 6-42. Table 6-42: Total number (in thousands) of persons affected due to various return p e r i o d f l o o d e ve n t s – T o n s s u b - b a s i n Sub-basin: Tons India Return Period Male Female Total 2-Year 110 101 211 5-Year 144 132 276 10-Year 167 152 319 25-Year 190 173 363 50-Year 205 187 392 100-Year 218 199 417 Economic Losses: The total PML, combining losses for all the exposure classes under study for the Tons sub- basin, varies from INR 513 million to INR 1,143 million due to 2-year and 100-year return period flood events respectively. Similarly, AAL of all the exposure classes is estimated at around INR 314 million. The risk assessment analysis shows that residential buildings and commercial buildings bear the maximum losses (86.1% combined). Among all the exposure classes, residential and commercial buildings have a maximum AAL of INR 115 million and INR 156 million that is 36.4% and 49.6% of total AAL respectively. Wheat crop is the third highest contributor to AAL with 4.0% contribution. Maize crop has the least contribution to total AAL with less than 0.1% contribution. Estimated PML and AAL for various exposure classes for the sub-basin due to different return period flood events are shown in Table 6-43. T a b l e 6 - 4 3 : P M L a n d A A L d u e t o f l o o d e ve n t s o f d i f f e r e n t r e t u r n p e r i o d s f o r va r i o u s exposure classes – Tons sub-basin Losses (Million INR): Tons Sub-basin Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Tons Sub-basin 512.68 697.64 826.29 965.70 1,057.41 1,143.19 314.46 Residential 189.02 252.46 298.18 346.29 376.96 406.56 114.60 Commercial 251.47 348.76 415.04 487.75 534.97 579.44 156.12 Industrial 3.55 4.91 5.84 6.85 7.50 8.12 2.20 Building Education 1.00 1.26 1.46 1.65 1.78 1.91 0.58 Health 0.53 0.71 0.84 0.98 1.08 1.16 0.32 Others 18.62 25.18 29.74 34.61 37.80 40.80 11.37 Final Report: Volume I Confidential Page 231 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Losses (Million INR): Tons Sub-basin Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Road 8.03 10.52 12.25 14.54 16.76 18.28 4.84 Infrastructure Railway 0.80 1.91 3.04 4.79 6.71 8.22 0.86 Rice 18.23 23.93 27.62 31.49 34.07 36.32 10.85 Agriculture Wheat 21.37 27.93 32.21 36.68 39.68 42.28 12.69 Maize 0.06 0.07 0.08 0.08 0.09 0.09 0.03 A loss exceedance curve (LEC) has been generated using PML and their respective exceedance probabilities for all return period flood events for the sub-basin. Figure 6-27 shows the LEC for the sub-basin considering flood losses to all the exposure types. F i g u r e 6 - 2 7 : E x c e e d a n c e p r o b a b i l i t y c u r ve s h o w i n g t h e t o t a l l o s s e s d u e t o f l o o d f o r a l l e x p o s u r e t yp e s – T o n s s u b - b a s i n The sub-basin level loss numbers can be further divided at state/province levels to provide a better understanding of the state-level losses. Uttar Pradesh and Madhya Pradesh in India are the only states, which have exposure in this sub-basin. Out of these two states, Uttar Pradesh has the maximum AAL of INR 247 million, which is around 78.7% of sub-basin’s total AAL, followed by Madhya Pradesh (21.3%). Table 6-44 shows the state/province level break-up of the PML and AAL for the sub-basin. Final Report: Volume I Confidential Page 232 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 6 - 4 4 : S t a t e / P r o vi n c e w i s e b r e a k u p o f P M L a n d A A L – T o n s s u b - b a s i n State/Province wise Losses (Thousand INR): Tons Sub-basin Return Period Country State/Province AAL 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year Madhya Pradesh 124,220.22 139,515.44 152,948.65 161,803.01 169,903.88 176,764.54 67,097.27 India Uttar Pradesh 388,455.61 558,127.39 673,339.92 803,899.26 887,506.86 966,425.74 247,367.13 Final Report: Volume I Confidential Page 233 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia The study further identifies hotspot areas based on the significance of their contribution to AAL. Karchhana, Meja, Bara and Koraon blocks of Uttar Pradesh and Sirmour block of Madhya Pradesh in India are the worst affected blocks, which incur the highest annualized losses every year in the sub-basin. As maximum losses are incurred by the residential building class, the residential AAL map is plotted to identify the areas with maximum losses. The district/sub-district level residential AAL map is shown in Figure 6-28. Figure 6-28: AAL (million INR) due to floods for buildings: Residential – Tons sub- basin Final Report: Volume I Confidential Page 234 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 6.1.15 SIND SUB-BASIN Affected Population: The risk assessment for Sind sub-basin shows that the total number of persons affected due to a 2-year return period flood event is estimated at 286 thousand for India. This is around 3.6% of the total population of the area of India lying in the sub-basin. The total estimated number of persons affected in the sub-basin is given in Table 6-45. Table 6-45: Total number (in thousands) of persons affected due to various return p e r i o d f l o o d e ve n t s – S i n d s u b - b a s i n Sub-basin: Sind India Return Period Male Female Total 2-Year 154 132 286 5-Year 189 161 349 10-Year 207 176 383 25-Year 226 192 418 50-Year 238 203 441 100-Year 250 213 463 Economic Losses: The total PML, combining losses for all the exposure classes under study for the Sind sub- basin, varies from INR 570 million to INR 1011 million due to 2-year and 100-year return period flood events respectively. Similarly, AAL of all the exposure classes is estimated at around INR 329 million. The risk assessment analysis shows that residential buildings bear the maximum losses followed by commercial buildings. Among all the exposure classes, residential buildings have a maximum AAL of INR 201 million that is 61.1% of total AAL, while commercial buildings and wheat crop are the second and third biggest contributors in total AAL with 26.2% and 6.8% contributions respectively. Health facilities have the least contribution to total AAL (0.1%). Estimated PML and AAL for various exposure classes for the sub-basin due to different return period flood events are shown in Table 6-46. T a b l e 6 - 4 6 : P M L a n d A A L d u e t o f l o o d e ve n t s o f d i f f e r e n t r e t u r n p e r i o d s f o r va r i o u s exposure classes – Sind sub-basin Losses (Million INR): Sind Sub-basin Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Sind Sub-basin 570.49 722.46 803.98 896.57 953.66 1,011.33 329.21 Residential 350.09 441.46 489.68 544.16 577.01 610.56 201.31 Commercial 147.95 189.45 212.47 239.07 255.94 272.75 86.15 Industrial 1.83 2.32 2.59 2.89 3.08 3.26 1.06 Building Education 1.28 1.63 1.82 2.03 2.16 2.29 0.74 Health 0.54 0.70 0.78 0.88 0.94 1.00 0.32 Others 15.19 19.30 21.55 24.12 25.77 27.41 8.79 Final Report: Volume I Confidential Page 235 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Losses (Million INR): Sind Sub-basin Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Road 8.40 11.07 12.59 14.50 15.91 17.37 5.01 Infrastructure Railway 1.79 2.38 2.70 3.24 3.40 3.61 1.07 Rice 3.57 4.57 5.10 5.66 6.02 6.36 2.07 Agriculture Wheat 39.41 49.01 54.05 59.31 62.67 65.90 22.44 Maize 0.44 0.57 0.64 0.71 0.76 0.81 0.26 A loss exceedance curve (LEC) has been generated using PML and their respective exceedance probabilities for all return period flood events for the sub-basin. Figure 6-29 shows the LEC for the sub-basin considering flood losses to all the exposure types. F i g u r e 6 - 2 9 : E x c e e d a n c e p r o b a b i l i t y c u r ve s h o w i n g t h e t o t a l l o s s e s d u e t o f l o o d f o r a l l e x p o s u r e t yp e s – S i n d s u b - b a s i n The sub-basin level loss numbers can be further divided at state/province levels to provide a better understanding of the state-level losses. Madhya Pradesh and Uttar Pradesh in India are the only states which have exposure in this sub-basin. Out of these two states, Madhya Pradesh has the maximum AAL of INR 258 million, which is around 78.5% of sub-basin’s total AAL, followed by Uttar Pradesh (21.5%). Table 6-47 shows the state/province level break-up of the PML and AAL for the sub-basin. Final Report: Volume I Confidential Page 236 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 6 - 4 7 : S t a t e / P r o vi n c e w i s e b r e a k u p o f P M L a n d A A L – S i n d s u b - b a s i n State/Province wise Losses (Thousand INR): Sind Sub-basin Return Period Country State/Province AAL 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year Madhya Pradesh 454,254.00 564,974.16 620,646.85 683,106.27 722,786.48 762,395.91 258,526.38 India Uttar Pradesh 116,240.06 157,490.40 183,332.99 213,467.48 230,874.37 248,938.39 70,688.51 Final Report: Volume I Confidential Page 237 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia The study further identifies hotspot areas based on the significance of their contribution to AAL. Bhind, Morena, Porsa, and Ambah blocks of Madhya Pradesh and Madhogarh block of Uttar Pradesh in India are the worst affected blocks, which incur the highest annualized losses every year in the sub-basin. As maximum losses are incurred by the residential building class, the residential AAL map is plotted to identify the areas with maximum losses. The district/sub-district level residential AAL map is shown in Figure 6-30. Figure 6-30: AAL (million INR) due to floods for buildings: Residential – Sind sub- basin Final Report: Volume I Confidential Page 238 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 6.1.16 KEN SUB-BASIN Affected Population: The risk assessment for Ken sub-basin shows that the total number of persons affected due to a 2-year return period flood event is estimated at 113 thousand for India. This is around 2.1% of the total population of the area of India lying in the sub-basin. The total estimated number of persons affected in the sub-basin is given in Table 6-48. Table 6-48: Total number (in thousands) of persons affected due to various return p e r i o d f l o o d e ve n t s – K e n s u b - b a s i n Sub-basin: Ken India Return Period Male Female Total 2-Year 60 53 113 5-Year 79 70 149 10-Year 92 81 173 25-Year 107 94 201 50-Year 117 103 221 100-Year 126 111 237 Economic Losses: The total PML, combining losses for all the exposure classes under study for the Ken sub- basin, varies from INR 295 million to INR 750 million due to 2-year and 100-year return period flood events respectively. Similarly, AAL of all the exposure classes is estimated at around INR 187 million. The risk assessment analysis shows that residential buildings bear the maximum losses followed by commercial buildings. Among all the exposure classes, residential buildings have a maximum AAL of INR 112 million that is 59.6% of total AAL, while commercial buildings and wheat crop are the second and third biggest contributors in total AAL with 28.8% and 4.3% contributions respectively. Maize crop has the least contribution to total AAL (less than 0.1%). Estimated PML and AAL for various exposure classes for the sub-basin due to different return period flood events are shown in Table 6-49. T a b l e 6 - 4 9 : P M L a n d A A L d u e t o f l o o d e ve n t s o f d i f f e r e n t r e t u r n p e r i o d s f o r va r i o u s exposure classes – Ken sub-basin Losses (Million INR): Ken Sub-basin Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Ken Sub-basin 295.47 416.49 505.01 612.36 685.36 749.99 187.19 Residential 175.58 248.42 301.93 367.28 411.00 450.08 111.59 Commercial 85.04 120.58 145.81 176.83 197.49 216.13 53.99 Industrial 0.90 1.26 1.52 1.84 2.05 2.25 0.57 Building Education 0.73 1.02 1.23 1.48 1.66 1.81 0.46 Health 0.32 0.45 0.55 0.67 0.75 0.83 0.20 Others 5.54 7.71 9.24 11.07 12.35 13.48 3.47 Final Report: Volume I Confidential Page 239 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Losses (Million INR): Ken Sub-basin Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Road 9.25 13.17 16.89 20.84 24.35 27.10 6.04 Infrastructure Railway 1.23 1.66 1.98 2.21 2.70 2.91 0.75 Rice 3.31 4.26 4.87 5.65 6.15 6.57 1.95 Agriculture Wheat 13.48 17.83 20.85 24.33 26.67 28.65 8.12 Maize 0.10 0.13 0.15 0.16 0.18 0.19 0.06 A loss exceedance curve (LEC) has been generated using PML and their respective exceedance probabilities for all return period flood events for the sub-basin. Figure 6-31 shows the LEC for the sub-basin considering flood losses to all the exposure types. F i g u r e 6 - 3 1 : E x c e e d a n c e p r o b a b i l i t y c u r ve s h o w i n g t h e t o t a l l o s s e s d u e t o f l o o d f o r a l l e x p o s u r e t yp e s – K e n s u b - b a s i n The sub-basin level loss numbers can be further divided at state/province levels to provide a better understanding of the state-level losses. Uttar Pradesh and Madhya Pradesh in India are the only states, which have exposure in this sub-basin. Out of these two states, Uttar Pradesh has the maximum AAL of INR 104 million, which is around 55.9% of sub-basin’s total AAL, followed by Madhya Pradesh (44.1%). Table 6-50 shows the state/province level break-up of the PML and AAL for the sub-basin. Final Report: Volume I Confidential Page 240 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 6 - 5 0 : S t a t e / P r o vi n c e w i s e b r e a k u p o f P M L a n d A A L – K e n s u b - b a s i n State/Province wise Losses (Thousand INR): Ken Sub-basin Return Period Country State/Province AAL 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year Madhya Pradesh 136,881.71 181,338.06 213,745.22 245,759.79 274,016.84 296,486.03 82,643.26 India Uttar Pradesh 158,592.98 235,148.87 291,268.11 366,601.25 411,340.72 453,506.54 104,549.37 Final Report: Volume I Confidential Page 241 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia The study further identifies hotspot areas based on the significance of their contribution to AAL. Banda, Hamirpur, and Maudaha blocks of Uttar Pradesh and Damoh and Rajnagar blocks of Madhya Pradesh in India are the worst affected blocks, which incur the highest annualized losses every year in the sub-basin. As maximum losses are incurred by the residential building class, the residential AAL map is plotted to identify the areas with maximum losses. The district/sub-district level residential AAL map is shown in Figure 6-32. Figure 6-32: AAL (million INR) due to floods for buildings: Residential – Ken sub- basin Final Report: Volume I Confidential Page 242 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 6.1.17 BETWA SUB-BASIN Affected Population: The risk assessment for Betwa sub-basin shows that the total number of persons affected due to a 2-year return period flood event is estimated at 184 thousand for India. This is around 1.7% of the total population of the area of India lying in the sub-basin. The total estimated number of persons affected in the sub-basin is given in Table 6-51. Table 6-51: Total number (in thousands) of persons affected due to various return p e r i o d f l o o d e ve n t s – B e t w a s u b - b a s i n Sub-basin: Betwa India Return Period Male Female Total 2-Year 97 87 184 5-Year 123 109 232 10-Year 138 122 260 25-Year 156 139 295 50-Year 168 150 318 100-Year 179 159 339 Economic Losses: The total PML, combining losses for all the exposure classes under study for the Betwa sub- basin, varies from INR 416 million to INR 1,077 million due to 2-year and 100-year return period flood events respectively. Similarly, AAL of all the exposure classes is estimated at around INR 264 million. The risk assessment analysis shows that residential buildings bear the maximum losses followed by commercial buildings. Among all the exposure classes, residential buildings have a maximum AAL of INR 147 million that is 55.6% of total AAL, while commercial buildings and wheat crop are the second and third biggest contributors in total AAL with 31.3% and 7.8% contributions respectively. Maize and health facilities have the least contribution to total AAL (0.2% each). Estimated PML and AAL for various exposure classes for the sub-basin due to different return period flood events are shown in Table 6-52. T a b l e 6 - 5 2 : P M L a n d A A L d u e t o f l o o d e ve n t s o f d i f f e r e n t r e t u r n p e r i o d s f o r va r i o u s exposure classes – Betwa sub-basin Losses (Million INR): Betwa Sub-basin Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Betwa Sub-basin 416.01 592.80 712.33 860.14 971.53 1,077.39 264.52 Residential 229.35 330.16 399.52 484.80 548.78 609.86 147.07 Commercial 129.72 186.09 223.13 269.29 305.36 339.04 82.74 Industrial 1.43 2.03 2.43 2.92 3.30 3.66 0.90 Building Education 1.16 1.67 2.03 2.49 2.84 3.17 0.75 Health 0.71 1.02 1.23 1.49 1.70 1.89 0.46 Others 9.06 12.84 15.31 18.40 20.72 22.93 5.72 Final Report: Volume I Confidential Page 243 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Losses (Million INR): Betwa Sub-basin Return Period 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year AAL Road 6.03 10.02 13.36 17.43 20.05 22.79 4.39 Infrastructure Railway 1.87 2.65 3.24 3.94 4.49 5.13 1.20 Rice 0.47 0.63 0.72 0.84 0.91 0.97 0.28 Agriculture Wheat 35.38 44.68 50.26 57.36 62.15 66.64 20.53 Maize 0.83 1.01 1.09 1.19 1.25 1.30 0.46 A loss exceedance curve (LEC) has been generated using PML and their respective exceedance probabilities for all return period flood events for the sub-basin. Figure 6-33 shows the LEC for the sub-basin considering flood losses to all the exposure types. F i g u r e 6 - 3 3 : E x c e e d a n c e p r o b a b i l i t y c u r ve s h o w i n g t h e t o t a l l o s s e s d u e t o f l o o d f o r a l l e x p o s u r e t yp e s – B e t w a s u b - b a s i n The sub-basin level loss numbers can be further divided at state/province levels to provide a better understanding of the state-level losses. Uttar Pradesh and Madhya Pradesh in India are the only states, which have exposure in this sub-basin. Out of these two states, Uttar Pradesh has the maximum AAL of INR 196 million, which is around 73.9% of sub-basin’s total AAL, followed by Madhya Pradesh (26.1%). Table 6-53 shows the state/province level break-up of the PML and AAL for the sub-basin. Final Report: Volume I Confidential Page 244 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 6 - 5 3 : S t a t e / P r o vi n c e w i s e b r e a k u p o f P M L a n d A A L – B e t w a s u b - b a s i n State/Province wise Losses (Thousand INR): Betwa Sub-basin Return Period Country State/Province AAL 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year Madhya Pradesh 110,426.69 154,233.03 182,238.92 217,719.72 241,736.32 265,736.52 68,914.77 India Uttar Pradesh 305,587.12 438,571.23 530,086.55 642,422.70 729,793.14 811,653.82 195,601.37 Final Report: Volume I Confidential Page 245 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia The study further identifies hotspot areas based on the significance of their contribution to AAL. Hamirpur, Rath, Garautha, Moth, and Orai blocks of Uttar Pradesh in India are the worst affected blocks, which incur the highest annualized losses every year in the sub-basin. As maximum losses are incurred by the residential building class, the residential AAL map is plotted to identify the areas with maximum losses. The district/sub-district level residential AAL map is shown in Figure 6-34. Figure 6-34: AAL (million INR) due to floods for buildings: Residential – Betwa sub- basin Final Report: Volume I Confidential Page 246 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 6.1.18 CHAMBAL SUB-BASIN Affected Population: The risk assessment for Chambal sub-basin shows that the total number of persons affected due to a 2-year return period flood event is estimated at 699 thousand for India. This is around 1.9% of the total population of the area of India lying in the sub-basin. The total estimated number of persons affected in the sub-basin is given in Table 6-54. Table 6-54: Total number (in thousands) of persons affected due to various return p e r i o d f l o o d e ve n t s – C h a m b a l s u b - b a s i n Sub-basin: Chambal India Return Period Male Female Total 2-Year 364 335 699 5-Year 460 424 884 10-Year 514 473 987 25-Year 570 525 1,095 50-Year 609 562 1,171 100-Year 646 596 1,241 Economic Losses: The total PML, combining losses for all the exposure classes under study for the Chambal sub-basin, varies from INR 1,627 million to INR 3,162 million due to 2-year and 100-year return period flood events respectively. Similarly, AAL of all the exposure classes is estimated at around INR 960 million. The risk assessment analysis shows that residential buildings bear the maximum losses followed by commercial buildings. Among all the exposure classes, residential buildings have a maximum AAL of INR 439 million that is 45.7% of total AAL, while commercial buildings and wheat crop are the second and third biggest contributors in total AAL with 37.2% and 7.8% contributions respectively. Health facilities have the least contribution to total AAL (0.1%). Estimated PML and AAL for various exposure classes for the sub-basin due to different return period flood events are shown in Table 6-55. T a b l e 6 - 5 5 : P M L a n d A A L d u e t o f l o o d e ve n t s o f d i f f e r e n t r e t u r n p e r i o d s f o r va r i o u s exposure classes – Chambal sub-basin Losses (Million INR): Chambal Sub-basin 100- Return Period 2-Year 5-Year 10-Year 25-Year 50-Year Year AAL Chambal Sub-basin 1,627.32 2,122.68 2,408.87 2,720.28 2,944.51 3,162.23 960.47 Residential 732.66 976.74 1,119.19 1,275.16 1,388.72 1,498.74 439.00 Commercial 612.65 785.08 883.61 988.78 1,063.77 1,136.22 357.09 Industrial 13.71 16.92 18.79 20.81 22.24 23.63 7.81 Building Education 3.57 4.78 5.48 6.24 6.78 7.30 2.14 Health 2.07 2.68 3.04 3.42 3.69 3.96 1.22 Others 60.66 78.14 88.23 99.14 106.89 114.42 35.50 Final Report: Volume I Confidential Page 247 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Losses (Million INR): Chambal Sub-basin 100- Return Period 2-Year 5-Year 10-Year 25-Year 50-Year Year AAL Road 27.95 36.59 41.45 48.04 53.22 58.82 16.61 Infrastructure Railway 16.51 19.52 21.72 24.17 26.33 28.49 9.24 Rice 7.00 9.50 10.94 12.56 13.58 14.65 4.24 Agriculture Wheat 128.81 165.35 185.82 207.97 222.85 237.38 75.09 Maize 21.72 27.37 30.59 33.99 36.44 38.62 12.52 A loss exceedance curve (LEC) has been generated using PML and their respective exceedance probabilities for all return period flood events for the sub-basin. Figure 6-35 shows the LEC for the sub-basin considering flood losses to all the exposure types. F i g u r e 6 - 3 5 : E x c e e d a n c e p r o b a b i l i t y c u r ve s h o w i n g t h e t o t a l l o s s e s d u e t o f l o o d f o r a l l e x p o s u r e t yp e s – C h a m b a l s u b - b a s i n The sub-basin level loss numbers can be further divided at state/province levels to provide a better understanding of the state-level losses. Rajasthan, Madhya Pradesh, and Uttar Pradesh in India are the only states, which have exposure in this sub-basin. Out of these three states, Rajasthan has the maximum AAL of INR 541 million, which is around 56.3% of sub-basin’s total AAL, followed by Madhya Pradesh (31.1%) and Uttar Pradesh (12.6%). Table 6-56 shows the state/province level break-up of the PML and AAL for the sub-basin. Final Report: Volume I Confidential Page 248 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 6 - 5 6 : S t a t e / P r o vi n c e w i s e b r e a k u p o f P M L a n d A A L – C h a m b a l s u b - b a s i n State/Province wise Losses (Thousand INR): Chambal Sub-basin Return Period Country State/Province AAL 2-Year 5-Year 10-Year 25-Year 50-Year 100-Year Madhya Pradesh 493,233.54 668,212.06 767,715.11 875,093.65 958,235.60 1,032,871.19 298,267.12 India Rajasthan 920,326.38 1,187,166.50 1,348,613.14 1,530,274.10 1,659,010.02 1,788,180.95 540,808.74 Uttar Pradesh 213,764.31 267,306.35 292,537.68 314,916.26 327,264.79 341,174.08 121,394.90 Final Report: Volume I Confidential Page 249 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia The study further identifies hotspot areas based on the significance of their contribution to AAL. Phulera (Hq. Sambhar), Nawa, and Rawatbhata blocks of Rajasthan, Bah block of Uttar Pradesh and Manasa block of Madhya Pradesh in India are the worst affected blocks, which incur the highest annualized losses every year in the sub-basin. As maximum losses are incurred by the residential building class, the residential AAL map is plotted to identify the areas with maximum losses. The district/sub-district level residential AAL map is shown in Figure 6-36. Figure 6-36: AAL (million INR) due to floods for buildings: Residential – Chambal sub-basin Final Report: Volume I Confidential Page 250 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 7 Conclusions and Recommendations This chapter provides conclusions based on the findings of the risk assessment carried out as a part of this assignment. On the whole, the geographically affected areas in the basin would do well with further macro-level studies especially in the light of the general recommendations presented at the end of this section. Apart from this, some broad recommendations have also been provided that could be further investigated at a specialized-study level in the areas of agriculture, and railway and road network losses. Geographically affected areas in the basin: Ganges Basin has very high flood frequency. Certain states/provinces of India, Nepal, and Bangladesh in the basin witness very high losses due to floods every year and need immediate attention of the concerned authorities. States of Bihar, Uttar Pradesh, Delhi, and West Bengal in India witness maximum flood losses in Ganges Basin, whereas, in Bangladesh, the States of Kushtia and Rajshahi bear the maximum losses. Similarly, in Nepal, the Eastern province has the maximum share in losses, followed by the Mid Western and Western provinces. State of Bihar in India witnesses huge losses every year. These flood losses come from the various sub-basins the state is part of, i.e. Lower Ganges, Bagmati, Kosi, Gandak, Mahananda, Ghagra, Kamla-Balan, and Sone. Substantial losses are estimated in all these sub-basins, though the Lower Ganges and Bagmati sub-basins have the maximum share in terms of economic losses and the number of flood affected persons. In Lower Ganges sub- basin alone, Bihar is likely to see annual losses of around INR 7,994 million while the total annualized losses are estimated around INR 14,811 million, which is around 37% of the total losses incurred by India due to floods in Ganges Basin. Together, the residential and commercial sectors have the maximum share in total losses. Anytime once in a given 2-year period, 18.12 million persons are likely to get affected in Bihar due to floods, which is around 40% of the total flood affected persons (45.4 million) in the Indian part of the Ganges basin. The second most severely affected state in India is Uttar Pradesh (UP). Its exposure lies in almost all the sub-basins of Ganges i.e. Betwa, Chambal, Gandak, Ghagra, Gomti, Ken, Lower Ganges, Middle Ganges, Ramganga, Sind, Sone, Tons, Upper Ganges, and Yamuna. Middle Ganges, Yamuna and Ghagra sub-basins have the maximum share in losses and flood affected persons in Uttar Pradesh. In Middle Ganges sub-basin alone, UP is likely to witness annual losses of around INR 4,225 million while the total annualized losses are estimated to be around INR 13,747 million, which is around 34% of the total losses incurred in the Indian part of the Ganges Basin. Anytime once in a given 2-year period, 17.3 million persons are likely to get affected in this state due to floods, which, is around 38%.of the total flood affected persons in the Indian part of the Ganges basin. The third most severely affected state in India is Delhi, which lies completely in the catchment area of Yamuna sub-basin. Delhi is likely to witness annual losses of around INR 6,737 million, which is about 17% of the total losses incurred in the Indian part of the Ganges Basin. Anytime once in a given 2-year period 3.8 million persons are likely to are likely to get affected due to floods in Delhi, which is around 8.3% of the total flood affected persons in the Indian part of the Ganges basin. The state of West Bengal in India also witnesses substantial losses due to floods in the Lower Ganges and Mahananda sub-basins. It is likely to witness annual losses of around INR 2,039 million, which is about 5% of the total losses, incurred in the Indian part of the Ganges Basin, though only 14% of its area lies within the basin. Anytime once in a given 2- year period 3.5 million persons are likely to get affected due to floods in West Bengal, which is around 7.7% of the total flood affected persons in the Indian part of the Ganges basin. Final Report: Volume I Confidential Page 251 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Rajasthan is likely to witness an annualized loss of INR 714 million and 672 thousand persons are likely to be flood affected anytime once in a given 2-year period. It is followed by Haryana with likely annual losses of INR 549 million and 580 thousand persons who are likely to be affected by floods anytime once in a given 2-year period. In Bangladesh, States of Rajshahi and Kushtia bear the maximum flood losses among all the four states of Bangladesh in Ganges Basin. Rajshahi and Kushtia have estimated annualized losses of INR 926 million and 294 million, which is 71% and 23% of the total annualized losses of Bangladesh in the Ganges Basin. Pabna and Dinajpur states share 3% each in total annualized losses of Bangladesh respectively. In any given 2-year period, 998 thousand and 281 thousand people are likely to be affected due to floods in Rajshahi and Kushtia respectively, which is 70% and 20% of the total persons affected in the Ganges Basin in Bangladesh. In Nepal, the Mid Western province is at maximum risk followed by the Eastern province. They bear the maximum flood losses in Nepal’s part of the Ganges Basin. Mid Western and Eastern Provinces are likely to have annualized losses of INR 64 million and 52 million, which is 27% and 22% of the total annualized losses of Nepal in the Ganges Basin. Central, Western, and Far Western provinces share 19%, 19%, and 14% of total annualized losses of Nepal respectively. In any given 2-year period,147 thousand, 127 thousand and 58 thousand people are likely to get affected in Eastern, Central and Mid Western Provinces due to floods respectively, which is 34%, 29% and 13% of the total persons affected in the Ganges Basin in Nepal. Following are the most severely affected areas due to flood for various exposure classes: Affected population: Lower Ganges sub-basin needs immediate attention with regards to affected population. The table below (Table 7-1) shows the top seven blocks/districts that have significant flood affected populations for a 2-year return period flood event. T a b l e 7 - 1 : D i s t r i c t s / s u b - d i s t r i c t s h a vi n g m a x i m u m a f f e c t e d p e r s o n s Sr. No. Block District State Country Sub basin 1 Patna Rural Patna Bihar India Lower Ganges 2 Kustia Kushtia Bangladesh Lower Ganges 3 Dinapur-cum-khagaul Patna Bihar India Lower Ganges 4 Nawabganj Rajshahi Bangladesh Lower Ganges 5 Kaliachak - III Maldah West Bengal India Lower Ganges 6 Patepur Vaishali Bihar India Lower Ganges 7 Kurhani Muzaffarpur Bihar India Lower Ganges Residential buildings: Lower Ganges sub-basin has the maximum share in losses for residential buildings considering all return period flood events. The table below (Table 7-2) shows the top seven blocks/districts that have a significant number of flood-affected residential buildings due to 2- year return period. T a b l e 7 - 2 : D i s t r i c t s / s u b - d i s t r i c t s h a vi n g m a x i m u m r e s i d e n t i a l b u i l d i n g l o s s e s Sr. No. Block District State Country Sub basin 1 Patna Rural Patna Bihar India Lower Ganges 2 Dinapur-cum-khagaul Patna Bihar India Lower Ganges 3 Munger Munger Bihar India Lower Ganges Final Report: Volume I Confidential Page 252 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Sr. No. Block District State Country Sub basin 4 Mokameh Patna Bihar India Lower Ganges 5 Colgong Bhagalpur Bihar India Lower Ganges 6 Phulwari Patna Bihar India Lower Ganges 7 Patepur Vaishali Bihar India Lower Ganges Commercial buildings: For Commercial building also, the Lower Ganges sub-basin has the maximum share in losses considering all return period flood events. The table below (Table 7-3) shows the top seven blocks/districts that have a significant number of flood-affected commercial buildings due to 2-year return period. T a b l e 7 - 3 : D i s t r i c t s / s u b - d i s t r i c t s h a vi n g m a x i m u m c o m m e r c i a l b u i l d i n g l o s s e s Sr. No. Block District State Country Sub basin 1 Mohiuddinagar Samastipur Bihar India Lower Ganges 2 Patepur Vaishali Bihar India Lower Ganges 3 Mokameh Patna Bihar India Lower Ganges 4 Raghopur Vaishali Bihar India Lower Ganges 5 Dinapur-cum-khagaul Patna Bihar India Lower Ganges 6 Kustia Kustia Kustia Bangladesh Lower Ganges 7 Nawabganj Nawabganj Rajshahi Bangladesh Lower Ganges Industrial buildings: The pattern is repeated for industrial buildings where Lower Ganges sub-basin has the maximum share in losses considering all return period flood events. The table below (Table 7-4) shows the top seven blocks/districts that have a significant number of flood-affected industrial buildings due to 2-year return period. T a b l e 7 - 4 : D i s t r i c t s / s u b - d i s t r i c t s h a vi n g m a x i m u m i n d u s t r i a l b u i l d i n g l o s s e s Sr. No. Block District State Country Sub basin 1 Manikchak Maldah West Bengal India Lower Ganges 2 Mokameh Patna Bihar India Lower Ganges 3 Patepur Vaishali Bihar India Lower Ganges 4 Dinapur-cum-khagaul Patna Bihar India Lower Ganges 5 Pandarak Patna Bihar India Lower Ganges 6 Kustia Kustia Kustia Bangladesh Lower Ganges 7 Nawabganj Nawabganj Rajshahi Bangladesh Lower Ganges Road network: Table below (Table 7-5) shows the top seven blocks/districts that have significant lengths of flood-affected roads for a 2-year return period flood event. These blocks require measures such as creating protection walls or increasing the road elevation or realignment of the road network. A separate study is required to find the most feasible mitigation measures to reduce losses due to floods. Final Report: Volume I Confidential Page 253 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 7 - 5 : D i s t r i c t s / s u b - d i s t r i c t s h a vi n g m a x i m u m r o a d l o s s e s Sr. No. Block District State Country 1 Parbatta Khagaria Bihar India 2 Patepur Vaishali Bihar India 3 Manihari Katihar Bihar India 4 Colgong Bhagalpur Bihar India 5 Gopalpur Bhagalpur Bihar India 6 Ghoswari Patna Bihar India 7 Raghopur Vaishali Bihar India Rail network: Table below (Table 7-6) shows the top seven blocks/districts that have significant lengths of flood-affected railway lines for a 2-year return period flood event. These blocks require measures such as creating protection walls or increasing the railway line elevation or realignment of the rail network. A separate study is required to find the most feasible mitigation measures to reduce losses due to floods. T a b l e 7 - 6 : D i s t r i c t s / s u b - d i s t r i c t s h a vi n g m a x i m u m r a i l w a y l o s s e s Sr. No. Block District State Country 1 Vidyapati Nagar Samastipur Bihar India 2 Kaliachak - III Maldah West Bengal India 3 Mohiuddinagar Samastipur Bihar India 4 Sabour Bhagalpur Bihar India 5 Barauni Begusarai Bihar India 6 Kurhani Muzaffarpur Bihar India 7 Nathnagar Bhagalpur Bihar India Agriculture:  Rice and Maize Lower Ganges basin requires suitable interventions to reduce losses to rice and maize crops as this sub basin shows maximum crop losses for a 2-year return period event.  Wheat Ghagra sub basin needs suitable interventions to reduce losses to wheat crop due to future flood events. A separate study is required to finalize the best interventions from the list below: 1. Improvements to existing drainage systems 2. Construction of flood defense systems to protect crops 3. Changes in cropping patterns towards more water resistant crops. Concluding Remarks To summarize, the study indicates that Bihar and Uttar Pradesh states in India and areas of Bangladesh lying in the basin is highly vulnerable to the floods. At sub-basin level, Lower Ganges is the most severely affected sub-basin. Final Report: Volume I Confidential Page 254 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia The findings of this study can be useful in the identifying the priority sub-basins and/or states which need immediate attention for flood mitigation measures. The study can also be treated as a base study to identify the sub-basins for flood forecasting by the World Bank. These findings can help minimize the economic losses due to floods in the three countries. It can also help to reduce the impact of floods on the population. For further improvements, it is recommended to replicate the study for finer resolution at sub-basin level. Also, in order to estimate the probable damage to other vulnerable exposures, other remaining assets (bridge, pipelines, livestock, electric lines, other crops, etc.) at risk and loss due to business interruption should also be included in the study. Local consultants and experts from various fields should be hired in order to validate the various processes involved in the study. The social vulnerability of various ethnic groups and adaptive capacity of people in the flood prone areas of the three countries should also be surveyed and made part of a separate study. The study can be replicated for other basins and sub-basin of the region. Improvements in the data availability, employing process oriented hydro-agrological models, considering detailed aspects of socio economic vulnerability and adaptive capability of various community groups in the study area can further enhance the usefulness of such studies. The sector specific impact of hydro meteorological disasters on the economy of the countries can also be analyzed. Final Report: Volume I Confidential Page 255 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Appendixes Final Report: Volume I Confidential Page 256 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia 8 Appendix A: Hydrological Station Network of Ganges Basin F i g u r e 8 - 1 : H yd r o l o g i c a l S t a t i o n n e t w o r k o f G a n g e s B a s i n ( S o u r c e : I n d i a - W R I S ) Final Report: Volume I Confidential Page 257 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia T a b l e 8 - 1 : L i s t o f H yd r o l o g i c a l India Discharge Gauges India Discharge Gauges S t a t i o n l yi n g i n G a n g e s B a s i n Sr. No. Gauge Name Sr. No. Gauge Name 32 Sangod (GDQ) 68 Rishikesh (GDSQ & FF) (Source: India-WRIS) 33 Barod (GDSQ) 69 Haridwar (FF) 34 A.B. Road X-ing (GDSQ) 70 Garhamukteshwar (GDSQ) India Discharge Gauges 35 Khatoli (GDSQ) 71 Narora Barrage (U/S)(FF) Sr. No. Gauge Name 36 Pali (GD) 72 Kachlabridge (GDSQ) 1 Tuini (P)(GD) 37 Manderial (GD) 73 Fatehgarh (GDSQ) 2 Tuini(T) (GDQ) 38 Dholpur (GDSQ) 74 Moradabad (GDQ & FF) 3 Naugaon (GD) 39 Udi (GDSQ) 75 Bareilly (GDSQ & FF) 4 Bausan (GD) 40 Bhind (GD) 76 Dabri (GDSQ) 5 Haripur (GD) 41 PachauIi (GDQ) 77 Kannauj (FF) 6 Yashwant Nagar (GDQ) 42 Seondha (GDSQ) 78 Ankinghat (GDSQ & FF) 7 Paonta (GDQ) 43 Auraiya (GDSQ & FF) 79 Kanpur (GDSQ & FF) 8 Kalanaur (GDQ) 44 Kalpi (GD & FF) 80 Bhitaura (GDSQ) 9 Karnal (GD) 45 Lalpur (GD) 81 Lucknow (GDSQ & FF) 10 Mawi (GDSQ & FF) 46 Hamirpur (GDQ & FF) 82 Dalmau (FF) 11 Baghpat (GD) 47 Basoda (GD) 83 Shahjadpur (GDSQ) 12 Delhi Rly Bridge (GDSQ & FF) 48 Mohana (GD & FF) 84 Phaphamau (FF) 13 Galeta (GDQ) 49 Shahijina (GDSQ & FF) 85 Chhatnag (GDSQ & FF) 14 Mohana (GDSQ) 50 Chillaghat (FF) 86 Garrauli (GDSQ) 15 Mathura (FF) 51 Garhakota (GD) 87 Mejja Road (GDQ) 16 Agra (GDSQ & FF) 52 Gaisabad (GD) 88 Mirzapur (GDSQ & FF) 17 Arnota (GD) 53 Banda (GDSQ & FF) 89 Varanasi (GDSQ & FF) 18 Etawah (GDSQ & FF) 54 Kora (GDQ) 90 Neemsar (GDQ) 19 Chittorgarh (GD) 55 Rajapur (GD) 91 Sultanpur (GDQ) 20 Bigod (GD) 56 Pratappur (GDSQ) 92 Jaunpur (G & FF) 21 Tonk (GDSQ) 57 Naini (FF) 93 Raibareli (GDQ & FF) 22 Baranwada (GDSQ) 58 Uttarkashi (GDSQ) 94 Palla (GDQ) 23 Dhareri (GD) 59 Deoprayag (GD) 95 Maighat (GDSQ) 24 Tal (GDSQ) 60 Badrinath (GD) 96 Ghazipur (FF) 25 Ujjain (GD) 61 Joshimath (GD) 97 Buxar (GDSQ & FF) 26 Mahidpur (GDSQ) 62 Karanprayag (GD) 98 Ballia (FF) 27 Gandhi Sagar (FF) 63 Rudraprayag (GDSQ) 99 Masani (GD) 28 Mandawara (GD) 64 Rudraprayag (GD) 100 Dadri (GD) 29 Sarangpur (GD) 65 Srinagar (FF) 101 Dhansa (GD & FF) 30 Salavad (GD) 66 Deoprayag (GDSQ) 102 Ghat (GDQ) 31 Aklera (GDSQ) 67 Marora (D/S) (GD) 103 Tawaghat (GD) Final Report: Volume I Confidential Page 258 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia India Discharge Gauges India Discharge Gauges India Discharge Gauges Sr. No. Gauge Name Sr. No. Gauge Name Sr. No. Gauge Name 104 Jauljibi (GD) 140 Nandadih (GD) 176 Narayanpur (FF) 105 Zero Point (GDSQ) 141 Gaya (GDQ) 177 Bazarsaw(GD) 106 Nahargarh (GD) 142 Patna (Dighaghat) (FF) 178 Jamtara (GDSQ) 107 Tumri (GD) 143 Gokul Barrage (GQ) 179 Gheropara (FF) 108 Paliakalan (GDSQ) 144 Lakhisarai (GDQ) 180 Nutanhat (GDSQ) 109 Elginbridge (GDSQ & FF) 145 Munger (FF) 181 Katwa (Purbast hali) (GDSQ) 110 Ayodhya (GDSQ & FF) 146 Lalbegia ghat (GD & FF) 182 Berhampore (GDSQ) 111 Basti (GDQ) 147 Sikanderpur (GDSQ & FF) 183 Islampur (GD) 112 Bhinga (GD) 148 Samastipur (FF) 184 Palashi para (GD) 113 Balrampur (GDSQ & FF) 149 Benibad (GD & FF) 185 Chapra (GDSQ) 114 Kakarahi(GD) 150 Saulighat (GD) 186 Kalna (Ebb) (GDSQ) 115 Bansi (G & FF) 151 Kamtaul (FF) 187 Hanskhali (GDQ) 116 Regauli (GDSQ) 152 Ekmighat (GDSQ & FF) 188 H/R Farraka (GDSQ) 117 Birdghat (GDSQ & FF) 153 Hayaghat (GDSQ & FF) 189 Barkisuriya (GD) 118 Turtipar (GDSQ & FF) 154 Rosera (FF) 190 Maithon Dam (GD & IF) 119 Darauli (FF) 155 Khagaria (FF) 191 Ramgarh (GDQ) 120 Gangpur Siswan (FF) 156 Bhagalpur (FF) 192 Tenughat Dam (GD & IF) 121 Chhapra (FF) 157 Colgong/ Kahalgaon (FF) 193 Konar Dam (GD) 122 Maner (FF) 158 Azmabad (GDSQ) 194 Panchet Dam (GD & IF) 123 Kuldah Bridge (GDSQ) 159 Jainagar (GDSQ) 195 Durgapur Barrage (GD & IF) 124 Rewaghat (FF) 160 Jhanjharpur (GDSQ & FF) 196 Jamalpur (GDS) 125 Chopan (GDSQ) 161 Basua (FF) 197 Harinkhola (GDS & FF) 126 Duddhi (GDSQ) 162 Baltara (GDSQ & FF) 198 Simulia (GD) 127 Japla (GDSQ) 163 Kursela (FF) 199 Tusuma (GD) 128 Pupunki (GD) 164 Sahibganj (FF) 200 Rangagora (GD) 129 Inderpuri (FF) 165 Siliguri (GDSQ) 201 Kharidwar (GD) 130 Koelwar (GDSQ & FF) 166 Matigara (GDSQ) 202 Phulberia (GD) 131 Gandhi ghat (GDSQ) 167 Sonapur (GDSQ) 203 Kangsabati Dam (GD & IF) 132 Tribeni (GDSQ) 168 Dhengra ghat (GD & FF) 204 Mohanpur (GD & FF) 133 Khadda (FF) 169 Jhawa (FF) 205 Dheng Bridge (GDSQ) 134 Chatia (FF) 170 Barhait (GD) 206 Labha (GDQ) 135 Dumariaghat(GDS) 171 Farakka (GDSQ & IF) 207 English Bazar (GDQ) 136 Lalganj (GDSQ) 172 Maharo (GDQ) 208 Kufri SHO 137 Hazipur (FF) 173 Massanjore Dam (GD & IF) 209 Jubbal SHO-II 138 Hathidah (GDSQ & FF) 174 Tantloi (GD) 210 Jubbal SHO-I 139 Sripalpur (GDQ & FF) 175 Tilpara Barrage (GD & IF) 211 Hanuman Chetty SHO Final Report: Volume I Confidential Page 259 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia India Discharge Gauges Nepal Discharge Gauges Nepal Discharge Gauges Sr. No. Gauge Name Sr. No. Gauge Name Sr. No. Gauge Name 212 Harsil SHO 28 Bagasotigaon 64 Gaurighat 213 Auli SHO 29 Jalkundi 65 Budhanilkantha 214 Tajewala Weir (Hathnikund) (FF) 30 Dhakeri 66 Tika Bhairab 215 Madla (GD) 31 Masurikhet 67 Chovar 216 Kalna (Flow) (GDSQ) 32 Kusum 68 Khokana 33 Kalimati 69 Darkot-Markhu (Source: DHM, Nepal) 34 Butwal 70 Lamichaur Nepal Discharge Gauges 35 Mangalghat 71 Kulekhani Sr. No. Gauge Name 36 Nayapul 72 Bhorleni 1 Harsingbagar 37 Setibeni 73 Padharadovan 2 Nayalbadi 38 Andhimuhan 74 Karmaiya 3 Panjkonaya 39 Borlangpul 75 Uwagaun 4 Patan 40 Ansing 76 Tumlingtar 5 Lalighat 41 Kotagaun 77 Pipaltar 6 Nagma 42 Lahachowk 78 Turkighat 7 Diware 43 Phoolbari 79 Simle 8 Asaraghat 44 Damauli 80 Barbise 9 Benighat 45 Shisaghat 81 Jalbire 10 Chhanna 46 Khudibazar 82 Helambhu 11 Mattada 47 Bhakundebesi 83 Pachuwarghat 12 Chitra 48 Bimalnagar 84 Panauti 13 Bhasme 49 Goplingghat 85 Busti 14 Gautada 50 Gharmbesi 86 Rasnalu 15 Gopaghat 51 Arughat 87 Khurkot 16 Bangga 52 Brtrawati 88 Sangutar 17 Rimna 53 Betrawati 89 Beni 18 Samaijighat 54 Belkot 90 Salme 19 Jamu 55 Kalikhola 91 Rabuwabazar 20 Chisapani 56 Devghat 92 Kampughat 21 Pahalmanpur 57 Rajaiya 93 Hampchuwar 22 Daradhunga 58 Manahari 94 Majhitar 23 Chepang 59 Lothar 95 Mulghat 24 Bargadha 60 Chitrasari 96 Chatara 25 Nayagaon 61 Sundarijal 97 Rajdwali 26 Chernata 62 Sundarijal 98 Sajbote 27 Kalimatighat 63 Shyamdado 99 Mainachuli Final Report: Volume I Confidential Page 260 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia References Ahmed, F., Rafii, M.Y., Ismail, M.R., Juraimi, A.S., Rahim, H.A., Asfaliza, R. and Latif, M. A. (2013). Waterlogging Tolerance of Crops: Breeding, Mechanism of Tolerance, Molecular Approaches, and Future Prospects, BioMed Research International, vol. 2013, Article ID 963525, 10 pages, doi:10.1155/2013/963525. BDA (2008), The Orissa Gazette, Bhubaneswar Development Authority (Planning and Building Standards), Regulation 2008 Bihar 2004 Flood Event - Relief Web (://www.reliefweb.int/library/documents/2004/undp- india-13aug.pdf) Bihar 2007 Flood Event Disaster Management Cell, Government of Bihar (http://disastermgmt.bih.nic.in/PPT%20Slides/Disaster-2005%20Slides.ppt) Boyer, J. S. (1982). Plant productivity and environment (crop genetic improvement). Science, vol. 218, no. 4571, pp. 443–448. Chow, V.T., 1959,Open-channel hydraulics, McGraw- Hill Book Co., 680 p. Chow, V.T., Maidment, D.R., and Mays, L.W. (1988). Applied Hydrology. Tata McGraw-Hill Education (reprint edition). 572 pages. Collaku, A. and Harrison, S. A. (2002). Losses in wheat due to waterlogging. Crop Science, vol. 42, no. 2, pp. 444–450. Content-to-Structure Value Ratios (CSVRs) in Support of the Jefferson and Orleans Flood Control Feasibility Studiesâ€?, Final Report, prepared by Gulf Engineers & Consultants for the USACE New Orleans District. Disaster Management Department, Government of Bihar http://disastermgmt.bih.nic.in/ Disaster Management Department, Government of Uttar Pradesh http://rahat.up.nic.in/rahat.htm FEMA (2009). Hazards U.S. Flood Technical Manual MR4, pages 8-1 to 8-6. GEC (1996). “Depth –Damage Relationships for Structures, Contents, and Vehicles and Hazarika, Manzul (2006). Education, Research, Training and Capacity Building Activities in AIT. A Case Study from Nepal, Geoinformatics Center, Asian Institute of Technology, pages 47-50. Herath, S. (2003). Flood Damage Estimation of an Urban Catchment Using Remote Sensing and GIS. International Training Program on Total Risk Management. Pp. 55. http://www.adrc.asia/publications/TDRM2003June/10.pdf Hosking, J. R. M. (1990). L-moments: analysis and estimation of distributions using linear combinations of order statistics. Journal of the Royal Statistical Society, Series B, 52, 105- 124. http://apy.dacnet.nic.in/crop_fryr_toyr.aspx http://archives.deccanchronicle.com/130816/news-current-affairs/gallery/monsoon-attack- india http://cbs.gov.np/?p=2191 http://directives.sc.egov.usda.gov/OpenNonWebContent.aspx?content=17757.wba http://eands.dacnet.nic.in/Publication/pub-2013/Index.xls http://eands.dacnet.nic.in/publications.htm http://indiarailinfo.com/news/post/construction-of-a-new-broad-gauge-line-between-pirpainti- jasidih-mohanpur-indian-railways-news/167878 http://india-wris.nrsc.gov.in/wrpinfo/index.php?title=Ganga http://pib.nic.in/newsite/erelease.aspx?relid=103623 http://srtm.csi.cgiar.org/SELECTION/inputCoord.asp Final Report: Volume I Confidential Page 261 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia http://unhabitat.org/?wpdmact=process&did=MzM5LmhvdGxpbms http://www.bbs.gov.bd/DistrictWisePopulation.aspx?MenuKey=147&distName=t http://www.bbs.gov.bd/Home.aspx http://www.censusindia.gov.in/DigitalLibrary/Tables.aspx http://www.core.indianrailways.gov.in/view_section.jsp?lang=0&id=0,294,302,536 http://www.dartmouth.edu/~floods/AMSR-E%20Gaging%20Reaches/203.htm http://www.dartmouth.edu/~floods/hydrography/E80N30.html http://www.diva-gis.org/gdata http://www.fncci.org/text/ind_fact.pdf http://www.hec.usace.army.mil/software/hec-hms/ http://www.hec.usace.army.mil/software/hec-ras/ http://www.indexmundi.com/commodities/?commodity=wheat&months=12¤cy=npr http://www.ireeindia.org/RE%20Booklet.pdf http://www.itbhuglobal.org/chronicle/archives/2008/09/helping_victims.php http://www.masshousingcompetition.org/sites/default/files/bd-dha-536summary_report.pdf http://www.moadwto.gov.np/noticefile/Year_%20Book_2013%20_Whole_Book_139686844 4.pdf http://www.nepalpropertymarket.com/index.php?linkID=178 http://www.nhai.org/index.asp http://www.nr.indianrailways.gov.in/view_detail.jsp?lang=0&dcd=3265&id=0,4,268 http://www.numbeo.com/food-prices/country_result.jsp?country=Nepal http://www.projectstoday.com/News/CCEA-approves-new-broad-gauge-lines-in-India http://www.sv.uio.no/esop/english/research/news-and-events/events/guest-lectures- seminars/esop-seminar/dokumenter/soderbom.pdf http://www.tehelka.com/theuttarakhandfloods (Press Trust of India). https://lta.cr.usgs.gov/LULC Huke, R. E. and Huke, E. H. (1997). Rice Area by Type of Culture South Southeast, and East Asia a Revised and Updated Database, International Rice Research Institute, Los Banos, Philippines. Jha, Abhas K., Robin Bloch, Jessica Lamond, 2012, Cities and Flooding - A Guide to Integrated Urban Flood Risk Management for the 21st Century, The World Bank, 1818 H Street NW, Washington DC 20433. Kirpich Z.P. 1940 (Kirpich, 1940). The Time of Concentration in Small Agricultural Watersheds, Civil Engineering 10 (6) 362. Martín-Vide, J. P., D. Nierola, A. Bateman, A. Navarro, and E. Velasco, 1999: Runoff and sediment transport in a torrential ephemeral stream of the Mediterranean coast. J. Hydrol., 225, 118–129. Mohanty, H. K. and Khush, G. S. (1985). Diallel analysis of submergence tolerance in rice, Oryza sativa L. Theoretical and Applied Genetics, vol. 70, no. 5, pp. 467–473. MSME (2012), Brief Industrial Profile of Khordha District, MSME-Development Institute, Ministry of MSME, Govt. of India, www.msmedicuttack.gov.in Musgrave, M. E. and Ding, N. (1998). Evaluating wheat cultivars for waterlogging tolerance,â€? Crop Science, vol. 38, no. 1, pp. 90–97. Nash, J. E. and J. V. Sutcliffe (1970), River flow forecasting through conceptual models part I — A discussion of principles, Journal of Hydrology, 10 (3), 282–290. NDMA (2013). Compilation of Catalogue of Building Typologies in India, Seismic Final Report: Volume I Confidential Page 262 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia Vulnerability Assessment of Building Types in India, Seismic Vulnerability Assessment Project Group, http://www.ndma.gov.in/images/disaster/earthquake/Catalogue%20of%20Building%20Type s%20in%20India.pdf NDMA (2013). Technical Document (Tech-Doc) on Typology of Buildings in India, Seismic Vulnerability Assessment of Building Types in India, Seismic Vulnerability Assessment Project Group, http://www.ndma.gov.in/images/disaster/earthquake/Building%20Typology%20Report.pdf NWS, 2002. Advanced hydrologic prediction services – Concept of services and operations. Report U.S. Department of Commerce – NOAA – NWS. Rathore, T. R. and Warsi, M. Z. K. (1998). Production of maize under excess soil moisture (waterlogging) conditions in Proceedings of the 2nd Asian Regional Maize Workshop PACARD, Laos Banos, Philippines. UNISDR (United Nations International Strategy for Disaster Reduction). 2004.“Guidelines for Reducing Flood Losses.â€? Geneva: UNISDR. http://www.unisdr.org/files/5353_PR200402WWD.pdf. USACE, 2010, Hydrologic Engineering Centre – River Analysis System (HEC –RAS), Application Guide Version 4.1, January 2010. USACE. 2000. Hydrologic Engineering Center. Hydrologic Modeling System, HEC-HMS: Technical Reference Manual. USACE Hydrologic Engineering Center. Davis, CA, USA USACE. 2005. Hydrologic Modeling System: User‟s Manual Version 3.0.0, USACE Institute for Water Resources. USACE. 2009. Hydrologic Modeling System, HEC-HMS: User’s Manual. USACE Hydrologic Engineering Center. Davis, CA, USA. Uttar Pradesh 1973 Flood Event - Irrigation Department Uttar Pradesh (http://irrigation.up.nic.in/past_flood.html). Veissman, W. and Lewis, G.L. (2003). Introduction to Hydrology. 5th Ed. Prentice Hall: Upper Saddle River, NJ. 612 p. Bangladesh Crop Wholesale price (www.mofood.gov.bd) Final Report: Volume I Confidential Page 263 of 264 The World Bank Flood Risk Assessment for the Ganges Basin in South Asia ---End of Final Report: Volume I--- Final Report: Volume I Confidential Page 264 of 264