CORFU PROJECT EXPERIENCE FROM THE RAPIDLY DEVELOPING CITY OF DHAKA, BANGLADESH

Transcription

1 CORFU PROJECT EXPERIENCE FROM THE RAPIDLY DEVELOPING CITY OF DHAKA, BANGLADESH S M Mahbubur Rahman 1,*, David M Khan 1, Abu Saleh Khan 1, M Monowar Hossain 1 1 Institute of Water Modelling, Bangladesh; * House # 496, Road # 32, New DOS, Mohakhali, Dhaka 1206, Bangladesh; Tel: +88 (02) ; smr@iwmbd.org ABSTRACT Dhaka is one of the nine cities considered as case study areas in the CORFU project. The objective of CORFU is to identify flood resilience measures which are suitable for varying conditions, in case of CORFU cities in Asia and Europe. There are four major Work Packages (WP) in CORFU. WP1 consists economic growth modelling, urban growth modelling. District wise panel data was collected from various sources and was analyzed to conduct econometric modelling and finally projection was made up to Urban Growth Modelling requires extensive land use information. The eastern side of the city is underdeveloped and UGM was very useful to predict the development for this area. In WP2 several models were developed for varying condition and 1D-2D coupled model was developed for the first time for Dhaka City for better understanding of inundation process. During model development it was decided 3 different kinds of models will be utilized for Dhaka and data requirement was identified to improve accuracy of the model. A damage tool has been developed as part of CORFU to assess impact in WP3 which requires damage curves for different sectors. Damage curves are being developed for Dhaka City as a part of the research. Survey was also conducted in Dhaka City to assess flood risk perception. The survey results provided insight into flood risk perception in two very different locations. In WP4 it was identified the flood management framework for Dhaka City is at the basic level. Majority of measures considered are structural in nature. Flood Resilience Measures of CORFU were assessed for one of the busiest and important areas of Dhaka City named Shegunbagicha. The evaluation provided insight in to developing a framework which can be applied in different kind of urban systems. KEYWORDS Flood resilience; urban flooding; flooding in Dhaka; flood modelling; flood impact 1. INTRODUCTION During the inception of the CORFU project cities were selected carefully from Asia and Europe so that the selected cities can address the complexity and differentiation between the overall flood management processes of two varying continents. Dhaka was chosen as it is a rapidly developing mega city and the capital Bangladesh, a country which lies in the deltaic plain of Ganges Brahmaputra and Meghna basin. On economic scale it is a developing country growing fast and considering climatic condition it is one of the most vulnerable countries in the world due to climate change. This makes Dhaka city a very interesting candidate for research in the CORFU project. This paper describes the different activities conducted as a part of the Work Packages in CORFU and shares the experience gained from Dhaka with respect to these work packages. CORFU aims to conduct collaborative research and tries to identify how strategies in one country can be utilized in another country. This follows the important items identified in CORFU for flood management and the appropriateness of any adaptation devised to fit the condition of Dhaka city. 2. DESCRIPTION OF DHAKA CITY 2.1 General description of city Dhaka being the capital of Bangladesh has seen unprecedented growth in the last couple of decades. This is partly due to the overall economic growth that the country has experienced and partly due to very centralized growth which saw Dhaka accounting for almost 30% GDP of Bangladesh. At city level it also sees similar focus where the City is growing vertically. Government has taken initiatives to

2 control this vertical growth which creates enormous pressure on existing infrastructure. Dhaka has about 28% people in slum areas fuelled by an estimated 300,000 to 400,000 migrants annually (CBSG, 2010). Greater Dhaka covers nearly 400 km 2 and has a population of 12.5 million. The Greater Dhaka is bounded by the Balu River on the east, the Tongi Khal on the north, the Turag- Buriganga Rivers on the west and the Dhaka-Demra-Chittagong Road cum embankment on the south. This area includes both flood protected western Dhaka and unprotected eastern Dhaka. Ground elevation of the Dhaka city varies from 0.5m to 12 mpwd. About 60% to 70% of the city area is within the elevation of 0.5 to 5 mpwd (IWM, 2006). The eastern part of the Dhaka city includes low lands, abandoned channels and depressions. 2.2 Flooding history There are two kinds of floods in Dhaka city: river flooding and water logging. River flooding is caused by high water levels, which overtop river banks and also the embankments. The country experienced disastrous floods in 1954, 1955, 1968, 1971, 1974, 1987, 1988, 1998 & 2004, which also affected Dhaka City. Water logging is mainly caused by heavy rainfall and drainage congestion. Sometimes these two types of flooding combine, e.g. high river stages at drain outfalls cause backwater effects in the drainage system leading to water logging. During every monsoon season, large portions of Dhaka city are inundated due to its low topography. In 2004, about 124 sq. km area, approximately 176,000 families and around 900, 000 people were affected by the 2004 floods (Hossain et al 2006). 3. DRIVERS OF FLOODING IN DHAKA The natural drainage system of a locality is affected when urbanization takes place. Urbanization is driven by population growth and the economic activity of the region. Thus within CORFU the characteristics of economic growth is captured through economic growth modelling. The affect of economic growth and also other factors drive the urban growth, which is rationalized and related through urban growth modelling 3.1 Economic growth modelling Data was collected from different sources in order conduct the economic growth modelling fro Dhaka City, as Economic Growth Modelling requires panel data. Lot of the data required does not exist at Dhaka City level, but rather on district or national level. So district level data was used rather than city level. Bangladesh has 64 districts and data were collected for these districts. The boundary of these districts also changed from time to time, so historical records had to be adjusted to accommodate the boundary change. The data collected from various sources were checked and analyzed for consistency, which are given below. GDP Total and by sector - GDP by sector and by district (shown in Figure 1) Total Employment - by education level, by sectors, in knowledge Intensive industries and services Population - by education level, by age, by gender Other demographics - fertility by age of mother, infant mortality Household number and distribution by size Consumer price index Industry wise gross district product Data were collected from the following sources: Bangladesh population census 1991, 2001 Population and Housing Census Preliminary results, 2011 Statistical Year Book, 1997, 1999, 2002, 2005, 2009 Labor Force Survey 1995, , , 2005, Report on monitoring of employment survey, 2009 Gender Statistics of Bangladesh, 2008 World Bank

3 Figure1: Historic GDP collected from records Using the above data, team from HWWI projected future populations for Dhaka district with three different growth rates. Using the EGM economic development for different sectors were also projected. Figure 2 shows some projections from EGM. Figure 2: Future scenarios ( ) for GDP (left) and total population (right) for Dhaka district. 3.2 Urban growth modelling Land use data collection and preparation Dhaka City has developed from a small town of about 1 sqkm with 0.2 million people during the pre Moghul period (1600s). The city now has a population of 12.5 million and is likely to grow into one of the 10 mega cities in the world with a population of over 20 million by The City Development Authority (RAJUK) in , completed a Detail Area Plan (DAP) for Dhaka city, which is intended to capture the policies and recommendations of the Structure Plan and Urban Area Plan of DMDP. DAP provides aggregated land-use for different parts of the City. Within CORFU both aggregated land-use or object based land use could be utilized. Object-oriented data records individual properties and buildings, whereas aggregated data, by its nature, records aggregated information of more or less homogeneous areas. No central agency is responsible for keeping record of object oriented data. So data were collated from various sources. These data were then updated so that they can be used together. Figure 3 shows the land use data collected and prepared for Dhaka case Study. The Object oriented data for Dhaka Case Study contained actual building footprints and associated land-use information. The associated land use information were: Building use 1 st tier Structure type of building Building use 2 nd tier No of stories Building use 3 rd tier Construction year

4 (a) (b) (c) (d) Figure 3: DAP land-use for 2004 (a) and 2010 (b) and Cadestral land-use for 2004 (c) and 2010 (d) The land-use in Dhaka City is changing very fast. So it is important to know the time period which the data represents. The prepared data represents Figure 4 shows changing landscape of Dhaka City in a small period of time. In order to account for future changes in land-use urban growth modeling will be performed as a part of the research. The intent of the data collection was to find out whether sufficient data is available to perform urban growth modeling for City like Dhaka with reasonable confidence. Urban growth modelling provides a mean to assess the spatial characteristics of urban growth behaviour for a specific metropolitan area. This behaviour is driven by economic and demographic development while guided by local characteristics. Within the CORFU-modelling framework, the aim of UGM is to express future spatial consequences of socioeconomic scenarios, flood impacts and responses for the given urban case study locations, in this case Dhaka City.

5 Figure 4: Urban growth in Dhaka City The data used in Dhaka case study was able to validate the various scenarios in the future. The results from urban growth modelling was compared with proposed plans and based on the comparison decisions were taken considering merits and demerits. It was also possible to develop scenarios based on these comparisons which took into account flooding for different storm events. 3.3 Climate change A1FI and B1 are the two scenarios that were considered for Dhaka case study for climate change assessment. Specifically A1FI scenario is important, as its gives more adverse impact of climate change. The scenario has very high economic growth, primary energy use is very high (primarily fossil fuel sources, low change in land use). The projected climate change variables are: precipitation change with regards to return period sea level rise flow from U/S catchments Table 1 shows the climate change factors for Dhaka City for A1FI and B1 scenario (University of Tokyo, 2008). Table 1: climate change factors for Dhaka City Dhaka ΔTlocal (A1FI) [K] ΔTlocal (B1) [K] ΔPmean/ Ppresent mean (A1FI) [%] (June-July-August) ΔPmean/ Ppresent mean (B1) [%] (June-July-August) ΔPextreme/ P present extreme (A1FI) [%] ΔPextreme/ P present extreme (B1) [%] Sea Level Rise (SLR) is not a major climate change contributor for Dhaka. The city is 200 km away from the coastal area. During dry season some tidal influence is observed, but the influence is insignificant during monsoon due to huge flow from the Ganges Brahmaputra Meghna (GBM) basin which pushes the sea waterfront back.

6 4. FLOOD INUNDATION MODELLING International Conference on Flood Resilience: The study area for Dhaka City covers part of densely populated central area and part of the eastern area which is rapidly developing. The drainage characteristics of these two locations are also different. The central part of protected from river flooding and has pipe network and box culverts to drain storm water. The eastern is affected by river flooding and the drainage system consists of open channels. The flood inundation modelling for these two areas were done using different kind of models. 4.1 Modelling approach The eastern Dhaka drainage model is developed based on IWM s existing North Central Region Hydrodynamic (NCRHD) Model to simulate the monsoon season flows and water levels of the river. The NCRHD is developed using Mike 11. The calibrated NCRHD model was truncated and the khals (natural channels) in the Eastern Dhaka were incorporated into the model. Surface runoff flows generated from rainfall was estimated for each catchment using a rainfall-runoff model. This flow will then used as input to the hydrodynamic model to simulate the flow of water through drainage channels and surrounding rivers. Mike Urban model was selected for the hydraulic model of Central Dhaka as it can simulate the various components drainage system consisting of pipe network. There are four components of the hydraulic model, which are: Drainage network consisting pipes, box culverts and open channels Lakes which work as retention ponds Three sluice gates/regulators Two pump stations Figure 5: The components of Hydraulic model and their properties

7 The runoff generated from rainfall flows through the pipes and box culverts and eventually drains to lakes and canals by several outlets. Run-off from some of the catchments flows to several retention ponds which are intricately connected. Hatirjheel is the main retention pond in the system. The excess water from Hatirjheel flows through a regulator. If the water level at rises above certain height the excess water is pumped out through a number of pumps of different capacities. The control structures and the operational setup for the hydraulic model are shown in Figure 5. Water level data extracted from the Mike 11 model of Eastern Dhaka were used as boundary condition at the outfalls of the Central Dhaka model. There considerable amount of sediment observed in the Dhaka drainage system, so 20 percent of the total depth of box culvert was provided as sediment depth. The percentage of sediment is determined by field survey. After preparing the hydraulic model, the model was coupled with the DEM (Digital Elevation Model) with cell size of 25m. Mike Flood was used to perform the modelling exercise for both the models. The manhole elevations were checked with DEM to ensure proper connectivity between 1D and 2D model. In the 1D model Mike Urban uses features like basin to represent the retention ponds. In the 2D model the retention ponds were replaced by the pond bathymetry and surrounding connectivity were established accordingly. Similar exercise was also conducted in the eastern Dhaka Model where the channel cross sections were checked with the surrounding DEM and the lateral links were established accordingly. 4.2 Flood modelling Hydraulic modelling runs and results To study the effect of urban flood three different scenarios were simulated. These are: The actual event of 9 th September to 17 th September of A 1 in 10 year rainfall event. A 1 in 30 year rainfall event Eastern Dhaka is unprotected and influenced by down stream conditions. The model run included average water level at the down stream and design storms with three return periods. As the area is influenced by river flooding the inundation extent and depth does not vary too much the different return periods. The Central Dhaka area is protected by embankment, so the river flooding does not have any influence inside. The outfalls have sluice gets which isolates from down stream flooding and excessive run-off is pumped outside during storm events. Flooding in this area is dictated by rainfall and internal drainage condition and the model was setup accordingly. Similar to eastern side three flooding events were simulated for central Dhaka with different design storms. It can be seen from flood depth and extent shown in Figure 6 that overall extent of flooding is quite similar with some differences in flooding depth. It was found from the model results that there were small differences in the extent of flooding among these events and the differences were more significant considering the duration of flooding.

8 5. IMPACT ASSESSMENT 5.1 Direct Damage Assessment Figure 6: Flood maps for 10-yr and 30-yr return period Damage is the negative result of the spatial and temporal impact of an event on societal elements (people, buildings, etc.), societal processes (interruption of production, services, etc.) and the environment (Vetere Arellano et al., 2003). To understand the damage or losses that floods can cause, the CORFU damage model comprises three damage model components: Direct tangible damage Indirect tangible damage Intangible damage In order to assess the direct damage due to flooding a damage tool has been developed as part of CORFU. The damage tool was tested for Dhaka case study. The tool requires land-use and flooding extent data in raster format. Different gird size to create the raster data and their affect on using the damage tool was evaluated. Land-use information greatly affects the damage assessment process and cell size affects land use aggregating the land-use or sometimes even negating the land-use classification. Same can be said for overland flow modeling and damage assessment. Overland flow process can be affected by cell size, which can lead to overestimate or underestimate flooding depth and thus affecting the damage estimation. Computation time is also big issue for the models. For Dhaka Case study it s taking 3 hrs for 25 m grid runs. If the grid size is reduced it takes more time, not to mentions additional time for preparation of data. To assess the compatibility of the damage tool developed under CORFU, the land use and flood maps for Dhaka Case study were used in conjunction with the DDC s provided by University of Exeter. The results showed the Damage Tool can be useful means to determine building damage quantitatively. The difference in building damage for different return periods flooding are shown in the following figures. The damage scale is based on hypothetical number of 1 to 6, where 1 indicates less damage and 6 indicates higher damage. It can be seen how building damage increases with flooding with higher return period. The expected annual damage calculated based the 3 return periods is also shown in the Figure 7.

9 Figure 7: output from damage tool There is no standard DDCs established for Bangladesh. In a recent study for Khulna City (ADB, 2010) depth damage curves were developed for 5 land use categories. To fully utilize the capability of the damage tool, damage curve will be developed for Dhaka City. Depth Damage Curves (DDC) for different land-use categories will be developed based on household survey and focus group discussions. Survey questionnaire was developed to get information from flood prone areas of the City for past flooding events. The classification of different land-use and household types were determined based on spatial analysis. This in terms will help using the DDC incorporated in the damage tool and calculate damage for objects for different land-use category. 5.2 Indirect and intangible damage assessment As a part of CORFU indirect damage related to flooding will be also explored. Indirect damage requires sector specific damage data which is lacking for Dhaka City. Some major studies were done during 2004 and 1998 flood event. These studies are currently getting reviewed for indirect damage assessment for Dhaka City. In order to identify the health impact of flooding CORFU will explore historical records of diarrhoea incidence in various parts of Dhaka City. The spatial and temporal record of diarrhoea incidence will be analyzed the historic flooding records in order identify any correlation. Diarrhoea disease data were collected from ICDDR,B and demographic and other related data were collected from 2011 census data. Dhaka was also one of the case study areas which were considered for the social perception study and survey. The survey recorded 22 hours of discussion through in-depth semi-structured interviews in Dhaka with residents (urban poor communities) in Badda, which is an area experience annual river flooding. The interview also covered middle class residents and their experience within the protected area and also NGOs involved in flood education and relief activities in flood prone areas. The preliminary observation was that perception of risk due to flooding is completely different in Dhaka City compared to perception of residents in Whilmesberg, Germany. This provides valuable input to the ARL framework (Awareness-Relationships-Livelihood) considered for social resilience study within CORFU project.

10 6. DEVELOPMENT OF FLOOD MANAGEMENT STRATEGIES International Conference on Flood Resilience: After the catastrophic 1988 floods, the Government and international development partners began a five-year plan called the Flood Action Plan (FAP). FAP 8A (Master Plan for Greater Dhaka Protection Project) study aimed at developing a Master Plan on comprehensive flood control and storm water drainage in Dhaka Metropolitan area (JICA, 1991). The study area comprised of approximately 850 km 2 of Greater Dhaka and the proposed improvement included embankment, raised roads, regulators, retention ponds and pumps stations. The ADB funded a parallel study named Dhaka City Integrated Flood Protection Project (FAP 8B) (Lois Berger, 1991). In addition to flood protection from peripheral rivers, FAP 8B also provided guidance to improve internal drainage of Dhaka City. This included infrastructure improvements such as: cleaning and rehabilitation of major existing open drainage khals. The rehabilitation works consist of dredging and re-profiling to restore the khals to their design capacity, including lining where needed; rehabilitation and construction of piped drainage facilities to relieve drainage congestion in the built-up areas of Dhaka; cleaning and unblocking existing covered and open drains in the more densely populated parts of the City, together with a limited programme of construction of small capacity new drains; Flood management for Dhaka relies on several agencies. The primary responsible agency is Dhaka WASA. DWASA mandate clearly indicates the organization is engaged in improvement and operation of structural measures to reduce flooding within the City. Dhaka City Corporation (DCC) is the other organization responsible for maintaining City drainage. Where as DWASA is responsible for the major drains which consists of pipes and culverts, DCC is responsible to build the surface drains collecting storm water from the streets and building compounds. In addition BWDB is another agency which also plays a role in the flood protection of the City by operating and maintaining the embankment and associated structures on the western side of the City. As no formal policy exits for coordination between the agencies mentioned above, it creates bureaucratic obstacle each time an inter-agency service is required. For this reason coordination among the public agencies is currently inefficient. At the same time enforcement of existing policies and legislations related to urban flooding in Dhaka are in a dire state. Flood Forecasting and Warning Center (FFWC) works as the central organization to provide early warning and information on floods. It is the main non-structural flood management initiative for the City and provides early warning for the peripheral river flooding. This is especially useful for people in eastern part of Dhaka City which experiences river flooding. Within CORFU in order to do evaluate different flood risk management frameworks hazard, exposure and vulnerability were chosen to be the main elements for analysis of flood risk. For Dhaka Case Study these elements were assessed with 2004 flood event. It was evaluated which measures were in place at that time, the performance of the measures and also the impact due to flooding, the recovery and the main problems identified from the flooding experience. Based on the findings the flood management framework for Dhaka City was ranked as basic, which means the City only has the basic means to address flooding. CORFU flood management framework is build upon different spatial scale of the City. The measures to improve flood resilience were evaluated in different scales for Shegunbagicha district of Dhaka City. Five dimensions were considered for the resilience measures. These included natural, physical, economic, institutional and social dimensions. One of the initial observations was to know the appropriate weightings and points that can be assigned for each of these dimensions. It was found that several of the components might not be relevant for Dhaka for present condition but may be important as the level of resilience improves in the future. Figure 8 was used to identify building scale structures to analyze the FRM.

11 Figure 8: Building scale information used to determine flood resilience measures for Dhaka City 7. CONCLUSION In its objective the CORFU project states collaboration of research on flood management between Asia and Europe. The methodology in different work packages were developed based on experiences from Case study Cities and Dhaka had significant contribution in the development of the methodologies. Panel data is required for economic growth modelling, so district level national census data were successfully used to generate projections. Urban growth modelling is very useful for a developing city like Dhaka. It was possible to develop several scenarios for the City based on various objective criteria. Several models were integrated to develop the flood model for Dhaka City. During model development several challenge were overcome and the benefit of developing these state of the art models were established. For impact assessment a damage tool was developed, which was tested in Dhaka City. The tool requires depth damage curves (DDC) for which a DDC for Dhaka is in the process of development. Impact assessment also covers indirect and intangible damage assessment. The findings of flood risk perception survey showed the difference in perception between people in Europe and Asia. The flood resilience measures being considered in CORFU were evaluated for Dhaka City. Different dimensions of flood resilience index were evaluated and modifications were made to reflect actual condition. The experience from Dhaka City has made the methodologies considered in different work packages of CORFU more robust.

12 8. ACKNOWLEDGEMENTS International Conference on Flood Resilience: Research on the CORFU (Collaborative research on flood resilience in urban areas) project was funded by the European Commission through Framework Programme 7, Grant Number Support for the members of the CORFU team in different aspects of the Dhaka Case Study is highly appreciated. 9. REFERENCES Asian Development Bank (ADB), 2010, Strengthening the Resilience of Water Sector in Khulna to Climate Change, ADB, Manila Capacity Building Service Group CBSG (2010) Ensuring Services to Slum Dwellers - Dhaka WASA Organisation for Low Income and Slum Community Water Service Delivery, Water and Sanitation for Urban Poor (WSUP), UK. Hossain A.N.H.A., Azam K.A. and Serajuddin M. (2006) The Impact of Floods on Dhaka City s Water Supply, Sanitation and Drainage and Mitigation Options, in Options for Flood Risk and Damage Reduction in Bangladesh, Siddiqui K.U. and Hossain A.N.H.A. (eds.) The University Press Limited and Institute of Water Modeling, Dhaka. Institute of Water Modelling (IWM) (2006) Study on Drainage Master Plan for Dhaka City and DND Area, Final Report, Dhaka JICA (1991) Master Plan for Greater Dhaka Protection Project, FAP 8A Main and Supporting Report, Government of Bangladesh Louis Berger International, Inc. (1991) Dhaka Integrated Flood Protection Project, FAP- 8B, Final Report, A.D.B. T. A. NO BAN, Government of Bangladesh Rajdhani Unnayan Kartripakkha (RAJUK), Capital Development Authority, (2010) Detailed Area Plan (DAP) for Dhaka Metropolitan DevelopmentPlan (DMDP), Final Report, Dhaka University of Tokyo, Integrated Research System for Sustainability Science (2008), The Study on Climate Impact Adaptation and Mitigation in Asian Coastal Mega Cities, Final Report to JICA, Tokyo. Vetere Arellano, A. L., Nordvik, J.-P. & Ranguelov, B. Year. In search of a common methodology on damage estimation: from an European perspective. In In search of a common methodology on damage estimation - workshop proceedings, Veen, A. V. D., Arellano, A. L. V. & Nordvik, J.-P., eds., 2003 Delft.