69868 Preparing to Manage Natural Hazards and Climate Change Risks in Dakar, Senegal A Spatial and Institutional Approach Pilot Study Report June 30, 2009 Hyoung Gun Wang Economist, Spatial team FEU, SDN Marisela Montoliu-Munoz Head , Spatial team FEU, SDN The Geoville Group Ndèye Fatou D. Gueye Consultant, FUE, SDN TABLE OF CONTENTS ACKNOWLEDGEMENTS ................................................................................................ I ACRONYMS AND ABBREVIATIONS............................................................................II EXECUTIVE SUMMARY ................................................................................................. 1 1. Introduction ............................................................................................................. 9 2. Natural Disasters in Dakar, Senegal ..................................................................... 13 3. Conceptual framework for hazard, risk, and vulnerability ................................. 21 3.1 Definitions ................................................................................................... 21 3.2 Risk Factors and Related Information Sources.............................................. 24 3.3 Hazard types ................................................................................................ 25 3.4 Hazards and Climate Change........................................................................ 27 4. Spatial Analysis...................................................................................................... 29 4.1 Overview and the Definition of Peri-Urban Areas ........................................ 29 4.2 Map Generation............................................................................................ 32 4.2.1 Land Use Mapping ........................................................................... 36 4.2.2 Hazard Potential Maps...................................................................... 41 4.3 Spatial Analysis............................................................................................ 48 4.3.1 Methodology .................................................................................... 48 4.3.1.1 Overview ............................................................................ 48 4.3.1.2 Generation of Input Data.................................................... 51 4.3.1.3 Mapping and statistical analysis......................................... 55 4.3.2 Spatial Analysis Results.................................................................... 56 4.3.2.1 Land Cover Changes .......................................................... 56 4.3.2.2 Social Exposure and Vulnerability...................................... 59 4.3.2.3 Economic Exposure and Vulnerability ................................ 60 4.3.2.4 Built-up Areas exposed to Hazards ..................................... 62 4.3.2.5 Non-built-up Areas exposed to Hazards.............................. 66 5. Hotspot Characterization and Assessment of Institutional Capacity .................. 69 5.1 Overview...................................................................................................... 69 5.2 The “Primer� Survey (the City Typology and Risk Characterization Matrix) 69 5.3 The survey results ........................................................................................ 71 5.3.1 General data on the Departments....................................................... 71 5.3.2 Governance Structure Related to Disaster Risk Management ............ 72 5.3.3 Urban Planning and Land Use Regulations ....................................... 77 5.3.4 Exposure of Political and Economic Assets to Disasters.................... 79 5.3.5 Climate Change Preparedness ........................................................... 80 5.3.6 Disaster Response System................................................................. 80 5.3.7 Peri-urban Areas ............................................................................... 81 5.4 Knowledge and Capacity Gaps for disaster management and climate change impacts......................................................................................................... 82 6. Moving Forward: Lessons Learned and Action Plans to Ramp up Natural Hazard and Climate Change Risk Management Practices in Dakar, Senegal .... 85 6.1 Guiding principles: Hyogo Framework for Action........................................ 85 6.2 More pro-active approach to informing, motivating, and involving people in their own local communities......................................................................... 86 6.2.1 Informing, motivating, and involving people in their own local communities. .................................................................................... 87 6.2.2 Strengthening local institutional capacity and coordination. .............. 87 6.2.3 Policy reforms and investments for improved hazard resilience and preparedness at the local level........................................................... 88 6.3 Replication of the Pilot Study ....................................................................... 89 BIBLIOGRAPHY ............................................................................................................. 91 TABLE Table 1: Components and Factors of RISK and Related Information Sources and Assessment Results ................................................................................................................ 24 Table 2: Types of Natural Hazards, Potential Anthropogenic Causes and Relations to Other Hazards ............................................................................................................... 26 Table 3: Overview of the Produced Maps ............................................................................ 33 Table 4: Available Multi-temporal Satellite Scenes ............................................................. 36 Table 5: List of Thematic Classes for land use change mapping........................................... 38 Table 6: Key for classifying and reporting relative natural hazard potentials at aggregated resolution (250 meter) ......................................................................................... 41 Table 7: Classes Contained in the Multi-Hazard Map .......................................................... 45 Table 8: List of Maps Produced for Spatial Analysis ........................................................... 49 Table 9: Source Data, Preparation Methods, and Resulting Input Data for the Spatial Analysis ............................................................................................................................ 51 Table 10: Land Cover 1988 and 2008 and Land Cover Changes .......................................... 57 Table 11: Land Price Value Exposed to High Hazard Potentials (million US$).................... 61 Table 12: City Typology and Risk Characterization Matrix ................................................. 70 Table 13: Population in Authorized and Non-authorized Human Settlements in the Dakar Metropolitan Area ............................................................................................... 72 FIGURE Figure 1: The Classification of Urban, Peri-Urban, and Rural areas in the Dakar Metropolitan Area ...................................................................................................................... 2 Figure 2: Hazard Mapping of the Dakar Metropolitan Area ................................................... 3 Figure 3: Hotspots of Social Exposure with High Population Growth and High Hazard Potentials VS Population Growth (1988 – 2008) in Areas with Hazard Potentials.. 4 Figure 4: Flooding and Coastal Erosion Threaten Life and Resources of Dakar ................... 14 Figure 5: Risk Triangle, Crichton 1999................................................................................ 22 Figure 6:: Risk as a Function of Hazard, Vulnerability, Exposure, and Resilience................ 23 Figure 7: The Dakar Metropolitan Area, Covering the Departments of Dakar, Guédiawaye, Pikine, and Rufisque ........................................................................................... 30 Figure 8: The Classification of Urban, Peri-Urban, and Rural areas in the Dakar Metropolitan Area .................................................................................................................... 32 Figure 9: Detailed Land Use Map of Dakar, 2008................................................................ 37 Figure 10: Comparison of Detailed and Aggregated Land Use Map for Year 2008 .............. 38 Figure 11: Land Use Changes, 1988, 1999, and 2008 .......................................................... 40 Figure 12: Relative Flood Potential Layer for Dakar, Senegal Aggregated to 250m Cells. ... 42 Figure 13: Coastal Erosion Potential Layer for Dakar, Senegal Aggregated to 250m Cells... 43 Figure 14: 1m Coastal Inundation Potential for Dakar, Senegal Aggregated to 250m Cells.. 44 Figure 15: 5m Coastal Inundation Potential for Dakar, Senegal Aggregated to 250m Cells.. 44 Figure 16: Multi-Hazard Potential in Dakar, Senegal, Including Relative Flood Potential, Coastal Erosion and 1m Coastal Inundation Potential Scenarios .......................... 46 Figure 17: Multi-Hazard Potential in Separate Maps............................................................ 47 Figure 18: Synchronizing Land-Cover/Use Data with the Census Data................................ 53 Figure 19: Comparison of Census-Based (Department) and Spatially Disaggregated Population Distribution ....................................................................................... 54 Figure 20: Hazard Potential Zoning Applied to the Satistical Analysis Maps ....................... 55 Figure 21: Detail of Hazard map.......................................................................................... 56 Figure 22: Development of the Built-up Area, 1988, 1999, and 2008................................... 58 Figure 23: Total Share of Built-up Areas within Urban, Peri-urban, and Rural Communes in 2008 .................................................................................................................... 58 Figure 24: Population Growth 1988 - 2008 in Areas with Different Hazard Potentials ......... 59 Figure 25: Hot Spots of Social Exposure Given by High Population Growth between 1999 and 2008 and High Hazards Potentials................................................................. 60 Figure 26: Economic Exposure to Hazards Expressed by Land Price Values (US$ per sq km) ............................................................................................................................ 61 Figure 27: Built-up Area in 2008 Threatened by Hazards, with Detail Shown in the Upper Part ..................................................................................................................... 63 Figure 28: Share of Industrial/Commercial/Traffic Areas and Exposure to Hazards in 2008 64 Figure 29: Non-built-up Areas 2008 Exposed to High Hazard Potentials ............................. 66 Figure 30: Hazard Potentials in Non-built-up Areas............................................................. 67 BOX Box 1: Natural Hazard and Disaster Risk Overview for Dakar............................................. 14 Box 2: Natural Hazards in the Dakar Metropolitan Area...................................................... 17 Box 3: Justification for Selection of Specific Natural Hazards ............................................. 29 ACKNOWLEDGEMENTS Many people deserve acknowledgement for their valuable contributions to this study. First of all, we thank Madhu Raghunath, who set the foundations for this project during her assignment at the Spatial and Local Development Team, and Christian Diou and Sylvie Debomy (AFTU2), who have provided useful guidance and support from the perspective of the Africa Region. Mr. Diou offerd invaluable advice and facilitation of the validation and dissemination workshops in Dakar, from his position at the World Bank Senegal Country Office. We are also grateful for the frank and helpful feedback of the peer reviewers, Franck Bousquet (AFTU2), Uwe Deichmann (DECRG), Daniel Hoornweg (FEU), and Madhu Raghunath (from her new position in MNSSD), whose comments have helped us strengthen this pilot study and consider ways to enhance future replications of this work. We would like to express our sincere appreciation for the support and ground-truthing comments received from the workshop participants in Dakar, Senegal, on March 5th, 2009, including Alioune Badara Diouck, Mamadou Dieng, Abdou Birahim Diop, Mme Madeleine Diof Sarr, Mme Badiane Reine Marie Coly, Demba Ba, Adama Guèye, Aboubacry Sadikh Niang, Oumar Cissé, and Denis Jean-Jacques Jordy. A particular note of thanks is due to Gracia Sorensen (FEU) and Astou Diaw-Ba (AFCF1) who provided tireless and effective logistic assistance. Last but not least, we are grateful to the Global Facility for Disaster Response and Recovery (GFDRR) for providing generous funding for this project. Saroj Kumar Jha and C.Y. Ollero offered resources and encouragement, and Andrea Zanon and Segio Dell’Anna played an important facilitation role in the implementation of this study. The findings, interpretations, and conclusions expressed in this report can, nevertheless, only be attributed to the authors, and do not necessarily represent the views of the World Bank, it Executive Directors, or the countries they represent. i ACRONYMS AND ABBREVIATIONS ADM The Municipal Development Agency AOF Afrique Occidentale Française (French West Africa) CAPC The Auxiliary Committee of Civil Protection CEP Coastal Erosion Potential CIP Coastal Inundation Potential CRED Center of Research on the Epidemiology of Disasters CRPC The Regional Committee of Civil Protection CSE Centre of Ecological Monitoring CSPC The High Commission of Civil Protection DASSE Direction of Social, Health and Educative Affairs (Direction des Affaires Sociales Sanitaires et Educatives) DEM Digital Elevation Model DLR Deutsches Zentrum für Luft- und Raumfahrt (German Aerospace Center) DPC The Division of Civil Protection DRM Disaster Risk Management DRR Disaster Risk Reduction DST Technical Services FEU, SDN The Finance, Economics and Urban Development Department, The Sustainable Development Network GFDRR The Global Facility for Disaster Reduction and Recovery GIS Geographic Information Systems GNSP Fire Department staff (Groupement National des Sapeurs Pompiers) IAGU African Urban Management Institute IPCC The Intergovernmental Panel on Climate Change MSL Mean Sea Level Rise NGO Non-Governmental Organization ORSEC Organization des Secours (the National Emergency Organization Plan) PDAS Master Plan for Urban Planning and Preservation of the Niayes and the Green Zones of Senegal PDU Urban Master Plan PDUD Urban Mobility Plan for the Agglomeration of Dakar PRDI The Regional Integrated Development Plan PRSP Poverty Reduction Strategy Paper ii RFP Relative Flood Potential SAR African Oil Refinery Company SPOT Satellite Pour l’Observation de la Terre SRAT Regional Land-Use Planning SRTM The Shuttle Radar Topography Mission UN/ISDR The United Nations International Strategy for Disaster Reduction UNDP The United Nations Development Programme iii EXECUTIVE SUMMARY Introduction 1. This report describes a pilot study of natural risk hazards in the peri-urban extension areas of the Dakar Metropolitan Area, Senegal.1 The area subject of this study stretches across 580 square kilometres, covering less than 1% of the national territory, but housing about 50% of Senegal’s urban population. Much of the rapid population growth of the Dakar Metropolitan Area is taking place beyond the boundaries of the Department of Dakar (the city center), in peri-urban areas that combine two disquieting features: they present significant vulnerability to some natural hazards, and they have unclear administrative and governance arrangements, often being out of the direct oversight of established urban and rural local governments. Situations like this are not unusual in developing countries, and call for more systematic attention to hazard risk management in peri-urban areas, including a better understanding and awareness of the nature of the hazards that they face as well as of the institutional capacities and measures that would be necessary to manage them better. 2. The objective of this pilot study is, therefore, two-pronged. First, the study intends to propose a new methodology for quick assessment of natural hazard risks at a metropolitan- region scale, using new tools of spatial analysis based on geographic information systems (GIS) data. Second, the study aims to apply the principles and diagnostic questionnaire of the Climate Change City Primer developed by the East Asia Region of the World Bank to get a comprehensive view of the institutional framework for climate change-related hazard risk management existing in the city at this time. Bringing the spatial and the institutional analyses together, the study proposes and starts to develop a number of dissemination and awareness-raising tools that can help to inform different stakeholders about the general parameters of the natural hazard risks facing the Dakar Metropolitan Area. The pilot study concludes with a broad action plan for Dakar, to ramp up disaster management practices, as motivation for a stakeholder debate to define subsequently a set of specific and viable actions. Natural Hazards in the Dakar Metropolitan Area 3. The pilot study focuses sharply on three types of natural disaster hazards affecting the Dakar Metropolitan Area and its surrounding regions: flooding, coastal erosion, and sea level rise. Flooding is one of the most severe hazards threatening Senegal, and in recent decades it has become a frequent and enduring reality. Senegal is also one of the African countries that suffer most from coastal erosion. In addition, the phenomenon of sea level rise – although less imminent and visible today – is attracting increasing attention as part of the ongoing discussion of the impact of climate change in coastal areas. These three types of natural hazard risks coincide in the Dakar Metropolitan Area, located in a low-lying, peninsula-like territory with a long coastal line. The area concentrates significant social, economic and political assets that are vulnerable to these hazards. These three types of hazards were chosen as focus of this pilot study because of their relevance and, especially in the case of flooding and coastal erosion, their current prevalence and impact in the area under study. The 1 The Dakar metropolitan area is also called the region of Dakar (la r gion de Dakar). 1 extension of the study to other types of natural hazards – e.g. land subsidence, also present in the area – was limited by the time and resources available for this pilot study. Peri-Urban Expansion Areas 4. The definition of peri-urban areas varies across countries and can sometimes be a source of confusion. This study started by defining the broad geographic scope of the analysis, which covers an area of 580 square kilometers including and around the city center of Dakar, and stretching the Departments of Dakar, Guediawaye, Pikine, and Rufisque (see map below). Using a combination of quantitative measurements (distance to city center, population density) and qualitative observations and know-how from local experts, the study classified the communes in this broad area into three segments: Urban (Dakar center, and communes with high density of urban economic and industrial activity), Rural (communes conventionally classified as “rural� according to Senegal administrative system), and Peri- Urban (communes lying in-between, exhibiting mixed land uses and relatively lower densities than the urban communes.). Figure 1: The Classification of Urban, Peri-Urban, and Rural areas in the Dakar Metropolitan Area Spatial analysis 2 5. The spatial analysis combines results from hazard mapping (see maps below) with population maps, land price data, and land cover information in order to derive the exposure of different variables in different locations to the three selected natural hazards. The scale of the spatial analysis is regional/metropolitan, a level of detail that is relevant for the awareness-raising and institutional engagement purposes intended here. Extensions to this analysis, not covered at the current pilot stage, may include (a) consideration of other natural hazards; (b) more thorough evaluation of the potential economic impact of natural hazards, taking into account direct and indirect damages; and (c) additional analysis of more detailed geographic information systems (GIS) data, cadastres, and first-hand ground-truthing, to ascertain detailed risks faced by specific populations and built areas in the broad areas indicated here as under risk. Figure 2: Hazard Mapping of the Dakar Metropolitan Area 6. Spatial analysis helps to discern the spatial relationship between the natural hazard potentials and the exposed variables, i.e. population, land use units, and economic assets. Specifically, it generates statistical results and maps containing the following information: • Potential hotspots and risk areas illustrating type and degree of vulnerability for the different risk categories and scales of risk defined previously, • Built-up areas exposed to risk: general characterization of the urban land use (i.e. residential, industrial/commercial, etc.) including type of vulnerability, and • Non-built-up areas exposed to risk: general characterization of the land use, topography, and soil, including type of vulnerability. 3 7. Spatial analysis shows that land cover in the area under study has changed significantly in the past twenty years – the surface of urbanized areas has increased by over 25% in the period, about 1% per year. Population growth over the period between 1988 and 2008 took place to a significant extent in areas that are prone to a moderate or high hazard potential. In particular, peri-urban areas have the highest percentage of population growth in hazard prone areas. In communes that are defined as peri-urban, almost 40% of new population has settled in areas with significant hazard potential from inland flooding, coastal erosion, or sea level rise. This rate is twice as high that of urban (19%) and rural communes (23%) in the area under study. Figure 3: Hotspots of Social Exposure with High Population Growth and High Hazard Potentials VS Population Growth (1988 – 2008) in Areas with Hazard Potentials Hotspots of Social Exposure with High Population Growth and High Hazard Population Growth (1988 – 2008) Potentials in Areas with Hazard Potentials 100,0% 90,0% 80,0% 70,0% 60,0% P n 50,0% P 40,0% h 30,0% 100,0% 20,0% Hot-spot of population 90,0% increase in high risk zones 10,0% 80,0% 0,0% Urban Peri-urban Rural Dakar (average) 70,0% Population growth in areas with low or 60,0% no hazard potential Population growth in areas with 50,0% high/moderate hazard potential 40,0% 8. A broad calculation of the exposure and vulnerability of economic assets in the area under study was inferred from the spatial analysis of land price values. Using this method, the study estimated that the Dakar Metropolitan Area represents a total land value of $44 billion. Out of this total value, over $2 billion or 5% is exposed to high natural hazard potentials. Given the imperfect functioning of land markets in Dakar, this is only a very rough approximation to the exposure of economic assets to natural hazards. Further extensions of this work should refine the methodology for the calculation of economic risks due to natural hazards. Nevertheless, these early estimates indicate the importance of appropriate planning to mitigate the city’s economic vulnerability to these risks. Hotspot characterization and assessment of institutional capacity 9. A survey developed by the World Bank’s East Asia Region (the City Typology and Risk Characterization Matrix, or simply the “Primer� Questionnaire) was adjusted to the needs of this analysis and conducted to collect information on the Dakar Metropolitan Area’s human and built environment characteristics, potential impacts of natural hazards and climate change risks, and local-level disaster management institutional capacity. The survey analysis reviewed (i) general information on four administrative Departments of the Dakar Metropolitan Area: Dakar, Guédiawaye, Pikine, and Rufisque, (ii) governance structure related to disaster risk management, (iii) urban planning and land use regulations, and (iv) 4 other factors such as political and economic impacts of disasters, climate change preparedness, disaster response system, and peri-urban areas. 10. The survey found that the implementation framework for disaster risk management (DRM) is ambiguous and complex at the local level, even though Senegal has been actively pursuing Disaster Risk Reduction (DRR) strategies at the national and regional levels. For example, in the case of flooding hazard, the local mayor is often responsible for disaster response but has very little influence on mitigation as those policies are often determined at the national level. Local governments lack adequate funding for infrastructure investments and service delivery, and face the challenge of skills shortage and lack of technical capacity to deal with complex issues like climate variability risks. Furthermore, planning instruments like land use planning, which has the ability to influence the urban-rural footprint, remains under the influence of the national government. Using Lessons Learned to Assist Awareness-Raising 11. Guiding principles: Hyogo Framework for Action. The Hyogo Framework for Action is adopted as guiding principle for designing the recommendations from the current study. The Hyogo Framework is a global blueprint for disaster risk management implementation during the decade 2005-2015, adopted by the Member States of the United Nations. It underscores: (i) the need for a more pro-active approach to informing, motivating and involving people in all aspects of disaster risk reduction in their own local communities; and (ii) the scarcity of budget resources allocated to risk reduction objectives, as compared to the significant potential to better exploit existing resources and established practices for more effective disaster risk reduction. Consistent with these principles, the recommendations emerging from this study include measures that improve awareness of disaster risks, as well as a frank assessment of institutional resources and capacity for disaster management and prevention in the Dakar Metropolitan Area (see summary table below). 12. Informing, motivating, and involving people in their own local communities. The most immediate recommendation from this pilot study is to develop a general awareness campaign – which has already started with the process of validation and dissemination of the findings of the study. Local agencies and local communities should play a pivotal role in disaster management practices, and develop demand for improved land use planning and disaster response. In turn, empowered local communities and agencies can play a key role in identifying areas and structures under risk, and monitoring and evaluating the implementation of the measures. Alignment with the Hyogo Framework requires a consultative and participatory process that ensures sustainability and ownership of the measures proposed. 13. Strengthening local institutional capacity and coordination. The study also reveals failures of capacity, accountability and coordination among local agencies and between them and agencies at other administrative levels. These are issues that require further examination and discussion among the relevant stakeholders. This study limits itself to pointing to some of the key institutional capacity and coordination issues identified. It suggests some that may be on the critical path for progress and deserve special attention, such as the identification and empowerment of an institutional champion for disaster risk management and prevention at the metropolitan level, and the development of a local database on hazards and the training of local agency staff to use it effectively. 14. Policy reforms and investments for improved hazard resilience and preparedness at the local level. Beyond the immediate findings and recommendations of this study, stakeholders in Dakar need to consider what substantive policy and investment measures may 5 be considered over time. Ultimately, policy reforms that influence behavior to promote better risk management and investments that strengthen resilience at the local level would be needed to improve the situation of Dakar and its peri-urban expansion areas. The findings of this study suggest the importance to focus on better local land use planning and management, and infrastructure, and the last segment in the table below summarizes possible measures to be considered. This study does not make specific recommendations in these areas, however, as this would require a more detailed analysis. Selection of viable choices would also depend importantly on stakeholder consultation. 6 Avenue for Action plan engagement General Awareness • Organize local knowledge and information dissemination Campaign activities, targeting local public agencies and local (Already initiated in communities, on the seriousness of the natural hazards and the context of this pilot climate change impacts on their own lives, with focus on the study) behaviors that the population can control and improve on. • Arrange collaboration and joint activities with various local agencies, academic and research institutions, non-profit organizations (NGOs) to pursue these campaigns of sensitization. Strengthening Local • Identify viable and well-recognized institutional champion at Institutional Capacity the Metropolitan Level. and Inter-agency • Initiate discussion at the highest political level for institutional Coordination strengthening and coordination and reforms: • Initial focus: (a) development of early warning and quick- response system, paying attention to currently under-served peri-urban areas; and (b) improve local organization and capacity to enforce urban zoning and regulations to reduce vulnerability to natural hazards, with special focus on currently under-served and fast-growing peri-urban areas. • Medium-term focus: (a) adequate resourcing of key agencies; (b) policy reform; and (c) reallocation of public expenditure and investment to local disaster risk mitigation and prevention. (See further below.) • Develop a spatial database for local disaster management in the Dakar Metropolitan Area, and ensure broad access and hands-on training for local agency staff. • Promote local communities’ engagement and participation in disaster prevention measures. Policy Reform and • Improve local land use planning and management: Investment (a) improvement of land property right assignment and enforcement, with special focus on peri-urban areas; and (b) consultative development of metropolitan development plan, including identification of disaster hotspots and corridors for urban expansion, and potential land acquisition plans to support urban growth corridors. • Strengthening resource base for local authorities, including through proposed betterment taxes that take advantage of improved land management plans and corridor development. • Invest in climate- and disaster-proofing infrastructure and housing stock: retrofitting existing infrastructure and housing in hazard-prone areas; improving infrastructure planning and monitor quality of investments. 7 Replication of the Pilot Study 15. This study was intended as a pilot to test new methodologies, and identify how the approach could be enhanced in case of replication. Interest in replicating the approach exists for other African cities, as well as for a selection of Asian cities. This pilot study may also provide the foundation for the development of a city vulnerability index to be applied to a large number of cities. In replications of this study, it is proposed that the following enhancements be considered: • Consideration of a broader range of natural hazards, beyond the three included in this study. • Progress toward a more robust definition of peri-urban areas. • More detailed and better documented analysis of the economic impact of hazards. • More detailed discussion of the methodology for population density imputation, possibly considering different relationships between building density and population density, depending on whether the area is formal or informal. • Addition of information (layering) on major infrastructure (roads, electricity, sanitation). • Utilization of more detailed GIS data sources, such as those captured in cadastres, to inform the more detailed economic and population analysis suggested above. 16. These extensions of the current pilot study may lead to slight increases in the cost of undertaking the study in the initial replications, until the methodology is refined and standardized. 8 1. Introduction 17. The Dakar Metropolitan Area (or the Region of Dakar) represents less than 1% of the national territory, but shelters 25% of its total population and 50% of its urban population. Since it was the capital city of the French West Africa (AOF), Dakar has been a hub; its strategic position was instrumental in its cultural and political leadership. Moreover, its economic, cultural, and political leadership was associated with touristic and military assets derived from a favorable geographical context. 18. The Dakar Metropolitan Area has become a sprawling conurbation. This evolution was made possible by the dynamism and natural growth of its population, its role as a main exchange point, and its role as host area for migrants. With the excessive concentration of population in a small portion of space, Dakar agglomeration is a real “Human Anthill�. The rapid urbanization of the region remains uncontrolled resulting in huge problems in terms of meeting the basic needs in housing, land, mobility, life environment, and urban poverty. 19. The urban structure is marked by imbalances and lack of integration as a result of ineffective land management and lack of strategic planning at the local and regional levels. The Government’s urban investments are mostly concentrated in the city center (the Department of Dakar) at the expense of the peri-urban areas – labor supplying areas where half of the total population of the region lives. Yet this social and infrastructural disparity promotes the spillover of migration from the city center, its peri-urban areas, and its close hinterlands resulting in degradation of the life environment. 20. Dakar’s housing and land markets face a number of interrelated challenges in terms of access and affordability; it has become increasingly hard to meet the rapidly increasing demand for these goods within the city’s perimeter, in the context of price inflation and land speculation. As a consequence, the suburbanization of the region has accelerated, creating new problems – for instance, those associated with property rights and management: ownership is often unclear or uncertified in peri-urban areas. As underscored by World Development Report 2009: Reshaping Economic Geography (World Bank, 2008a), well- functioning land markets are essential to facilitate the movement of the population toward economic opportunity, as well as the evolution of land uses from lower to higher value or, more importantly, the allocation of land to the most efficient uses. 21. The current territorial and administrative organization of the Dakar Metropolitan Area is complex, with interlocking of various forms of governance (Region, Department, Arrondissement, Commune, Rural communities). At the larger, regional level, the “Region� is headed by the governor (appointed), followed by the “Department,� headed by the Prefect (appointed). At the local level, there are three types of distinct territorial or administrative units: the “Arrondissement� is headed by the Sous-prefect (appointed), the “Communes� by the Mayor (elected), and the “Rural Communities� by the President of rural community (elected). 22. This institutional complexity and ambiguity generates a fragmentation of local policies and power conflict at various levels (administrative and institutional). In addition, the weight of the national and regional governments can often undermine local authorities and local public action. With the ongoing decentralization reform, administrative authorities and decentralized public services are in fact losing weight. The relationships between local authorities and the national and regional administrations are not very well defined, mixing autonomy, assistance, and supervision. The excessive “agencization� of the Government 9 activities and key projects results in higher density and more complex government structure, with new ad-hoc agencies carrying out the missions the ministries are expected to take. 23. The Dakar Metropolitan Area is characterized by a lack of dialogue between the administration services, the administration and local authorities, and between the local authorities. The populations, the civil society, and the private sector are only weakly involved in urban management. 24. Dakar has been a hub of cultural and political leadership and the destination of large rural-urban migrations since the 1960s. Recent satellite imagery reveals contiguous expansion of urban built-up areas over the decades, with a large proportion of the urban growth occurring in the peri-urban areas (outer suburbs) of the city center. As the Dakar Metropolitan Area experiences rapid urban expansion, it faces increased vulnerability to natural hazards and climate change risks. 25. Overlaid on top of this confluence of political and demographic factors, the natural hazard risks faced by the Dakar Metropolitan Area make for a truly complex picture in terms of ensuring safe livelihoods and an enabling environment for economic productivity. In order to meet those challenges and to manage rapid urbanization in the areas likely to be hit by natural hazards or undergo climate change impacts, the knowledge and response capacity of the local authorities need to be significantly improved. 26. The World Bank’s Spatial and Local Development Team (FEU, SDN) and the World Bank-housed Global Facility for Disaster Response and Recovery (GFDRR), in collaboration with colleagues from the World Bank Senegal Country Office, have launched an innovative pilot study. The main objectives of this study are two pronged: 27. First, it develops a new generic methodology combining spatial and institutional analyses at reasonable costs, which would serve a standard assessment template ready to replicate to other cities and other countries. This will benefit global disaster management practitioners and communities. 28. Second, it provides action plans for Dakar, Senegal to ramp up disaster management practices. We aim at providing an overarching strategic framework in bottom-up and sensitization perspectives, rather than a list of extensive top-down directions and micro- management recommendations. We believe the latter is counter-productive, unsustainable, and inconsistent with our guiding principle: informing, motivating, and involving people in all aspects of disaster risk reduction in their own local communities. This element will benefit directly local communities in the Dakar Metropolitan Area as well as Senegal governments in general. 29. In this regard, this study (i) develops state-of-the-art spatial analysis tools to spatially evaluate natural hazard and climate change risks, and (ii) addresses the critical knowledge and capacity gaps of local governments in dealing with rapid peri-urban expansion into areas that may face vulnerability to natural hazards, including those risks associated with climate variability. Three pillars of activities are being implemented sequentially. Activity 1: Spatial Analysis of Natural Hazard and Climate Change Risks and Hotspot Characterization in Peri-Urban Expansion Areas of Dakar, Senegal 30. Activity 1 develops a methodology for the generation of hazard and vulnerability maps for the Dakar Metropolitan Area. Three types of hazards are selected for detailed spatial analysis: Flooding (inland flooding of depressions), Coastal Erosion, and Sea Level Rise. The occurrence of flooding within the area of Dakar has recently been increasing for climatic and 10 anthropogenic reasons, and constitutes a major threat especially for newer settlements in unsuited low-lying terrain. Coastal erosion as a more steadily acting, but very hazardous process in the area is included in the detailed analysis along with scenarios of sea level rise. 31. Satellite imagery at different time points (1988, 1999, and 2008) and hazards information from various sources are collected, geo-processed, and integrated for the thematic map generation. These maps include land use profiles, geology, hydrology, single and multi natural hazard maps (flooding, coastal erosion and sea level rise scenarios). Spatial analysis combines all these information, and produces hazard risk profiles in the Dakar Metropolitan Area. The concepts and methods follow best-known standards and scientific approaches. Activity 2: Hotspot Characterization and Assessment of the Institutional Capacities at the Local Level 32. Activity 2 implements a local field survey for the hotspot characterization of the four departments in the Dakar Metropolitan Area: Dakar, Pikine, Guédiawaye and Rufisque, and identifies the inter-relationship between governance structure, urban-rural characteristics, disaster risks and climate change preparedness. The survey also examines the institutional capacity at the local level. The activity combines survey results and assesses gaps in the prevention and mitigation capacity of local governments with respect to the particular risks that they face as their cities expand beyond their current jurisdictions. Activity 3: Action Plans to Ramp up Natural Hazard and Climate Change Risk Management Practices in Dakar, Senegal 33. Activity 3 proposes action plans of Dakar to ramp up natural hazard and climate change risk management practices. The guidelines and action plans draw on the findings of Activities 1 and 2 and the international consensuses on best-practice exercises. 34. The guiding principle is “more pro-active approach to informing, motivating, and involving people in their own local communities,� as articulated in the Hyogo Framework for Action 2005-2015. Local agencies and local communities should play a pivotal role in disaster management practices, and correspondingly empowered to ramp up disaster management practices. The empowerment should follow four venues of engagements. • First, local agencies and local communities should be better informed. Construction of a spatial database for local disaster management and sensitization is highly recommended. • Second, local agencies and local communities should be provided more resources and administrative support (from national and regional governments) to implement and enforce disaster-mitigating land-use regulations effectively. • Third, disaster management in peri-urban expansion areas, which are often politically and economically neglected, should be addressed in long-term inclusive local development perspectives. • Fourth, Sensitization activities to wake up public awareness on natural hazards and climate change risks are a key trigger for more sustainable and inclusive disaster management. 11 2. Natural Disasters in Dakar, Senegal 35. Flooding is one of the most severe hazards threatening Dakar, and in the last years it has become a frequent and enduring reality. The underlying causes are complex and involve not only the recent increase of rainfalls, but in particular the whole socio-economic process of an out-of-control urban sprawl. The consequences are devastating: “Three months since 183,000 people in Dakar were affected by severe floods resulting from torrential rains, many houses and schools are still floating in water.�2 36. Moreover, Senegal is one of the countries that suffer most from coastal erosion: “Currents sweep away the sand from Rufisque’s coast and deposit it further south. The beach is slowly being hollowed out and the shore is receding. Abdoulaye Ndiayhe, an old fisherman who offers his services as guide, insists that more than 100 meters once separated the sea from a house whose ruins are now being lapped by the waves.�3 37. Around Dakar, the maritime domain is crowed with hotels -- already operating or in construction – buildings touching the sea which may result in rock slides and tidal waves. There are various causes of coastal erosion, whether they are natural or entropic. But global warming generated by greenhouse gas may increase sea level. Coastal erosion can therefore be regarded as a phenomenon in progress. Surveys on the impacts of climate change in the Senegalese coastal areas (Denis et al. 1995; Niang-Diop et al. in press) show that the increase of the sea level can result in floods in the lower coastal areas and increased salinization of lands and of surface and ground waters. 38. In the Cap-Vert Peninsula, coastal erosion is felt in both beaches and rocky shores. The main identified erosion areas include: the area of Camberene-Yoff, the West and East ledges of Dakar, and the bay of Hann. It is worth mentioning the Mbeubeuss quarries which show serious signs of erosion. Coastal erosion is particularly felt in the area of Rufisque- Bargny with the narrowing of the beach of Rufisque, particularly along the center of the city, resulting, in the sectors of Merina and Thiawlene, in overhanging buildings and stripped roads.4 39. The rocky sector of the tip of the Cap-Vert Peninsula is an unstable area. Based on the observation of the Madeleines, Fall et al. (1996) proposed annual rates between 0.8 and 1.4 m for the loamy cliffs of the Hospital reaching 4m at the volcanic tuffs of the Pasteur beach. For all the West and East ledges, Diop (2000) indicates that between 1980 and 1997 the annual gradual disappearance rate of the littoral is 0.45 and 2.7m. The most affected sector being the Madeleine and the Rebeuss beach. 40. Thus, the below shown pictures are only small glimpses of the extent of the hazards that are threatening and afflicting the Dakar Metropolitan Area. The flood map shown in Figure 3 gives an impression of the dimension of the 2008 flooding. Both, flooding and coastal erosion, are to a large part caused by human behavior and are exacerbated by Climate Change. 2 http://www.planuk.org/wherewework/westafrica/senegal/floodsindakar/. 3 http://portal.unesco.org/en/ev.php- URL_ID=30499&URL_DO=DO_PRINTPAGE&URL_SECTION=201.html 4 Rapport sur l’état de l’environnement au Sénégal. Edition 2005. Centre de Suivi Ecologique. 13 Figure 4: Flooding and Coastal Erosion Threaten Life and Resources of Dakar Upper left: A desperate impression of the recent flood http://www.solarpanel.co.za/global- warming.htm Upper right: September 2008 in Dakar http://wow.gm/africa/senegal/dakar/article/200 8/9/10/senegal-flooding-spreads-as-rains- continue Lower left: Senegal: Coastal Zones at the Mercy of the Waves http://portal.unesco.org/en/ev.php- URL_ID=30499&URL_DO=DO_PRINTPAG E&URL_SECTION=201.html Box 1: Natural Hazard and Disaster Risk Overview for Dakar Overview (Source: http://ww2.unhabitat.org/programmes/uef/cities/summary/dakar.htm5 Dakar, the political and economic capital of Senegal, has a metropolitan population of over 1.8 million; thus, it concentrates 22% of the country’s population on 0.1% of its land. The metropolitan region enjoys a temperate coastal climate. It is located on a peninsula that can be divided into three zones: a) an eastern section of volcanic outflow, sands and a large plateau; b) a central region of dunes and depressions; and c) a western section of undulating hills and plateaus. 5 UN-Habitat (no date given): Identifying Geographic and Thematic Environmental Issues through Consultation. http://ww2.unhabitat.org/programmes/uef/cities/summary/dakar.htm 14 Economy (Source: http://ww2.unhabitat.org/programmes/uef/cities/summary/dakar.htm Economically, Dakar generates 68% of Senegal’s GDP and its workers account for 66% of the national wage bill. 80% of the country’s industries are located in the metropolitan region. In resource use, Dakar consumes three-quarters of Senegal’s piped water and 40% of its charcoal. The unemployment rate is 30% and 40% of the active work force is employed in the informal sector. About 13% of the population is estimated to be living below the country’s poverty line. Settlement Structures (Source: http://ww2.unhabitat.org/programmes/uef/cities/summary/dakar.htm Spatially, urban development is characterized by a preponderance of informal settlements with limited access and poor transportation, and inadequate infrastructure and services. Two of every three newly-established households in the metropolitan region are sub-standard. Unplanned settlements have made it difficult to gain access to and to connect neighborhoods, and have put additional strain on the existing transportation network. Low- income districts tend to be the most underserved; for example, in the central city 61% of households have piped water but this figure drops to 16% in the poor district of Rufisque. Hazards (Source: http://ww2.unhabitat.org/programmes/uef/cities/summary/dakar.htm Problems of environmental health - malaria, gastrointestinal diseases and upper respiratory tract infections are prevalent in Dakar’s poorer neighborhoods and are linked with poor sanitation and waste management. Water pollution - Dakar’s most important aquifer is located close to the surface, making it vulnerable to malfunctioning sanitary systems and other domestic wastewater. The risk of bacteriological contamination is greatest for those who draw water from those who draw water from the aquifer’s 5000 unregulated wells. Natural risks - the city is exposed to frequent and constant natural risks. Flooding, though infrequent, can be devastating. The last major flood affected 10,000 families; it is estimated that the next major flood will cause nearly $9 million worth of damage. Damage from coastal erosion is a more constant risk. Industrial risks - Dakar’s citizens have experienced a series of fatal industrial accidents. These are linked to poor workplace safety, proximity of industrial activities and residential areas, a low level of public information, and lack of enforcement of safety measures and other rules. Recent climatic developments and implications for Dakar (Source: http://findarticles.com/p/articles/mi_go2454/is_/ai_n29342505 ) Senegal benefits from a Sudano-Guinean climate and thus from more favorable ecological conditions in the southern part of its territory. Furthermore, the temperate climate of the Atlantic seaboard has played an important role in the establishment of most of its major cities. The resulting imbalances between North and South, East and West, both eco- geographically and with respect to economic potential, significantly influence the internal mobility of the population. Reflecting these factors, Senegal’s current population of almost 15 11 million is distributed very unevenly across the country’s total area of 196 722. Changing climatic conditions have stimulated this mobility and contributed to reinforcing the role of urbanization. Indeed, like most West African nations, particularly those of the Sahelian region, Senegal has experienced a complete upheaval in its climatic norms since the mid-1960s, when a long period (1950-67) of surplus rainfall caused the rapid growth of the overall population and an augmentation of rural population density. Jean Leborgne points to three distinct periods of rainfall deficit (1970-3, 1976-7, 1983-4) punctuating the long period of drought in the Sahelian region between 1970 and 1990. The climate in Senegal is regulated by a rainy season of 3-4 months and a dry season of 8-9 months, and the natural world proves extremely vulnerable to climatic variations (Michel 1990). Because the Senegalese economy is based principally on agriculture (peanuts, millet, sorghum, rice, cotton, manioc, sugar cane and niebe), the chronic drought since the 1970s has had a traumatic effect. The crisis of the economic motor that is agriculture inspired a massive migration toward the cities and increasingly to foreign countries. The capital, Dakar, has been the principal destination for the rural exodus and the principal site for observing the most pertinent problems caused by this migration and accelerated urbanization. The development of Dakar has manifested itself by a rapid spatial expansion, resulting from intrinsic demographic dynamism (natural increase) and migratory influx from regions of the interior (net immigration). The Dakar agglomeration, with a population now estimated at some 4 million, thus constitutes the most important site of socialization and invention of Senegalese society. As a result of the enduring internal migration, Senegal’s population was already over 48% urban in 2001, one of the highest levels of urbanization in mainland sub-Saharan Africa. Over 76% of this urban population lived in areas classified as slums, which were growing at over 4% p.a. (UN-HABITAT 2006, 188). The degradation of marine ecosystem conditions through climatic change and other factors has left Senegalese fishermen, especially the youngest age cohorts, destitute and thus drives them toward massive emigration to Spain on traditional fishing boats, as now regularly featured in international news reports. This dramatic phenomenon sees thousands of immigrants searching for a better life through suicidal voyages from the Senegalese coastline. (Highlighted text by author of this study) 16 Box 2: Natural Hazards in the Dakar Metropolitan Area Hazard Type Information, maps, and data provided by various References and Hazard sources characteristics Flooding Scientific paper on the relationship between Urban sprawl Mbow, C., Local and flooding at Yeumbeul suburb. In depth analysis of the Diop, A., occurrence in natural, human and external causes of flooding. Diaw, A.T., depressions “The most important factor is the human factors with the Niang, poor management of land and the occupation of (2008)6 Human factors are very depressions�. important Scientific paper on Problems associated with flooding in Lo, P.G., factors of Dakar, western Senegal: influence of geological setting and Diop, M.B.7 flooding town management Recent “The increase of both flooding in some districts and the increase due to environmental damage that is taking place year by year are increased closely linked to the geological setting of the area, notably rainfall the topography and geomorphologic processes. These Flooding factors are so important that decision makers must take full closely linked account of them when considering any new proposals for to the development�. geological setting of the “The last major flood affected 10,000 families; it is UN-Habitat8 area estimated that the next major flood will cause nearly $9 million worth of damage� Flood map of Western Africa from Sept. 30, 2008 ReliefWeb9 Aug.Sept. 2005 Dartmouth Flood Heaviest rains in Dakar in two decades. Worst affected Observatory were Pikine and Guediawaye shanty towns outside Dakar. 10 Worst floods in 20 years in Senegal’s suburbs. "reconstruction required after the floods would cost some 45 billion CFA francs" 6 Mbow, C., Diop, A., Diaw, A.T., Niang (2008): Urban sprawl development and flooding at Yeumbeul suburb (Dakar-Senegal). African Journal of Environmental Science and Technology Vol. 2 (4), pp. 075-088 http://www.academicjournals.org/AJEST/PDF/pdf%202008/April/Mbow%20et%20al.pdf 7 Lo, P.G., Diop, M.B. (2000): Problems associated with flooding in Dakar, western Senegal: influence of geological setting and town management. Bull. Eng. Geol. Env. 58, p 145-149. 8 UN-Habitat (no date given): Identifying Geographic and Thematic Environmental Issues through Consultation. http://ww2.unhabitat.org/programmes/uef/cities/summary/dakar.htm 9 ReliefWeb: Floods in Western Africa (as of 30 Sep 2008). http://www.reliefweb.int/rw/rwb.nsf/db900SID/LPAA-7JZBJQ?OpenDocument&rc=1&cc=sen 10 Dartmouth Flood Observatory, Global Active Archive of Large Flood Events. http://www.dartmouth.edu/%7Efloods/Archives/2008.xml 17 Hazard Type Information, maps, and data provided by various References and Hazard sources characteristics Floods reported for Dakar: 2008, 2007 (2003, 2002, 2000) GLIDE Disaster Data Base11 “Thousands have been affected by flooding in more than SENEGAL: 40 neighborhoods across Senegal, including 21 Dakar Flooding suburbs according to the Senegalese fire rescue services. spreads as AFP has reported at least one drowning�. rains continue12 “Dakar neighborhoods affected include Pikine, Guédiawaye, Thiaroye and Diamaguène, according to Mayé Konate, spokesman the National Association of Firefighters�. But Youcef Ait Chellouche, disaster management coordinator at the IFRC, warns the problems are only beginning. “We need to be ready for potential epidemics – like cholera – to break out as the water starts to subside. We are closely monitoring this situation.� “Flooding worsened a cholera epidemic already underway in Dakar in 2005, which eventually infected more than 23,000 people�. “The two districts are Guédiawaye and Pikine, both of SENEGAL: which are in coastal wetland zones. Particularly affected is Thousands the neighborhood of Wakhinane Nimzatt, which means displaced “dig and drink� in the local language Wolof�. from their “Building on wetlands exacerbates the flood situation�, Dakar said Sysall. “These wetlands are not viable for homes13 construction.� The rural exodus to Dakar has contributed greatly to the flood problem, as newcomers build houses illegally, but the government is planning a tougher stance. 11 GLIDE Disaster Data Base, http://www.glidenumber.net/glide/public/search/search.jsp 12 SENEGAL: Flooding spreads as rains continue. http://wow.gm/africa/senegal/dakar/article/2008/9/10/senegal-flooding-spreads-as-rains-continue 13 SENEGAL: Thousands displaced from their Dakar homes. http://wow.gm/africa/senegal/dakar/article/2008/10/7/senegal-thousands-displaced-from-their-dakar-homes 18 Hazard Type Information, maps, and data provided by various References and Hazard sources characteristics Coastal Erosion “Analysis of the results shows that the slides were Fall, M., Permanent influenced by the geotechnical properties of the soil, the Azzam, R., process weathering, the hydrogeological situation, and the erosion Noubactep, by waves�. C.14 Cliff retreat rates up to 2m “It was demonstrated that the tuff cliffs retreated 10 to 60 p.a. m between 1953 and 1981. This agrees with the poor geotechnical properties of the volcanic tuffs. The recession Human factors of the loam cliffs in this period is 10 to 20 m. The basalt contribute, e.g. cliffs show the lowest recession because of their high via sand strength�. extraction “The village is facing a serious coastal erosion problem; UNESCO- the outer row of fisher folk houses has already been Dakar destroyed by the sea and thus abandoned by the population. Office15 Assistance from specialists in coastal dynamics/processes is an urgent requirement�. “Like many other coastal cities in West Africa, Dakar is Sane, M., prone to several coastal hazards. The most serious among Yamagishi, these hazards is coastal erosion�. H.16 “The effect of nature is through two modes of strong and constant swells and the action of tidal waves during rough- sea seasons. The contribution of human beings is due to a cumulative effect of poor city planning and overpopulation�. “This concentration of people led to a construction boom, and then, to excessive sand extraction from beaches and dunes�. Drought Droughts are a common meteorological phenomenon in the Mbow, C., Major hazard Sahel zone, and are caused by recurring, large-scale shifts Diop, A., of the area in the general global circulation. They may be aggravated Diaw, A.T., when regarded by climate change, or climate variability may further Niang, .. at a continental increase with climate change. (2008)17 14 Fall, M., Azzam, R., Noubactep, C. (2006): A multi-method approach to study the stability of natural slopes and landslide susceptibility mapping. Engineering Geology 82 241– 263. http://cat.inist.fr/?aModele=afficheN&cpsidt=17446593 15 UNESCO-Dakar Office: Socio-cultural survey of the Yoff coastal community, Senegal: conserving biological and cultural diversity. http://www.unesco.org/csi/act/dakar/projec4e.htm 16 Sane, M., Yamagishi, H. (2004): Coastal Erosion in Dakar, Western Senegal, Journal of the Japan Society of Engineering Geology, Vol. 44; No. 6; p. 360-366. http://sciencelinks.jp/j- east/article/200407/000020040704A0134339.php (abstract) 17 Mbow, C., Diop, A., Diaw, A.T., Niang (2008): Urban sprawl development and flooding at Yeumbeul suburb (Dakar-Senegal). African Journal of Environmental Science and Technology Vol. 2 (4), pp. 075-088 http://www.academicjournals.org/AJEST/PDF/pdf%202008/April/Mbow%20et%20al.pdf 19 Hazard Type Information, maps, and data provided by various References and Hazard sources characteristics scale. Droughts decrease agricultural yields and led in the past Drought is an decades to the tremendous migrations of people to Dakar indirect driver and other urban areas. Thus, droughts lead indirectly to the for severe other problems and hazards within the Dakar Metropolitan Area hazards (e.g. itself. flooding) in the Dakar Metropolitan Area. Earthquakes The earthquake potential is low in Dakar according to the Munich Re, Nathan data base of Munich Re, though not totally absent. Nathan18 Tornado There is a medium probability of tornados according to Munich Re, Munich Re, on a four-point scale ranging from low over Nathan medium to high and very high. Hail Storm The Hailstorm probability is low (from low – medium – Munich Re, high) according to Munich Re. Nathan 41. At the environmental level, despite the efforts made, pollution is prevailing in the bays surrounding Dakar such as the Bay of Hann. The other major environmental concern remains air pollution which derives from public transportation, old cars, and cement factories. The gradual colonization and degradation of the “Niayes� and other sensitive ecological areas may turn out to be the worst ecological disaster.19 The environmental problems cannot be addressed without mentioning the “Mbeubeuss� dump, an environmental disaster-in-the- making, surrounded by human settlements and agricultural lands. 18 NATHAN Data Base by Munich Re, http://mrnathan.munichre.com/ 19 “Niayes� means lowlands between parallel sand dunes. The “Grande Niayes� is located between Dakar and Pikine. The water table is near the ground surface (ie located under the superficial sand zone). 20 3. Conceptual framework for hazard, risk, and vulnerability 3.1 Definitions 42. There is a variety of definitions of hazards, risk, vulnerability, and related terms, as listed in Thywissen (2006). The definitions and understanding of these terms reflect the attitude towards the underlying causes and factors of risk and, at the same time, influence concepts and strategies of disaster management. As Thywissen points out, there has been a shift from regarding disasters as extreme events caused by natural forces, to viewing them as manifestations of unresolved (human) development problems. Hazards 43. Among the many definitions of ‘Hazard’, the one of the United Nations International Strategy for Disaster Reduction (UN/ISDR) is frequently found in the relevant literature and is taken here as a standard definition: A potentially damaging physical event, phenomenon or human activity that may cause the loss of life or injury, property damage, social and economic disruption or environmental degradation. Hazards can include latent conditions that may represent future threats and can have different origins: natural (geological, hydrometeorological and biological) or induced by human processes (environmental degradation and technological hazards). Hazards can be single, sequential or combined in their origin and effects. Each hazard is characterized by its location, intensity, frequency and probability. (UN/ISDR, 2004). 44. This study is confined to Natural Hazards. Nevertheless, their assessment includes the consideration of anthropogenic factors, as these may aggravate the natural factors and in some cases even turn hazardous situations into actual disasters. This applies for instance to the flooding events in Dakar. In other cases, anthropogenic factors trigger extreme natural events, which would otherwise not materialize and remain hazards- potential events. Hazard versus Event, Disaster and Risk 45. While Hazards have been and may be differently defined in different contexts, it should be stressed that the term ‘Hazard’ needs to be differentiated from ‘Event’, ‘Disaster’ and ‘Risk’. Shortly, Hazards are no events, but only potential harmful events, whereas a Disaster is the impact of a hazard that has materialized to an event, on a community, human assets and/or natural resources. 46. Risk, according to an UNDP definition is “The probability of harmful consequences, or expected loss of lives, people injured, property, livelihoods, economic activity disrupted (or environment damaged) resulting from interactions between natural or human induced hazards and vulnerable conditions. Risk is conventionally expressed by the equation (UNDP, 2004). Risk = Hazard × Vulnerability. 47. Another concept that is widely applied in disaster research is risk being composed of the three components Hazard, Exposure, and Vulnerability (Schneiderbauer and Ehrlich in Birkmann 2006) illustrated with the Risk Triangle by Crichton (1999). 21 Figure 5: Risk Triangle, Crichton 1999 48. The risk concept presented in Figure 5 is based on a separation of the impact of hazard events into exposure and vulnerability. 49. Exposure refers to the quantity of the exposed elements: “Elements at risk, an inventory of those people or artifacts that are exposed to a hazard.� (UNDP, 2004). One may add environmental/natural assets to the list of potentially exposed items, although these are more difficult or often impossible to quantify. 50. Vulnerability refers to “The conditions determined by physical, social, economic, and environmental factors or processes, which increase the susceptibility of a community to the impact of hazards� (UN/ISDR, 2004). 51. While several definitions of vulnerability integrate or equate exposure and vulnerability (see Thywissen 2006), the authors of this study believe that the separation of the two terms makes definitely sense. Exposure in the above cited definition can be measured or estimated by studies such as this or similar approaches. Vulnerability, on the other hand, is a very complex and often in many respects intangible property that cannot be directly measured and hardly be quantified in absolute terms. Yet, the decrease of vulnerability (and the increase of coping capacities and resilience) has moved in the focus of disaster management strategies and superseded the mainly technical prevention or mitigation of disasters. In this respect the development of Vulnerability Indicators is of importance. 52. Figure 6 presents a third concept of risk. Here a longer-term perspective is explicitly taken, by adding the resilience of a system to the factors that determine the size of the risk. 22 Figure 6:: Risk as a Function of Hazard, Vulnerability, Exposure, and Resilience20 53. To understand the logic behind this concept, a short digression is made to the terms Coping Capacity and Resilience following Thywissen (2006), who puts all these disaster related terms and their manifold definitions in a logical and coherent synthesis. According to Thywissen, • Coping capacity encompasses those strategies and measures that act directly upon damage during the event by alleviating or containing the impact or by bringing about efficient relief, as well as those adaptive strategies that modify behavior or activities in order to circumvent or avoid damaging effects. • Resilience is all of these things, plus the capability to remain functional during an event and to completely recover from it. So resilience includes coping capacity but at the same time goes beyond it. 54. Thywissen (2006, p.38) points out: “If the extent of the damage or harm is defined also by the duration of the adverse effects and by its repercussions on people’s poverty, economy, or awareness, then vulnerability has to include coping capacity and resilience�. Analogy to illustrate the Risk Concept 55. For a better understanding, let’s take the following analogy: A hundred persons are exposed to the same bacteria (Hazard) during the same time period let’s say in a train compartment. Thus, the Exposure of these 100 people is identical. But yet they may not be infected by the bacteria to the same degree – some persons may stay healthy, the other ones may get infections depending on their individual degree of susceptibility (or Vulnerability). In addition, some of the infected people may stay ill for two weeks, some others only for 3 days, or similar, depending on their Resilience. 56. However, Vulnerability can in certain instances depend on the degree of exposure. (In the above analogy that might be the case if the people in the compartment are exposed to the bacteria over an extremely long time span, so that almost nobody can resist). In addition, Vulnerability and Resilience are interrelated as they partly depend on the same factors, but overall they include also many different factors. 20 Thywissen (2006) 23 3.2 Risk Factors and Related Information Sources 57. A comprehensive risk profile for a specific region and time frame would under the above definition of risk include the items listed in Table 1. Here, not all of these items are requested, and only for a part of them can information be derived in the frame of this study. The respective items are highlighted in yellow. Thus, Table 1 illustrates the relation between a more comprehensive Risk assessment and the elements that are provided in this study. Table 1: Components and Factors of RISK and Related Information Sources and Assessment Results Risk Components Factors of the Information Sources Resulting and Definitions used Risk (indicative examples) Information here Components (generic type, examples) Hazard Location Single hazard factor maps Spatial differentiation A potentially (e.g., DEM, geology, soil of single and multi- damaging physical sealing) hazard factors event, phenomenon Intensity Superposition of single and Spatial distribution of or human activity multi-hazard factors relative hazard that may cause the intensity loss of life or injury, Frequency E.g., climate data, disaster Statistics, property damage, data bases, etc. Assessments social and economic Probability Long term data series Damage-frequency disruption or Scientific assessments, relationship, environmental Monitoring of Current Return period degradation Trends (UN/ISDR, 2004) Vulnerability Physical E.g., Climate data to show Statistics, Maps, The conditions factors and recurring adverse or Assessment determined by processes deteriorating climate physical, social, conditions and economic, and consequences environmental factors Social factors Social data, e.g., education Statistics, Maps, or processes, which and processes Demographic Assessment increase the developments E.g., Maps of recent susceptibility of a urban expansion community to the Economic E.g., Individual Income Statistics, Maps, impact of hazards factors and data Assessment (UN/ISDR, 2004) processes Financial Situation of the public sector Adverse Economic Trends Environmental E.g., Maps of very E.g., coral reef maps, factors and sensitive elements, Statistics on climate- processes habitats, etc.; sensitive resources Monitoring of Habitat degradation Exposure Exposed Land Use Maps, census Distribution of Elements at risk, an people data residential areas and inventory of those population 24 inventory of those Exposed Land use maps Distribution of people, artifacts, and property residential and other environmental built-up areas resources that are Exposed Land use maps Distribution of exposed to a hazard. livelihood agricultural, (UNDP, 2004) industrial, “environmental commercial etc. resources� added by activities author Exposed Land cover maps Land cover types environment Resilience Direct relief Information about external Assessment, statistics Includes coping assistance and relief capacity and the needed and provided capability to remain through humanitarian functional during an organizations during or event and to following hazard events completely recover Adaptive Review of political E.g., Maps of from it. strategies that developments, educational hazardous zones for According to modify measures; Information land development Thywissen (2006) behavior about technical, social and that are to be avoided economic conditions and developments, etc. Prevention Information about planned E.g., maps showing measures and concrete measures dams, sewers, under way protection against erosion; Hazard-proof buildings, etc. Capability to Monitoring and Descriptions, remain documentation of assessments, maps of functional reconstruction, protection, those measures; during an prevention and early Time series of socio- event and to warning measures, and of economic data completely socio-economic factors covering the period recover from before and after the it hazard event Note: The contents highlighted in yellow are covered by this study. 3.3 Hazard types 58. Hazards are often grouped into three main categories according to their causes (Schneiderbauer and Ehrlich in Birkmann (ed.) 2006): • natural, • technological, and • anthropogenic or social causes. 59. The allocation of a hazard to one cause may be difficult, as hazards may have interrelated causes, and frequently one hazard is triggered by another. 25 60. Table 2 contains an overview of natural hazards and combined natural/anthropogenic hazard types. Four types of hazard events (tropical cyclones, earthquakes, floods and droughts) were responsible for 94 percent of the deaths triggered by natural disasters between 1980 and 2000 worldwide (UNDP, 2004). Many of these so-called natural hazards have also anthropogenic causes or may be triggered by human factors. Only a few hazards are purely natural. Table 2: Types of Natural Hazards, Potential Anthropogenic Causes and Relations to Other Hazards Natural and combined Hazard Types Potential Anthropogenic Major relations to natural/anthropogenic factors or triggers other hazards Hazards Geological Hazards Earthquake Tsunamis, fire Volcanic Land slides eruption Land Slide and Soil destabilization as a Mud flow result of deforestation, agricultural, technical, or leisure activities Subsidence Excessive extraction of Flooding water or other substances Coastal Extraction of sand or other Flooding erosion materials Hydrologic Hazards Flood Soil sealing, especially Epidemic, famine rapid urbanization Canal contamination River regulations Snow and ice Leisure or technical avalanche activities Deforestation Tsunami May be caused by earth quake Storm surges Potential increase with Flooding, coastal climate change erosion 26 Natural and combined Hazard Types Potential Anthropogenic Major relations to natural/anthropogenic factors or triggers other hazards Hazards Atmospheric Hazards Meteorological Soil degradation in Famine Drought combination with dry Subsequent severe meteorological conditions soil erosion may lead to edaphic droughts Tropical Potential increase with Flooding cyclone climate change Tornado Potential increase with climate change Hail storm Potential increase with climate change Heat wave Potential increase with climate change Wildfire Hazards Carelessness; Intended fires running out of control Criminal actions Biospherical Hazards Epidemic Lacking education Poor sanitation and waste management Famine Complex socio-economic- May cause or natural reasons, mainly trigger human related to poverty besides actions that in turn natural conditions and increase the armed conflicts probability of further hazards 3.4 Hazards and Climate Change 61. “The most vulnerable industries, settlements and societies are generally those in coastal and river flood plains, those whose economies are closely linked with climate- sensitive resources, and those in areas prone to extreme weather events, especially where rapid urbanization is occurring.� (IPCC, 2007). This statement of the 4th IPCC Assessment Report fully applies to the Dakar Metropolitan Area. Natural hazards cannot be regarded separately from climate change and its variability. It is widely known that the number and severity of natural disasters has increased during the past decades. 27 62. Climate change is interrelated with natural hazards in several ways. In this context, it is important to look at an area in a greater spatial context. While sea level rise is the most direct and prominent climate-change related threat for low lying coastal cities, deteriorating climatic conditions in the wider surroundings of those cities will in many places increase human migration into the cities and thus accelerate the already rapid urbanization. Unmanaged urban growth aggravates or even triggers natural hazard events, as urban expansion tends to occur in hazard-prone peri-urban areas which attract poor people with a cheaper land price. At the same time, climate change and its variability weaken the resilience capabilities of those cities. 28 4. Spatial Analysis 4.1 Overview and the Definition of Peri-Urban Areas 63. This section aims at providing the conceptual and methodological basis for the generation of hazard and vulnerability maps for Dakar, Senegal. While all types of “natural� hazards relevant for the area are considered, a selection of the following hazards for a detailed spatial analysis was made: Flooding (inland flooding of depressions), Coastal Erosion, and Sea Level Rise. Local subsidence over time was not considered in this study. The occurrence of flooding within the Dakar Metropolitan Area has recently been increasing for climatic and anthropogenic reasons, and constitutes a major threat especially for newer settlements in unsuited low-lying terrain. Coastal erosion as a more steadily acting, but very hazardous process in the area was included in the detailed analysis along with scenarios of sea level rise. 64. The generation of all map inputs needed for the risk analysis are described in detail. These maps include land use at three different points of time (1988, 1999, and 2008), Digital Elevation Model (DEM) based data layers, geology, hydrology, single-hazard maps (flooding, coastal erosion and sea level rise scenarios), and a multi-hazard map integrating all three hazards analyzed. 65. The Spatial Analysis combines the hazard mapping results with population maps, land price data, and land cover information in order to derive the exposure of these items to the hazards. As an important intermediate step, the population data from the census are disaggregated to the much finer level of 250 meter grid cells using housing density data derived from the satellite data. 66. The scale of the spatial analysis is regional, a level of detail that is relevant for the awareness-raising and institutional engagement purposes intended here. Extensions to this analysis, not covered at the current pilot stage, may include (a) consideration of other natural hazards; (b) more thorough evaluation of economic impacts of hazards taking into account direct and indirect damages; and (c) additional analysis of more detailed geographic information systems (GIS) data, cadastres, and first-hand ground-truthing, to ascertain detailed risks faced by specific populations and built areas in the broad areas indicated here as under risk. Box 3: Justification for Selection of Specific Natural Hazards According to the historical data (see previous chapter) available for the Dakar Metropolitan Area and surroundings, flooding, coastal erosion and drought are identified as the most frequent, prevalent and significant natural hazards. While inland flooding and coastal erosion are direct risks to urban development in Dakar, the natural risk of drought must be viewed in wider geographic perspective. In Senegal, drought is documented to be a driver of migration to Dakar, but cannot be regarded as direct hazard to urban development within the Dakar Metropolitan Area. Sea level rise has been selected as it is the most pertinent climate-change related natural risk for the coastal city of Dakar. 29 Study area 67. The Area of Interest covers a surface of approximately 582 square kilometers. It is shown in the map and satellite images below: Figure 7: The Dakar Metropolitan Area, Covering the Departments of Dakar, Guédiawaye, Pikine, and Rufisque21 Peri-urban areas 68. There is no consensus on a definition of peri-urban (urban fringe) areas. Peri-urban areas have been defined in a qualitative way, for instance, as a zone of transition in land-use 21 Sources: (2008) Africa’s Urbanization for Development: Understanding Africa’s Urban Challenges and Opportunities Built-up area in 2008 (white line), as observed from aerial photography—Authors’ manual analysis. Administrative area (blue line)—maplibrary.com. ground imagery—GoogleEarth & embedded sources. 30 lying between the continuously built up urban and suburban areas of a central city (Pryor, 1968); a transition zone between fully urbanized land in cities and areas in predominantly agricultural use (Rakodi, 1998); or an area subject to urban expansion on the edge of cities and environmentally fragile urban area which is unstable and unfit for occupation (World Bank, 2007). Buxton et al. (2006) considered proximity to primary metropolitan areas and to their physical structure and form, especially types of land use. 69. In addition, there have been attempts to define peri-urban areas quantitatively, for example in terms of the contiguous territory to a central place having a density of at least 1,000 persons per square mile (US Census Bureau).22 However, there is a general agreement that peri-urban areas cannot be defined uniformly only by one measure. As Webster (2002) notes, contiguous built up (peri-urban) areas sometimes extend as far as 150 km from the core city – or, in the case of Chinese cities, as far as 300km. 70. This study adopts an eclectic peri-urban definition, combining quantitative and qualitative local information and taking into consideration know-how of local experts. The communes in the Dakar Metropolitan Area are classified into 3 subdivisions: (i) Urban areas, including communes in the Department of Dakar (the city center) and four communes (Rufisque, Bargny, Diamniadio and Sébikotane) and three joint districts (Rufisque Ouest, Rufisque Nord and Rufisque Est) in the Department of Rufisque, which are in general, areas with strong urban and industrial economic activities; (ii) Rural areas, including two large communes in the Department of Rufisque defined locally as “Rural Communes,� Sangalkam, and Yène; and (iii) Peri-urban areas, which are the areas lying in between these other types of areas, including mixed land use and relatively lower densities (Figure 8).23 22 US Census Bureau 1995 at: http://www.census.gov/population/censusdata/urdef.txt. 23 Specifically, Urban areas include Biscuiterie, Cambérène, Dieuppeul-Derklé, Fann-Point E-Amitié, Gorée, Grand Yoff, Grand-Dakar, Gueule Tapée-Fass-Colobane, Hann-Bel Air, HLM, Médina, Mermoz-Sacré Cœur, Ngor, Ouakam, Parcelles Assainies, Patte d'Oie, Plateau, Sicap Liberté, and Yoff (the Department of Dakar); Bargny, Rufisque Est, Rufisque Nord, and Rufisque Ouest communes (the Department of Rufisque). Rural areas are Sangalkam, and Yène rural communities (the Department of Rufisque). Finally, Peri-urban areas cover Golf Sud, Ndiarème Limoulaye, Sam Notaire, and Wakhinane Nimzatt (the Department of Guédiawaye); Dalifort, Diamaguène Sicap Mbao, Djidah-Thiaroye Kaw, Guinaw Rail Nord, Guinaw Rail Sud, Keur Massar, Malika, Mbao, Pikine Est, Pikine Nord, Pikine Ouest, Thiaroye Gare, Thiaroye Sur Mer, Tivaouane Diaksao, Yeumbeul Nord, and Yeumbeul Sud communes (the Department of Pikine). 31 Figure 8: The Classification of Urban, Peri-Urban, and Rural areas in the Dakar Metropolitan Area Some remarks on the limitations of the study 71. It can be seen that this study compiles and spatially analyses many components of hazard (or disaster) risks, especially with regard to the spatial distribution and interaction of major risk factors. On the other hand, the study is limited in the following respect: With the available input information only the spatial distribution of relative hazard potentials can be estimated. The absolute damage-frequency relationship and probabilities and their spatial differentiation cannot be derived. So for instance absolute inundation levels and probabilities cannot be determined, and it is not possible with a remote study to infer concrete figures about coastal erosion. Likewise, the projected flooded areas in connection with the assumed seal level rise scenarios are the result of simplifying assumptions (the only possible in such a study), and do by no means mean that the flooded area will actually be of that extension. 72. With regard to vulnerability and exposure, mainly exposure-related information can be derived. Vulnerability, as argued below, cannot be measured directly, but only inferred with the help of indicators. In the context of this study we derive spatial indicators of vulnerability from hot spots of exposure, assuming that extreme concentrations of exposure will lead to locally increased vulnerability. 73. Finally, for this study no information on protection and prevention measures that may be taken or planned by local authorities are known and included. 4.2 Map Generation 74. The map generation described in this chapter comprises 32 • Land use and land use change for two periods (1988-1999, 1999-2008) • Single hazard factor maps • Single hazard potential maps, and • Multi-hazard potential maps. 75. In Table 3, an overview is provided about these maps, including input data used, methods, map contents, and references for their production. Table 3: Overview of the Produced Maps Map Type Input Data Approach Contents References Land Use and Land use Change Maps Land use and Optical satellite Image Co- Current detailed Modified land use change data registration land use (2008) CORINE land and cover legend Visual Image interpretation Land use changes 1999 – Change analysis 2008 and 1988 – Topology 1999 checks and Legend: 13 and validation 7 classes, respectively Single Hazard Factor Maps Corrected and Hybrid (based DEM correction Hydrologically ANUDEM adapted on X and C based on correct, program by Digital Band) SRTM imposed depression-less Michael Elevation (Shuttle Radar drainage DEM Hutchinson24 Model (DEM) Topography condition; Mission) Adaptation of seashore level to high resolution satellite imagery DEM derivate: Corrected DEM Average Slope in degrees Burrough and maximum McDonell25 Slope technique 24 Hutchinson, M. F. (1993): Development of a continent-wide DEM with applications to terrain and climate analysis. In Environmental Modeling with GIS, ed. M. F. Goodchild et al., 392–399. New York: Oxford University Press. 25 Burrough, P. A., McDonell, R.A. (1998): Principles of Geographical Information Systems (Oxford University Press, New York), p. 190. 33 DEM derivate: Corrected DEM State of the Art Aspect in ArcGIS Aspect GIS software degrees: software 0° = N, 180° = S DEM derivate: Corrected DEM Derivation of Eight direction Jensen and Flow direction steepest descent flow model; Dominingue26 (or maximum Flow direction drop) among in degrees eight directions DEM derivate: Corrected DEM Flow derived as Flow Jensen and Flow accumulated accumulation Dominingue accumulation weight of all expressed as no. cells flowing of cells into each contributing to downslope cell the accumulated in the output water at a given raster location DEM derivate: Corrected DEM Catchment areas ARC GIS software Watersheds Geology Detailed Map Digitizing and Geological map Lo and Diop27; by Lo and Diop merging the with 11 Centre de Suivi (1:25.000), and two input maps geologic units Ecologique Geologic map (CSE, 1:200.000 http://www.cse. sn/) Rocky Cliffs Topographic Visual map Map containing Institute Map 1 :50.000 interpretation coastal stretches Geographique (1983) and delineation with rocky cliffs National France of respective (IGN, features http://www.ign.f r/) Soil Sealing 2008 Land use Map re-coding Binary map map derived in containing this study sealed and non- sealed surfaces Hydrologic Topographic Visual map Map containing: Institute Features Map 1 :50.000 interpretation Geographique 26 Jenson S. K., Domingue, J.O. (1988): Extracting Topographic Structure from Digital Elevation Data for Geographic Information System Analysis. Photogrammetric Engineering and Remote Sensing 54 (11): 1593- 1600. 27 Lo, P.G., Diop, M.B. (2000): Problems associated with flooding in Dakar, western Senegal: influence of geological setting and town management. Bull. Eng. Geol. Env. 58, p 145-149. 34 Features (1983) and delineation Wetlands/Ponds National France (without of respective ; (IGN, permanent features Temporary http://www.ign.f water bodies) flooded; r/) Temporary water channels Permanent Topographic Visual map Map containing: Institute water bodies Map 1 :50.000 interpretation Geographique Permanent (1983) and delineation National France water surfaces; High resolution of respective (IGN, features, Permanent http://www.ign.f Satellite data merging water channels r/) 2008 features from both sources Single Hazard Potential Maps Flood Potential DEM and Re- Map of Relative Lo and Diop; Map derivates ; classification of Flood Potential: Mbow et al.28 Geological all map input five levels from Map ; layers using no risk to very five-point high plus Soil Sealing scales and Map permanent additive water bodies Hydrological merging features Coastal Geological Re- Map of Relative Fall et al.29 Erosion Map ; classification of Coastal Erosion: Potential all map input Aspect of the five levels from coast facing the layers using no risk to very predominant five-point high wave impact; scales and additive Slope of the merging coast Coastal DEM Connectivity Map of flooded Poulter and Inundation analysis for 2 areas in Halpin30 Potential seal levels rise different sea scenarios level rise scenarios 28 Mbow, C., Diop, A., Diaw, A.T., Niang, .. (2008): Urban sprawl development and flooding at Yeumbeul suburb (Dakar-Senegal). African Journal of Environmental Science and Technology Vol. 2 (4), pp. 075-088 http://www.academicjournals.org/AJEST/PDF/pdf%202008/April/Mbow%20et%20al.pdf. 29 M.Fall, R.azzam,C.Noubactep. 2006. A multi-method approach to study the stability of natural slopes and landslide susceptibility mapping. Engineering Geology 82: 241-263. 35 Multi - Hazard Potential Maps Multi Hazard Relative Flood Re- Map of showing Khatsu and Potential Potential classification of potential for v.Westen31 Coastal Erosion all map input natural hazards Potential layers and in Dakar merging Coastal Inundation Potential 4.2.1 Land Use Mapping 76. These base maps should include urban areas and rural communes, show built-up and non-built up areas, and highlight areas and direction of observed rapid urban expansion based on comparison of time series data. 77. The objective of “map generation� is to generate multi-temporal base GIS maps of the Dakar Metropolitan Area from satellite imagery, depicting built-up areas, the nature of major infrastructure and areas with rapid urban expansion as well as non-built up areas such as agriculture and forests. The project should generate comparable information as input to diagnostics for urban extent and peri-urban expansion in the Dakar Metropolitan Area. GIS maps should show the status and development of land cover/use within the area for three points in time, i.e. 1988, 1999, and 2008. In order to create multi-temporal base GIS maps for the Dakar Metropolitan Area three different acquisition dates within the last 3 decades were considered. Table 4: Available Multi-temporal Satellite Scenes Acquisition Date Sensor Resolution 2008-09-08 / 2007-09- SPOT 5 2.5m 26 1999-11-04 / 1999-11- Landsat ETM 15 m (resolution 11 merge) 1988-03-10 Landsat TM 30m Base mapping 2008 78. Mapping started with the most recent satellite data with the highest spatial resolution (SPOT satellite image 2008 with 2.5 meter resolution). Technical procedures are based on automated object-based classification to detect specific land cover elements, complemented by GIS-based post-processing and visual image interpretation to add additional land 30 Poulter, B., Halpin, P.N. (2007): Raster Modelling of Coastal Flooding from Sea-Level Rise. Int. Journal of Geographical Information Science, Vol. 22, No.2, p. 167-182. 31 Khatsu, P., v.Westen, C.J, (2005): Urban multi-hazard risk analysis using GIS and Remote Sensing: A case study from Kohima Town, Nagaland, India, ACRS Proceedings, p9. . http://www.aars-crs.org/acrs/proceeding/ACRS2005/Papers/URM1-2.pdf 36 cover/use classes that need detailed on-screen investigation and verification. Additional sources for identification were scanned topographic maps and information extracted from online-sources (webmaps). 79. The current nomenclature of 13 thematic classes represents the maximum information content that can be realistically derived from SPOT satellite imagery. For subsequent mapping of land use changes an aggregated land use nomenclature was used to allow for reliable detection and recognition of land use changes using Landsat imagery. Figure 9: Detailed Land Use Map of Dakar, 2008 37 Figure 10: Comparison of Detailed and Aggregated Land Use Map for Year 2008 Detailed Land use classification 2008 (13 Aggregated Land use classification 2008 (7 classes) classes) Central business district, very dense urban fabric 80-100% Residential areas Mixed use, dense urban fabric 50-80% Non-residential urban areas Primarily residential, low to medium density urban fabric 10-50% Transport units Isolated rural structures Green urban areas Industrial / commercial areas Forest Roads (asphalted roads > 10m width) Water Railways Agricultural areas and other open non-urban land Port areas & Airports Mineral extraction sites Green urban areas Forests Water Agricultural areas and other open non-urban land Base mapping 1988 and 1999 and changes 80. Mapping of the historical land use for the reference dates 1988 and 1999 was done by using lower resolution Landsat imagery (15m merged image for 1999, 30m image for 1988). Methods used comprise automated land cover derivation and visual interpretation. Table 5: List of Thematic Classes for land use change mapping Common classed land cover/use maps 1988, 1999 Additional classed only for and 2008 land cover/use map 2008 Central business district, Urban areas Residential areas very dense urban fabric 80- 100% Mixed use, dense urban fabric 50-80% 38 Common classed land cover/use maps 1988, 1999 Additional classed only for and 2008 land cover/use map 2008 Primarily residential, low to medium density urban fabric 10-50% Isolated rural structures Industrial / commercial areas (incl. water supply Non-residential urban infrastructure, sea walls and areas flood defenses) Mineral extraction sites Roads (> 10m width) Transport units Railways Port areas & Airports Green urban areas Forests Non-urban areas Water Agricultural areas and other open non-urban land 39 Figure 11: Land Use Changes, 1988, 1999, and 2008 40 Residential areas Non-residential urban areas Transport units Green urban areas Forest Water Agricultural areas and ther open non b 4.2.2 Hazard Potential Maps Single Hazard potential maps 81. The purpose of the single hazard factor maps is to provide the necessary input data layers for a spatial representation of the natural hazards in Dakar. As also human factors contribute to these hazards, they are also contained in this map array (i.e. soil sealing). For each hazard type we classified the available input parameter for risk potential according to peer reviewed published scientific literature of related studies in the area and merged the appropriate parameters using best practice knowledge and GIS techniques. The input parameters, spatial analysis and results are described for each natural hazard hereafter. Input data maps are also shown in Annex. 82. In order to facilitate analysis for the Dakar Metropolitan Area and to accommodate the different resolutions and accuracies of the input data in the final multi hazard map, the single hazard risk potential grids were aggregated to 250 x 250 meter cells using a nearest neighborhood method. Comparability of all hazard potential maps is ensured by using a homogeneous classification scheme for each hazard map as shown in the following table: Table 6: Key for classifying and reporting relative natural hazard potentials at aggregated resolution (250 meter) High Moderate Low No risk Relative Flood Potential (RFP) 83. To derive a relative flood potential map of Dakar all factors were considered that were in past studies identified to be significant and/or relevant parameters contributing to flooding by rain and for which a data layer was available. These are: • Elevation (from absolute, depressionless DEM) • Geology as an indicator for the permeability of the ground • Degree of soil sealing • Slopes of the topography • Degree of potential flow accumulation 41 • Hydrological features • Permanent water bodies 84. Each parameter was reclassified for relative flood potential according to the studies by Lo and Diop (2000) and Mbow et al.(2008), and to best practice knowledge by our scientific experts. Figure 12: Relative Flood Potential Layer for Dakar, Senegal Aggregated to 250m Cells. Coastal Erosion Potential (CEP) 85. In order to derive a coastal erosion potential map of Dakar, Senegal, all factors were incorporated that were in past studies identified to be significant and/or relevant parameters contributing to coastal erosion and for which a data was readily available in an adequate resolution. These are: • Geology as an indicator for the stability of the coast • Aspect of the coast facing the pre-dominant wave impact • Slope of the coast • Information on the direction of pre-dominant wave impact, based on Satellite Radar data and obtained through the database of http://www.surfline.com. 86. Each parameter was re-classified for coastal erosion potential based on the studies by Fall et al. (2006) and according to best-practice knowledge by our scientific experts. 42 Figure 13: Coastal Erosion Potential Layer for Dakar, Senegal Aggregated to 250m Cells. Sea Level Rise 87. Two scenarios were assumed: A mean sea level rise (MSL) of 1m and a mean sea level rise of 5m. These two scenarios are not considered to be likely in the short term, but representing them in the maps of the city can give stakeholders and policy makers a sense of the magnitude of extreme climate change impact. The projection of the sea level rise was based on the DEM. 88. Sea level rise can be expected to cause coastal erosion and inundations. During that time an unforeseeable sequence of coastal erosion, protection measures and actual inundation will occur, which will vary depending on various conditions, especially topography, geology and human factors. Meteorological factors during the period of sea level rise, such as storm frequency, can also not be projected. 89. For the technical approach a comprehensive study of Poulter and Halpin (2007) has been consulted, which shows the high variability of resulting flood extents when using varying horizontal DEM resolutions and different assumptions made on hydrological connectivity. 43 Figure 14: 1m Coastal Inundation Potential for Dakar, Senegal Aggregated to 250m Cells Figure 15: 5m Coastal Inundation Potential for Dakar, Senegal Aggregated to 250m Cells 44 Multi hazard potential map 90. The single hazard potential maps aggregated to 250m grid level were merged with the objective to obtain a map that reflects both, the type and the potential of all three analyzed natural hazards for a given location (250x250m cell). In order to achieve this, the single hazards and their potential were combined to a new set of hazard potentials that includes possible combinations of several hazards, single hazards, where only one hazard type occurs, and areas with no hazard potential. 91. For multi-hazard maps, this is a standard approach (see for example Khatsu and v.Westen, 2005), even though different variants of the approach may be applied. The three single hazards • Relative Flood Potential (RFP) • Coastal Erosion Potential (CEP), and • Coastal Inundation Potential (CIP) 92. Were combined to a multi-hazard map containing both, the hazard types and their potential. However, due to the complexity of the map, not all original degrees of the hazard potentials could be maintained. Consequently, an aggregation of the hazard potential to the following two levels was applied: 1) Low and moderate potential of the respective hazard (aggregating the low and moderate potential) 2) High Potential of the respective hazard (remained the same as in the single hazard maps). 93. One exception from this scheme was made for the RFP as a single risk, as the aggregation of low and moderate RFP would result in a very large, undifferentiated aggregated class. For this reason, moderate and low RFP are depicted as individual classes. 94. As a result, the hazard classes listed in Table 7 were derived for the multi-hazard map. Figure 16 shows the resulting multi-hazard map with the classes listed in Table 7. Figure 17 shows another representation of the multi-hazard potential, where the single hazards are juxta-positioned instead of overlaid. Table 7: Classes Contained in the Multi-Hazard Map Hazard Classes derived for the Original Hazard Potential in Single multi-hazard map Hazard Map (possible alternatives) No Potential No Potential of any single hazard Low and Moderate CIP/CEP CIP low or moderate, CEP low or moderate Low and Moderate CIP/RFP CIP low or moderate, RFP low or moderate Low RFP RFP low, others no potential Moderate RFP RFP moderate, others no potential Low and Moderate CEP CEP low or moderate, others no potential 45 Hazard Classes derived for the Original Hazard Potential in Single multi-hazard map Hazard Map (possible alternatives) Low and Moderate CIP CIP low or moderate, others no potential High CIP/CEP CIP high, CEP high CIP high, CEP moderate or low CIP moderate or low, CEP high High CIP/RFP CIP high, RFP high CIP high, RFP moderate or low CIP moderate or low, RFP high High RFP RFP high, others no potential High CEP CEP high, others no potential High CIP CIP high, others no potential Figure 16: Multi-Hazard Potential in Dakar, Senegal, Including Relative Flood Potential, Coastal Erosion and 1m Coastal Inundation Potential Scenarios 46 Figure 17: Multi-Hazard Potential in Separate Maps 47 4.3 Spatial Analysis 4.3.1 Methodology 4.3.1.1 Overview 95. The objective of “Spatial Analysis� is to obtain information about the spatial relationship between the natural hazard potentials and the exposed “items�, i.e. population, land use units, and economic assets. For this purpose, different spatial analyses were to be performed, generating statistical results and maps containing the following information: • Potential hotspots and risk areas illustrating type and degree of vulnerability for the different risk categories and scales of risk defined previously (Task 1) • Built-up areas exposed to risk: general characterization of the urban land use (i.e. residential, industrial/commercial, etc.) including type of vulnerability (Task 2) • Non-built-up areas exposed to risk: general characterization of the land use, topography, and soil, including type of vulnerability (Task 3). 96. As an outcome of the theoretical part of this project and of the common clarification, instead of the term “vulnerability� the term “exposure� should be used in this context for the following reason: What we measure with remote sensing and auxiliary information in such a project is the kind and quantity of items, which are exposed to potential hazard events. We cannot infer conclusions about their vulnerability without further information. Such information would for instance include institutional capacities at different administrative levels, state of the infrastructure(s), public income and expenses, educational level, etc. This type of “vulnerability analysis� however is not covered in this study. 97. Instead, we relate the spatial distribution of the hazard potentials to different quantities and types of exposed items, i.e., people, land prices, and land use. Vulnerability is thus not directly derived, but a relative spatial increase of vulnerability is indicated in the map legends as a result of a concentration of exposed items and hazard potentials (see Figure 20). This follows the logic that a higher vulnerability of exposed items can be expected at locations where spatial concentrations (hot spots) of both hazard potential and degree of exposure occur. 98. The objectives stated in the three above listed items were first translated into a list of map products that can be generated on the basis of the available input data and results generated so far (Table 8). The relation of these map products to the original, above listed tasks is indicated in the table (Task No., column 1). 99. Along with the maps, statistical analyses were performed, using MS Excel. In the following sections, the methodology and results of the produced statistics and the maps are presented, structured according to the three above listed project tasks. 48 Table 8: List of Maps Produced for Spatial Analysis Task Map Name Map Contents Approach No. Social exposure and vulnerability 1 Social exposure No. of residential people per Disaggregation of residential and vulnerability hectare in 2008 within population figures to 250 m 2008 potential (multi)-hazard zones grid cells on the basis of land on a continuous scale use data and overlay with (multi)-hazard zones. The 1 Social exposure No. of residential people per population figures are shown and vulnerability hectare in 1999 within on a hectare basis. 1999 potential (multi)-hazard zones on a continuous scale 1 Social exposure No. of residential people per and vulnerability hectare in 1988 within 1988 potential (multi)-hazard zones on a continuous scale 1 Population Population development 1999 Derivation of the population development and - 2008 in absolute numbers on development on a 250m grid vulnerability 1999 a 250 m grid basis within basis and overlay with to 2008 (multi)-hazard zones on a (multi)-hazard zones continuous scale 1 Population Population development 1988 Derivation of population development and - 1999 in absolute numbers on development on a 250m grid vulnerability 1988 a 250 m grid basis within basis and overlay with hazard to 1999 potential (multi)-hazard zones zones on a continuous scale Economic exposure and vulnerability 1 Economic Land price values in $ /sq km The land prices are on a sq km exposure and and potential (multi)-hazard basis and were converted from vulnerability 2008 zones the Senagalese Franc into U.S. dollars Overall social and economic vulnerability 1 Overall social and Land prices and population Based on a classification of economic density overlaid with potential the land prices and population vulnerability 2008 (multi)-hazard zones density 6 combined classes for 49 vulnerability 2008 (multi)-hazard zones these two properties were built and overlaid with (multi)-hazard zones Built-up Areas exposed to Hazards 2 Built-up areas with Built-up land use classes in Overlay of built-up land use (multi)-hazard 2008 and potential (multi)- classes and (multi)-hazard potential 2008 hazard zones zones; (multi)-hazard zones are only depicted within built-up areas 2 Built-up areas with Built-up land use classes in (multi)-hazard 1999 and potential (multi)- potential 1999 hazard zones 2 Built-up areas with Built-up land use classes in (multi)-hazard 1988 and potential (multi)- potential 1988 hazard zones 2 Share of industrial/ Percentage of Derivation of percentage of commercial/traffic industrial/commercial/ traffic these economically used areas areas and areas within 250 m grid cells within 250 m grid cells, vulnerability 2008 overlaid by and (multi)-hazard shown together with (multi)- zones hazard zones (multi)-hazard zones are depicted for the total area Non-built-up Areas exposed to Hazards 3 Non built-up areas Non built-up land use classes Overlay of non built-up land with (multi)-hazard in 2008 and potential (multi)- use classes and (multi)-hazard potential 2008 hazard zones, soil types zones together with soil types; (multi)-hazard zones are only 3 Non built-up areas Non built-up land use classes depicted within non built-up with (multi)-hazard in 1999 and potential (multi)- areas potential 1999 hazard zones, soil types 3 Non built-up areas Non built-up land use classes with (multi)-hazard in 1988 and potential (multi)- potential 1988 hazard zones, soil types 50 4.3.1.2 Generation of Input Data 100. For the generation of the input data of the Spatial Analysis, several data sources were used, which were partly generated in this project, and partly derived from respective source data. A list of the data sources and the derived input data for the Spatial Analysis is contained in Table 9. The table includes a short description of the data preparation and generation methods. A more detailed description of the population disaggregation is added following Table 9. Table 9: Source Data, Preparation Methods, and Resulting Input Data for the Spatial Analysis Source data Vintage and Data Method of data Resulting input for source preparation/generation the Spatial Analysis Satellite data from Generated in this Automatic image Land cover data SPOT and project classification and visual with 10m raster size LANDSAT correction and 13 (detailed) and 7 satellites refinement (generalized) classes Visual change detection Share of industrial/ commercial/traffic areas Hazard Potentials Generated in this Various GIS intersection Hazard potentials project and modelling with 250m raster approaches size Transformation in 250m raster data sets Demographic data Land cover data Population Disaggregegated from different (1988, 1999, 2008) disaggregation population data: administrative generated in this modelling using in- Number of people levels; project house developed models per hectare, Land cover data Demographic data rasterized with 250 (generated in this (1988, 1998, 2005) The figures for 1999 and m grid resolution project) from: 2008 were extrapolated Population Ministry of based on averaged development economy and population growth rates between 1988 and finances of Senegal 1999, and 1999 and 2008 National agency for demography and statistics – Dakar 51 Source data Vintage and Data Method of data Resulting input for source preparation/generation the Spatial Analysis Digital Elevation 2000 Bilinear interpolation Elevation data with Model Hybrid X and for smoothing, 10m raster size DLR C Band SRTM- resampling; Data, 30m (90m for Filling gaps with 90m gaps) SRTM Topographic Maps 1983 Extraction of permanent Permanent water Dakar, Bargny, National Geographic water bodies in bodies as depicted in Kayar Institute of France combination with land the Spatial Analysis cover maps Maps 1:50.000 Soil map of Senegal 1965 Scanned, vectorized, Digital soil map, translation of soil types, used for 1:1.000.000 Institute of research geographical cartographic overlay in Dakar (centre coregistration, O.R.S.T.O.M) assignment of map colors Administrative 2008 Analogue borough Digital boundaries, various boundaries were administrative National agency for levels vectorized and partially boundaries demography and corrected; statistics - Dakar Derivation of district boundaries Land Price Values 2008 Currency translation Land prices in U.S. from Senegalese Franc $/sq km Local consultation to U.S. $ Derivation of the Disaggregated Population Data for 1988, 1999, and 2008 101. Information about the distribution of population within Dakar was essential to estimate the local social exposure to the hazards. The better the allocation of the population can be assessed, the better the exposure of the population to natural hazards can be estimated. Therefore it is of greatest interest to disaggregate the population from municipality or district level to a more detailed level such as 250 m grid cells. 52 102. For the study area census data was obtained for the years 1988, 1998 and 2005. While for 1988 the census data was available on district level, 1998 census data was only available for whole Dakar, and for 2005 census data was available on municipality level.32 103. To use the census data for the disaggregation procedure, which is based on the housing density (proportion of residential areas) derived from earth observation data, the census data has to be synchronized with the land cover data. Land cover data was derived from the satellite images that were available for 1988, 1999 and 2008. Figure 18: Synchronizing Land-Cover/Use Data with the Census Data 104. The following steps were performed for synchronizing land cover data and the census data (see Figure 15): • Extrapolation of overall Dakar population to 1999 based on population growth between 1988 and 1998. • Disaggregation of population on commune level based on population of 1999 and change of residential build up area between 2005 and 1999 • Disaggregation of population on commune level based on population of 1988 and change of residential build up area between 1999 and 1988 • Extrapolation of population to 2008 on commune level based on population growth between 1999 and 2005. 105. The population information on municipality level serves as input for the population disaggregation to 250m grid cells. Applying housing density (proportion of residential areas) as a proxy for population density allows estimating the local population distribution. This approach can be formalized as follows: Pdens = k Hdens (1) POP = (2) 32 Agence Nationale de la Statistique et de la Demographie 53 106. where Pdens and Hdens are the population and housing density respectively, POP is the total population of the municipality and Ai corresponds to the area of the housing density i. The factor k, representing the relationship between population and housing density, can be derived by solving equation (2). The local population density is then calculated from equation (1). The following assumptions were made when applying this approach: (i) the population density is proportional to housing density, (ii) no population occurs outside housing areas, and (iii) dependency between population and housing density is constant within a region. Figure 19: Comparison of Census-Based (Department) 33 and Spatially Disaggregated Population Distribution 50 and below inh/ha 51 - 100 inh/ha population per population per 101 - 250 inh/ha disaggregated population density disaggregated population density department statistics department statistics 251 - 500 inh/ha per 250m Grid cell (2008) per 250m Grid cell (2008) Above 500 inh/ha 3 dimensional population per 250m grid cell (2008) 33 Source : UN-Habitat (2007), Stratégie de développement urbain du Grand-Dakar. Diagnostic Territorial, Rapport Consolidé. 54 107. Figure 16 shows the effect of the disaggregation. After the disaggregation, uninhabited areas are separated from inhabited areas, and within the latter the population density is shown. This gives a significantly “sharper� image of the actual distribution of the resident population. 4.3.1.3 Mapping and statistical analysis 108. The Spatial Analysis Maps are raster based maps with a general raster size of 250m. This is the reference grid size used for the derivation of the hazard potential, the disaggregation of the population and for the derivation of the share of the industrial/ commercial/traffic areas shown in one of the maps covering. Exempted from the 250m raster size is the representation of the land cover classes, properties of the non-built-up areas (soil types, DEM), traffic lines, and district boundaries. These items are depicted in 10m resolution. 109. For the maps showing the spatial analysis results, the hazards potentials were generalized for cartographic simplicity. High Potential zones were kept separately, while Moderate and Low Potential zones were combined, which resulted in a total of six Hazard Potential zones (Figure 17).34 Figure 20: Hazard Potential Zoning Applied to the Satistical Analysis Maps 110. Cartographic overlay techniques using dots of different colors and size were used for the superposition of the exposed items and the hazard potentials (see Figure 18). The dot grids were shifted to each other in order to enable the representation of co-occurring hazard potentials. 34 For the Coastal Inundation Potential in the Spatial Analysis, only the 1m Sea Level Rise scenario was used. The latter comprises possible coastal inundations to over 3 to 4 meters, because spring-tides and storm waves have to be accounted for. A sea level rise of 1m is expected to occur within more than half a century to over three centuries (current estimates move towards the shorter end). The 5m sea level rise scenario was not included in the Spatial Analysis. This would go far beyond the temporal applicability of the other map parameters; further on there is complete uncertainty as to what time period may be needed for a 5m rise. 55 Figure 21: Detail of Hazard map Note: see Figure 20 for Legend 111. The statistical analysis was performed using MS Excel. All three Hazard Potentials, i.e. High – Moderate – Low where used separately for the analysis, while for the cartographic presentation Low and Moderate Hazard Potentials where combined. The statistics were derived separately for the following three spatial units: (i) Urban areas, (ii) Peri-urban areas, and (iii) Rural areas. The subdivision classification is based on the consultation with local experts (Figure 8). 4.3.2 Spatial Analysis Results 4.3.2.1 Land Cover Changes 112. The land cover changes are not part of the spatial analysis, but they are shown here in order to highlight the massive urban development that has taken place during the observed time period. Table 10 provides an overview about the major land cover changes between 1988 and 2008. In the last column the changes are indicated in percent of the land cover 1988: both residential and other urban areas have increased by over 25% each, meaning they grow at a rate of more than one percent per year. They grew mainly at the expense of green and other non- built-up urban areas, and to some extent agricultural areas. Forest areas experienced little change. 56 Table 10: Land Cover 1988 and 2008 and Land Cover Changes Change in Change in Area in km² km² % Thematic classes 1988 - 1988 - 1988 2008 2008 2008 Residential areas 81.58 102.04 + 20.46 + 25.1 Non residential urban areas* 29.91 38.07 + 8.16 + 27.3 Transport units 15.17 15.49 + 0.32 + 2.2 Green & other non-built-up 73.16 47.87 - 25.29 - 34.6 Forests 15.98 15.99 + 0.1 + 0.1 Water 4.88 6.37 + 1.49 + 30.6 Agricultural and other open 362.22 356.13 - 6.09 - 1.7 * Comprises all industrial/commercial areas including water supply infrastructure, construction & mineral extraction sites. 113. The development of built-up areas within the areas defined in this study as Urban, Peri- urban and Rural is shown in Figure 22. The highest relative growth in built-up area has occurred in the area defined as Rural, and the lowest in the Urban area. That means the Rural built-up land is developing most dynamically – although from a small initial base – while, in absolute terms, the Urban areas accommodate the largest growth of built-up areas. The current situation is shown in Figure 23: The Urban area has 77%, the Peri-urban area 55%, and the Rural Area 6% built-up land. 57 Figure 22: Development of the Built-up Area, 1988, 1999, and 2008 180 160 +9,6% Development of built-up areas 1988-2008 +10,0% 140 120 180 160 urban 100 +9,6% km² 140 +10,0% peri-urban 80 +12,2% 120 +10,4% rural 60 urban 100 km² peri-urban 40 80 +12,2% +12,4% +10,4% +14,3% rural 20 60 0 40 +12,4% +14,3% 20 1988 1999 2008 0 Figure 23: Total Share of Built-up Areas within Urban, Peri-urban, and Rural Communes in 2008 2008 Urban communes Peri-urban communes Rural communes 6% 23% 45% lt-up 55% 23% 77% 94% Total area built-up Total area non-built-up 58 4.3.2.2 Social Exposure and Vulnerability 114. This section examines the relation between the population growth during the time period 1988/1999/2008 and the areas experiencing hazards of inland flooding (RFP = Relative Flood Potential), Coastal Erosion (CEP), and Coastal inundation Potential (CIP). As highlighted in Figure 24, much of the population growth that happened in this period took place in areas that are exposed to a moderate or high hazard potential – and the Peri-urban zone has the highest percentage of population growth in hazard prone areas. In Dakar’s Peri-urban communes, almost 40% of new population has settled in areas with significant hazard potential from inland flooding, coastal erosion, or sea level rise. This rate is twice as high compared to urban (19%) and rural communes (23%) in the Dakar Metropolitan Area. Figure 24: Population Growth 1988 - 2008 in Areas with Different Hazard Potentials 100,0% 90,0% 80,0% 70,0% 60,0% P n 50,0% P 40,0% h 30,0% 20,0% 10,0% 0,0% Urban Peri-urban Rural Dakar (average) Population growth in areas with low or no hazard potential Population growth in areas with high/moderate hazard potential 115. Still another representation of the social exposure and vulnerability is shown in Figure 25, where the yellow ellipses mark areas with a high population growth 1999 – 2008 and high hazard potentials. 59 Figure 25: Hot Spots of Social Exposure Given by High Population Growth between 1999 and 2008 and High Hazards Potentials 100,0% Hot-spot of population 90,0% increase in high risk zones 80,0% 70,0% 60,0% 50,0% 40,0% 30,0% 4.3.2.3 Economic Exposure and Vulnerability 116. The links between economic exposure and vulnerability are based on information on land price values collected through local consultation. Land prices are expressed in U.S. $ per square km (Figure 26). The figures reveal that the Dakar Metropolitan Area represents a total land value of 44 billion USD, and that 4.9% of this land value, i.e. 2.15 billion USD, is exposed to high hazard potentials. 117. Further, Table 11 shows the land values exposed to high hazard potentials differentiated according to Urban/Peri-urban/Rural zone, and to hazard type: while in the Urban area, the largest exposure is to Coastal Erosion, in the Peri-urban and the Rural Areas the major hazard threatening land values is Inland Flooding. 60 Figure 26: Economic Exposure to Hazards Expressed by Land Price Values (US$ per sq km) Table 11: Land Price Value Exposed to High Hazard Potentials (million US$) The Dakar Hazard type Urban Peri-urban Rural Metropolitan Area (total) Inland Flooding 134 291 751 1,176 Coastal Erosion 501 115 97 257 Coastal Inundation 73 2 181 714 TOTAL 709 408 1,030 2,147 61 4.3.2.4 Built-up Areas exposed to Hazards 118. An example of Spatial Analysis output is shown in Figure 27. The hazard potentials are only overlaid on built-up areas, therefore the pattern of hazard potential present in the other maps is interrupted here. 119. This map type shows what kind of built-up land use classes are exposed to hazards, and the type and degree of hazard potential. In the detailed (upper) part of Figure 24 there is a widespread low to moderate inland flooding hazard potential within the respective area, and that areas under high hazard potential for inland flooding seem to be limited in extent. Likewise, in this area there is a low to moderate coastal erosion potential threatening the southern coastal region. 120. Thus, this type or similar maps could be used as the basis for infrastructure planning such as sewage disposal, or engineering measures to locally decrease hazard potentials. As a sort of economically relevant built-up land exposed to hazards, a further map type was generated within this category: It shows the kind and degrees of hazard potentials, and the amount of Industrial/Commercial/Traffic areas exposed to them (Figure 28). Again, a detail is extracted in the upper part of this figure. The map shows that areas containing large shares of economically relevant land cover classes are prone to at least a low to moderate inland flooding or coastal erosion hazard. 62 Figure 27: Built-up Area in 2008 Threatened by Hazards, with Detail Shown in the Upper Part 63 Figure 28: Share of Industrial/Commercial/Traffic Areas and Exposure to Hazards in 2008 64 65 4.3.2.5 Non-built-up Areas exposed to Hazards 121. The final type of Spatial Analysis was applied to investigate hazard potentials in non- built-up areas. The result for 2008 is shown in Figure 29. Especially in the enlarged upper part of the figure, low elevation areas (depressions) can be recognized, which are obviously more prone to flooding than surrounding areas with larger slopes. This type of map can potentially be used for advising municipal administrations in the question of land use zoning. Areas with high hazard potential, for instance, should not be assigned as building zones. 122. The scope of this study does not allow for detailed area or structural assessments, but it provides the basis for planning more in-depth research, including in-situ work to develop land use zoning maps. Figure 30 shows the percentage of non-built-up areas in the Urban, Peri- urban, and Rural communes that are exposed to high hazard potentials. Figure 29: Non-built-up Areas 2008 Exposed to High Hazard Potentials Urban communes (Open land: 23,78 km² = 23%) 4% Non-built-up area (high RFP) Peri-urban communes Non-built-up area (high CEP) 91% High Risk 9% 3% (Open land: 44,74 km² =45%) Non-built-up area (high CIP) Non-built-up area (no hazard potential) 2% N 10% N 83% High Risk 17% N Rural communes 7% N (Open land: 357,79 km² =94%) 0% 5% Non-built-up area (high RFP) Non-built-up area (high RFP) 7% 5% Non-built-up area Non-built-up area (high CEP) (high CEP) 93% High Risk 7% Non-built-up area (high CIP) Non-built-up area (high CIP) 1% 1% Non-built-up area (no hazard potential) Non-built-up area (no hazard potential) 1% 1% 66 Figure 30: Hazard Potentials in Non-built-up Areas 67 5. Hotspot Characterization and Assessment of Institutional Capacity 5.1 Overview 123. Senegal has been actively pursuing a Disaster Risk Reduction (DRR) strategy at the national and regional levels as part of its Poverty Reduction Strategy (PRS) process (IMF 2007). The strategy, which was developed as part of the Africa Regional Strategy for Disaster Risk Reduction, has initiated several programs at the national level, including the establishment of flood and disaster management committee under the coordination of the Ministry of the Interior. The Ministry of the Interior is also responsible for the plan of the national disaster relief organization (ORSEC). Furthermore, a special flood management plan, also known as the JAXAAY plan, has been established with a view to increasing the effectiveness of the services provided in flood prone areas. However, the implementation framework for disaster risk management (DRM) is ambiguous, complex, and unclear at the local level. 124. For example, in the case of flooding, the local mayor is often responsible for disaster response but has very little influence on mitigation, as those policies are often determined at the national level. In Senegal, six Departments at the national level are in charge of flooding issues: the Ministère de l’Urbanisme, the Ministère de l’Environnement, the Ministère de l’Aménagement du Territoire, the Ministère de l’Intérieur, the Ministère de l’Hydraulique, and the Ministère de la Prévention (Mbow et al., 2008). The coordination mechanism is weak among the ministries, which in turn weakens coordination at the local level further. 125. Even though Senegal has devolved extensive responsibilities to local governments as part of its decentralization process, local governments lack adequate funding for infrastructure investments and service delivery. They also confront skills shortages in general; the lack of technical capacity is even more poignant in relation to dealing with complex issues like climate variability risks. Furthermore, planning instruments like land use planning, which has the ability to influence the urban-rural footprint, remains pretty much under the control of the central government. Ex-ante disaster risk management (DRM) measures like land-use and infrastructure planning that can reduce exposure of natural hazard risks at the local level are essential in a fast- growing city like Dakar. 5.2 The “Primer� Survey (the City Typology and Risk Characterization Matrix) 126. In order to assess DRM institutional capacity at the local level, a specially designed survey was implemented, namely the City Typology and Risk Characterization Matrix, or simply the Primer Questionnaire (World Bank, 2008b). The assessment exercise, developed by the World Bank, identifies the city’s human and built environment characteristics, and potential impacts of natural hazards and climate change risks.35 127. The Primer Questionnaire is designed to give an overview of all important issues and activities that could affect the city, and should be completed by a range of city stakeholders. It is 35 The assessment will also identify (i) local government prerogatives and authorities that would allow them to take action in dealing with potential climate change impacts and natural hazards, and (ii) main vulnerable and at-risk areas. This knowledge provides valuable information in order to define priority actions that move (or “cool down�) the city from becoming a “Hot Spot.� 69 divided into 11 categories of attributes (A through K) in four main areas. We extend the Questionnaire by adding additional modules of (i) Vulnerability assessment for different consequences of climate change, (ii) Preparedness and response to different natural hazards in peri-urban expansion areas, (iii) Institutional framework, and (iv) Peri-urban areas. Table 12: City Typology and Risk Characterization Matrix City description Category A the geographical location of the city the size and main characteristics of the city area and Category B population Category C governance structure and hazard management the responsibilities for disaster risk management and Category D climate change management Category E the financial resource of the city Category F the city’s built environment. Political and economic impacts Category G the political impact of a disaster affecting some cities the impact of disasters on the most relevant urban Category H economic activities of the city Natural hazards Category I the threat of natural hazards the disaster response system and existence of a city’s Category J emergency response plan Climate change impacts Category K climate change impact 128. The survey collected local DRM information from four administrative Departments of the Dakar Metropolitan Area: Dakar, Guédiawaye, Pikine, and Rufisque. It was implemented in December 2008 using the questionnaire in the Annex. The questionnaire is divided into three modules: 1) Definition of hotspots; 2) The institutional framework; and 3) The peri-urban areas. 70 A local consultant, with help of two professionals, conducted the survey. A range of decentralized authorities were interviewed, including Préfets, city authorities (Mayors, Deputy Mayors) technicians from the cities of Dakar, Guédiawaye, Pikine, and Rufisque, urban planners, land use specialists, financial experts, and hotel managers. 5.3 The survey results 5.3.1 General data on the Departments 129. The Dakar Metropolitan Area (the Region of Dakar) stretches only about 580 square kilometres or 0.28% of the total national territory, but is the highest population density area with 4,122 inhabitants per square kilometer. However, its population is unevenly distributed with 22,108 inhabitants per square km in the department of Guédiawaye and 842 inhabitants per square km in Rufisque.36 Given the fact that Rufisque is the least populated department and its rural characteristics, it has the largest land stocks particularly in the rural communities of Yène and Sangalkam. 130. The Dakar Metropolitan Area has a population of 2,167,893 inhabitants distributed as follows: Department of Dakar: 871,038; Guédiawaye: 259,972; Pikine: 767,046; and Rufisque: 269,737 inhabitants.37 According to the same source, the annual population growth rates in 2002- 2006 in the four departments are respectively: 2.96%; 2.9%, 3.9%; and 3.32% for Dakar, Guédiawaye, Pikine, and Rufisque. The department of Guédiawaye is the smallest one (12.9 sq/km and has the highest population density (22,108 inhabitants per square km), followed by Dakar with a population density of 13,366 inhabitants per square km living on 78.5 sq/km, Pikine (9,777 inhabitants per sq/km on 86.8 sq/km land, and Rufisque which is the largest department (371.8 sq/km) with the lowest density (842 inhabitants per sq/km). 131. Squatter settlements, which are defined as housing occupation in areas which have not been plotted out based on urban planning rules, is very common in the cities of developing countries. Statistics are not reliable, as shanty towns, slums, irregular housing, squatter settlements, and traditional villages integrated in the city tend to be confused. Some specialists assume that 30% of the Senegalese urban population, which is around 1,300,000 inhabitants, is living in squatter settlements. The largest concentration of squatter settlements in Senegal is in the Pikine-Guédiawaye area with a population of 600,000 inhabitants (Diagne, 2002), which in our definition of this study can be classified as peri-urban areas. 36 Source : Situation économique et sociale de la région de Dakar-année 2006, ANSD, Service Régional de la Statistique et de la Démographe de Dakar (SRSD), October 2007. 37 Source : Situation économique et sociale de la région de Dakar-année 2006, ANSD, Service Régional de la Statistique et de la Démographe de Dakar (SRSD), October 2007. 71 Table 13: Population in Authorized and Non-authorized Human Settlements in the Dakar Metropolitan Area38 Region % Dakar % Pikine % Rufisque % Housing (regular, 4674.1 62.28 2952 89.05 1482.3 42.95 239.96 32.5 Buildings, Villas) % 100 63.15 31.71 5.14 Squatter 1633.1 21.76 98.32 2.98 1464.2 42.42 70.58 9.57 (Spontaneous) % 100 6.02 89.65 4.33 Village Type 1196.7 15.96 264.36 7.97 504.74 14.63 427.63 57.93 % 100 22.09 42.17 35.74 TOTAL 7504 100 3314.7 100 3451.1 100 738.17 100 Source : enquête ménages CAUS- PDU de Dakar 2025 132. Given the percentage of the floating population39, the department of Dakar has the lowest rate (2.98%). The highest rates are in Guédiawaye, Pikine, and Rufisque. There is data variability for Guédiawaye, because our calculations based on the results of the latest population census carried out in 2002 give 37.5% whereas the environmental profile of Guédiawaye (IAGU, 2005) gives 70%. As far as Rufisque is concerned, the city Mayor gave the figure of 22.95%. For Pikine the rate is 64.2% (Audit Urbain de Dakar, 1999). 5.3.2 Governance Structure Related to Disaster Risk Management Current Disaster Risk Management System 133. Traditionally, the hazard and disaster management is assured through 2 types of mechanisms: The High Commission of Civil Protection (CSPC), which have subdivisions at regional and departmental levels, and the Division of Civil Protection (DPC). 134. The High Commission of Civil Protection (CSPC) is a special consultation body in the domain of civil defence, established by the Government for risk prevention. The Commission is chaired by the Minister of Interior and is represented at administrative level by the regional and auxiliary commissions of civil protection and it includes representatives of Presidence, 38 Enquête ménages CAUS-PDU de Dakar 2025 39 Floating population includes all people living in irregular zones, spontaneous occupations (in zones that have not been parcelled), and traditional villages. Our own calculation (for Guédiawaye) was made as follows: population of the irregular quarters (Medina Gounas and Wakhinane Nimzat) / Total department population × 100. 72 Primature, Parliament, sectoral ministries, local authorities, the private sector, and civil society organizations, etc. (Decree n° 99 – 158 du 22 février 1999). 135. The Division of Civil Protection (DPC) is responsible in time of peace or war for protecting people, facilities, resources, and public and private properties. It heads various departments of civil protection at all levels and can use the Fire Department staff (GNSP). DPC manages the permanent secretariat of the High Commission of Civil Protection (CSPC) and it includes: • Department of Surveys and Civil Defense Operations ; • Administrative and Financial Division; • The Permanent Secretariat of the High Commission of Civil Protection ; and • The Training and Refresher Training Centre for Civil Protection of Dakar. 136. Its mission in terms of prevention consists of: • Developing draft texts on civil protection ; • Organizing in partnership with Regional and Auxiliary Commissions of Civil Protection and Inter-Ministry Technical Committees, prevention visits in public settlements, classified facilities, high buildings, and other buildings at risk; • Ensuring follow-up of reports from prevention visits; • Providing security recommendations for construction projects of public settlements, classified facilities and high buildings. The recommendations are made prior to delivering a construction authorization ; and • Training, informing, and sensitizing on hazard prevention. 137. In terms of disaster management, its mission consists of: • Developing relief plans, for example the ORSEC Plan ; • Providing recommendations on Individual Intervention Plans (P.P.I.) and Internal Operational Plans (P.O.I.); • Developing regional resource files that can be mobilized when the ORSEC Plan is triggered. The Division of Civil Protection heads the ORSEC National Resource Management Committee; • Securing the file of the civil protection back-up units; and • Developing focussed hazard management programmes: rehabilitation and establishment of fire hose systems, lightning rods, etc. 138. The hierarchical structure between these civil protection structures is descending: at the top level we have the High Commission of Civil Protection (CSPC), then the Regional Committee of Civil Protection (CRPC) and the Auxiliary Committee of Civil Protection (CAPC). The Division of Civil Protection (DPC) plays a role of connecting those three organizations because it is coordinating all the means of the Plan ORSEC. 73 139. The functional relations between these civil protection organizations can be resumed as follow: in case of disaster occurring in Rufisque, for example, the Prefect, with regard to the extent of the disaster, initiates the ORSEC Plan at departmental level via the CAPC that (s)he is heading. However, before the starting of the departmental ORSEC Plan, a localising, identification and mobilisation of means must be done by the Prefect. If the disaster is beyond the capacities of the CAPC, or if it spreads to the other departments of the region, the Governor takes the matter in charge, under the instruction of the DPC. This latter DPC plays a role between the CRPC, the CAPC and the CSPC, of which it assures the permanent secretariat. If the disaster spreads too many regions, the DPC asks to the Ministry of Interior, via the CSPC to start the National ORSEC Plan. However, the Prefects and Governors are often surpassed because the means censing and update at departmental and regional levels are still not correctly done. The institutional framework includes also: • The National Flooding Prevention and Control Unit (CNPLI) chaired by the Minister of Interior (Decree n°2004-1153 du 18 août 2004 modifiant le décret n°2003-685 du 13 novembre 2003) ; • The Steering Committee of Risk Prevention and Hazard Management within the framework of the Charter signed in September 2002 by the Minister of Interior and the Chair of the National Employer Council (CNP), which is one of the most representative employer organizations in Senegal; • The Steering Committee of the Memorandum of Understanding on prevention of drowning in authorized beaches signed in March 2004 by the Minister of Interior and the Chair of the Association of Mayors in Senegal; and • The National Platform for disasters risks reduction. 140. The mechanism also includes the National Commission of Sustainable Development (CNDD), the National Meteorological Centre of Yoff, the Centre of Ecological Monitoring (CSE), the Ministry of Environment, Nature Protection, Water Storage Ponds, and Artificial Lakes, and recently the Disasters Risks Reduction Network of members of Parliament. Moreover, following consultations organized with the stakeholders involved, and within the aforementioned commissions, the Government defined the following: • A national action plan on risk prevention aiming at mainstreaming policies and programmes within the framework of a sustainable development strategy ; and • A short and medium term national programme of prevention and protection of vulnerable areas (2004-2007). The programme includes (i) building dykes against water swelling in vulnerable areas, (ii) building channels and pumping stations, (iii) developing of retention ponds, and (iv) transferring population from flooding prone sites to suitable sites. 141. Finally, between July and December 2004, the Government organized a consultation meeting with a group of Senegalese and international experts supported by ILO to define an industrial hazard control system. The system includes the following: • Establishment of a national industrial hazards control unit; • Consolidation of an industrial hazards legal and regulatory framework; 74 • Creation of a National Research Institute on major industrial hazards; • Capacity building on major industrial hazards control; • Definition of an industrial hazards management plan; and • Organization of a promotion campaign for the project with the stakeholders. 142. In the aforementioned multi-sector programmes, steps have been made to encourage prioritization of risk prevention in economic and social development plans, sustainable development programmes, poverty alleviation programmes, and national UN Millennium Development Goals implementation plans. 143. A number of academic and research institutions are working on disasters and climate change. Some of them are: • The Institute of Environment Science (ISE) in the Faculty of Science and Techniques of Cheikh Anta Diop University of Dakar (UCAD); • The Laboratory of Atmospheric Physics in UCAD Faculty of Science and Techniques; • The Centre for Ecological Monitoring (CSE); • The Regional Centre for Improvement of Adaptation to Drought (CERAAS/ISRA); • Adaptation to Climate Change in Africa (ACCA) of the International Development Research Centre (IDRC); • Institute of Earth Sciences (IST) ; • The National Meteorology Division ; • African Urban Management Institute (IAGU) ; • ENDA; • GREEN Senegal (Environment Research and Survey Group) ; and • The Federation of Senegalese NGOs (FONGS). 144. About the territorial administration, the Government is represented at departmental level by the Prefect who is appointed for an undetermined term. The city authority is the Mayor who has a five year term. In general, the municipalities do not have a division in charge of risk management. For Pikine, there is the “Direction des Affaires Sociales, Sanitaires et Educatives� (DASSE), which is responsible for public assistance to the population. DASSE has three equipped ambulances and fifty relief workers as a result of Italian cooperation. For Rufisque, there is a Security Commission within the Mayor’s Office in addition to the technical services in charge of prevention and compliance visits. The Secretariat is sheltered by the Fire Department. 145. In all the cities there is the Environment Department. At regional level there exists the Environment and Classified Settlements Division and the Regional Environment Service. In Guédiawaye and Pikine, there is within the Prefecture a department in charge of environment management headed by the sanitation services. 75 Financial Resources of the Cities of the Dakar Metropolitan Area 146. Revenues (47.4 billion CFA) as well as expenditures (47.3 billion CFA) of the Dakar Metropolitan Area are significant, but they represent only 2% of the regional GDP. The budget of the City of Dakar rose from 2 billion CFA in 1985 to 10 billion CFA in 1996, 24 billion in 2004, 26 billion in 2005, and over 28 billion in 2006, making 50% of the budgets of local authorities in the agglomeration. This shows a clear financial imbalance between the local authorities in the Dakar Metropolitan Area. The distribution of the expenditures confirms the predominance of operational expenditures over investment expenditures with respectively 59% and 41% (Source: CDS, IAGU, 2008). 147. The revenues are 94% recurring revenues whereas an important part of the expenditures (68%) are devoted to functioning, leaving some for investment. However, compared with the regional wealth, the levies at local levels are very low (2%) against 13% for the Government levies. But this situation is more due to the nature of taxes levied in each entity rather than to performance of local authorities. (Source: CDS, IAGU, 2008). 148. The survey shows the total budget of the administrative departments is 36.73 billion for Dakar, 841.21 million for Guédiawaye, 8.85 billion for Pikine, and 6.33 billion for Rufisque as of 2008. The budgets mainly include local taxes and levies (around 90%) for all the departments. Subsidies from central government remain minimal, less than 2%. The funds from internal and international markets are not often specified. Climate Change Management and Hazard Management at Administrative Department Level 149. The hazards related to climate variability and human action are increasing. The concentration of population in large agglomerations, their settlement in risky areas, and the development of economic activities in some sites, result in the population vulnerability to disasters and major hazards. 150. Senegal has experienced several disasters including the explosion of an ammonia tank in 1992 in the SONACOS factory in Dakar, the crumbling of buildings (old buildings and even building under construction, in some cases as a result of non compliance with the construction standards), flooding in some areas (lack of channels, works and structures undersizing against the cities’ demographic evolution). The accidents resulted in heavy human and property losses. The advent of such hazards showed the urgency of having proper knowledge of the nature of those hazards which threaten the populations, their location, and the resources needed. 151. In an effort to face natural disasters and those generated by human action, the Government authorities have adopted since 1993 the ORSEC Plan, which aims at coordinating relief operations in times of disaster. Within the framework of the hazard prevention effort, the Government established in 1993 a High Commission and Regional and Auxiliary Commissions of Civil Protection. Moreover, Commissions for specific disaster management were also established, such as the National Commission of Flooding Prevention and Control. 152. The activities implemented by those Commissions and the identification of risks carried out in October 2002 by the Division of Civil Protection in partnership with administrative authorities helped define the areas where risks are likely to take place at the expense of the populations, the nature of the risks, and the preventive measures to take. It is necessary to visualize in the maps all those data in order to help the authorities have a readable risk database 76 and existing resources to cope with it, take appropriate action, and strengthen intervention and disaster management resources. This study provides practical spatial database and action plans to support the initiatives. 153. As far as disaster risk management at department level is concerned, the responsibilities are clearly defined consistently with the terms of reference of the Division of Civil Protection both at depopulated and decentralized levels. However, in terms of climate change management, given its technical aspect, the Prefect refers to the Regional Department of the Environment and Classified Settlements. For the specific case of Rufisque, the Water and Forestry Department is very active. The authorities in charge of the contracts for the services are: the Prefect, the cities and communes Mayors, and the Chairs of rural communities (specific case of Rufisque). 5.3.3 Urban Planning and Land Use Regulations 154. With regards to the issue of urban planning, there are many planning documents which are partial and hardly integrated in the regional space. In other words, the (technical and regulatory) instruments which could help own urban development are yet to be implemented. 155. Urban Master Plans (PDU) are hardly implemented for proper spatial evolution in the region. The PDU is a reference document that aims to plan and program the development of an agglomeration in short and medium terms (10 to 20 years), taking into account the global objectives of the regional development. It specifies the socio-economic and demographic perspectives of the region and the agglomeration and determines the means and strategies to be implemented in order to reach a harmonious and sustainable development. From 1946 to 2001, there have been four urban Master Plans developed in 1946, 1961, 1967, and 2001. The latest Master Plan was developed in December 2006 (Dakar Horizon 2025), but it has not been validated yet for a lack of implementing order. As a matter of fact, its validation was postponed by the stakeholders until some data are updated. The data are updated but the implementing order remains to be signed. 156. The Urban Mobility Plan for the agglomeration of Dakar (PDUD) Horizon 2025 has been developed to address the critical problems of mobility in the Dakar Metropolitan Area and to reorganize the transport by promoting public transportation in Dakar. The PDUD, which dated back in September 2007 (it can be seen in the Division of Urban Planning in the Departmental Urban Service), is not forcibly taken into account in the construction of the major roadwork items which are often managed by specialized agencies. Guédiawaye as well as Pikine and Rufisque depends on PDUD. However, for Rufisque there is an accessible Subdivision Plan in the City Technical Services and in the Departmental Land and Estate Registry Services. 157. The Master Plan for Urban Planning and Preservation of the Niayes and the Green Zones of Dakar (PDAS) has not played its role yet, in terms of overall urban planning. The PDAS is a specific plan for the Niayes (depressions zones) and Green Zones at the level of the Dakar Metropolitan Area, and has been validated since 2004. 158. The Regional Land-Use Planning framework (SRAT) provides comprehensive guidelines which are not forcibly taken into account in the field. In fact, the SRAT is validated but the implementation does not really respect the prescriptions included in it. Finally, the actions planned in the Regional Integrated Development Plan (PRDI) finalized since 2004 are not implemented yet. 77 159. The survey finds that the percentage of population living in regular areas is 63% in Rufisque, 60% in Dakar, 35.8% in Pikine, and 30% in Guédiawaye. These data show the level of irregular settlements in the administrative departments. Guédiawaye and Pikine are the departments where a large majority of local people reside in irregular areas. The population density in the irregular areas of the department of Dakar is very high (375 inhabitants par square kilometer and around 200 inhabitants per sq/km in Rufisque. In Dakar, the population of the traditional and historic districts is around 100,000 inhabitants representing 11.5% of the total population of the administrative department, whereas for Rufisque, the population is less than 5% of the total department population. 160. With respect to the issue of the building code, Senegal has a Code of Urban Planning (Loi N° 88-05 du 20-06-1998) which regulates the construction standards and is being reviewed in an effort to integrate hazard aspects in building construction. This Code stipulates in Chapter 2, Article 69 that nobody can undertake a construction without authorization no matter what its nature is or carry out modifications in existing buildings in the Cities and in the agglomerations with over 5,000 inhabitants, and the other agglomerations defined by Decree, other agglomerations with important demographic growth, extension, and functions. This obligation is put upon public services and public service contractors, administrative departments and communes as well as private persons. 161. Construction authorization is delivered by the Ministry of Urban Planning and Housing, or by the Governor of the administrative region where the construction will take place, or by the Director in charge of urban planning in the conditions defined in the regulatory part of the Code. There is also the law (Loi N° 78-43 du 06 juillet 1978) which provides guidelines for architecture in Senegal. All the services of the Ministry of Urban Planning and Housing are in charge of their implementation. But at the city level, the DST (Technical Services) is involved. 162. The construction authorizations are instructed by the services of the Ministry of Urban Planning and Housing for the local collectivities. The request (comprising a number of documents) is addressed to the Mayor and the documents are transmitted to the urban planning service for registration and preliminary examination before its introduction in the approval circuit (Domains, Cadastre, Civil Protection, Division of Environment, etc.). Then the authorization ordinance is prepared and transmitted to the Mayor for signature and after signature, the documents are transmitted to the Prefect for approval before going back to the Urban Planning service which delivers it to the applicant. The duration of all this operation can be three months. In Rufisque, the DST has developed construction guidelines for the clay belt. 163. With regards to the level of compliance with the construction standards, in Dakar, Pikine, and Rufisque, there are no figures, because the compliance certificate is not demanded after completing their construction. For Guédiawaye, the percentage is estimated 10%. 164. As far as the vulnerability of the buildings to natural disasters is concerned, no data are available, but our sources have evaluated this aspect for the four administrative departments: for Dakar less than 5% of the buildings are highly vulnerable, for Guédiawaye and Pikine it is over 15%, and for Rufisque it is between 5 and 15%. For historic buildings, the percentage is everywhere less than 5% whereas for new formal constructions, the rate is under 1% for Dakar and Rufisque and over 5% for Guédiawaye and Pikine. However, as the estimates are not based on scientific quantitative analysis such as the spatial analysis of this study, those are only for reference. 78 5.3.4 Exposure of Political and Economic Assets to Disasters 165. Covering only 2.8% of the national territory, the Dakar Metropolitan Area hosts 75% of the country’s economic activity. Dakar’s GDP is worth around 2.724 billion CFA and represents 60% of wealth generated in the national territory. Its sector-based distribution shows the predominance of services with 69% of the GDP, far ahead the secondary and primary sectors. The revenues collected by local authorities on the wealth generated in the agglomeration remains poor despite the increasing growth of the budgets of communes and key local authorities of the Dakar Metropolitan Area. The national Government remains the main player of the economy in spite of a consolidated decentralization at institutional level. 166. We briefly discuss the locations of political and economic stakeholders and the size of economic values exposed to disaster risks.40 Exposure of Political Assets to Disasters 167. As the national capital city, a regional and departmental capital, Dakar shelters the political administration, 80% of Senegalese industries, and the majority of decision makers. Guédiawaye, Pikine, and Rufisque are departmental capitals with many decision makers as a result of its political and historical background. It is very likely that disasters influence political activity in urban as well as in peri-urban areas. Exposure of Economic Assets to Disasters 168. The administrative departments covered by the survey are situated in the Dakar Metropolitan Area which shelters the major national economic activities. The departments of Dakar and Rufisque are at various levels the main economic activity centers at national and regional levels, unlike Guédiawaye and Pikine. 169. The Dakar Metropolitan Area accommodates over 46% of the Senegalese officials, 97% of the trade and transport sectors employees, 96% of the bank employees, 95% of industries and businesses, and 87% of permanent jobs41. Industries in Dakar include high risk industries such as hydrocarbon tanks in the Port, factories like SONACOS, and petrochemical industries. 170. In Rufisque, there is the largest cement plant in West Africa (SOCOCIM Industries), pharmaceutical industries like VALDAFRIQUE, an oil factory (SENARH), a mattress factory (SPI), the electric power plant of Cap des Biches which produces 22.2% of the total production of SENELEC (National Electricity Company) with 167 MW, and another power plant in Kounoune. 171. In Pikine, the industry represents 9,111 jobs, 8.1 % of the total jobs. Industrial activities are located in the industrial zone with SAR (Africa Oil Refinery), a woodwork plant (La Rochette), RHONE POULENC, and the Free Zone (SAFCAC, SENECOR, VENUS, etc.) 172. The sector of services is represented by trade (35 markets the most important of which are the fish market and the stockyard), and transportation is also an important service sector. The sector of finance is developed in the Department of Dakar, but it is not important in other 40 This section is not intended to evaluate economic impacts of disasters systematically by taking into account direct and indirect damages. It is rather a broad description of the mass of economic activities exposed to disaster risks. 41 Plan de développement de la région de Dakar, conseil Régional, 2004 79 administrative departments. In Dakar, the tourism sector is the most important with 22% of the national capacity and 233,643 arrivals in 2006. 5.3.5 Climate Change Preparedness 173. In general, the climate change impacts are unknown by the survey interviewees. As for the sectors vulnerable to climate change, in all the administrative departments covered by the survey, the following have experienced the effects of climate change: • The built-up environment (the coast dwellers); • The cultural and religious heritage (in Rufisque, the Mosque, the Church, and the central market are threatened by the sea); • The local economy (trade, industry); • Electric power production and distribution (Cap des Biches Power Plant in Rufisque); • Access to healthcare; • Land use (resulting in Rufisque in displaced population to new land plotting); • Transportation (water taxi system planned for Rufisque to connect Dakar by the sea can be a solution to the transportation constraints); • Recreation sites; and • Tourism. 174. With regards to Dakar, the evaluation of climate change is based on surveys which were already carried out. As for Guédiawaye and Pikine, there has been no specific survey on climate change. However, the organizational, financial, and urban audits conducted by ADM and other surveys address the issue in general terms. As far as Rufisque is concerned, a perfunctory survey has been carried out with partners from the city of Nancy, France in addition to the alert made by the Chair of the Environment Commission of the Mayor’s office since 2004. But none of the cities has a specific climate change strategy. 5.3.6 Disaster Response System 175. In terms of climate change related national policy, since 1994, Senegal has developed a national climate change strategy completed by a 1997 National Communication which is a report dealing with policies and projects to mitigate the climate change impacts in vulnerable areas, but also aiming at adapting agriculture to climate change. The second reference document is the National Action Plan Against Climate Change developed in a participatory way. 176. The Dakar Metropolitan Area is prone to disasters such as flooding and coastal erosion. There is a system to respond to these phenomena. There is the ORSEC Plan (Relief Organization Plan) for flooding. For coastal erosion, there is urban planning of the Millennium Port from the ledges of Dakar to Gorée Island, but all those interventions are made at national level as a result of Dakar position. In Rufisque, it seems that the system which includes the police, the 80 Gendarmerie, the Mayor’s Office, and the Fire Department is still in stand-by but operating loosely. 177. At departmental level, there is a disaster response system that is a departmental commission coordinated by the Prefect and includes the technical services and the Fire Department which operates the Secretariat. There is also the ORSEC Plan triggered by the High Commission of Civil Protection (CSPC) which is in contact with the Governor who informs the Prefect and the latter contacting the Sous-prefects. Decentralized authorities (city mayors, commune and arrondissement mayors) and services and structures (technical services, the Fire Department, etc.) are involved in disaster management. The Police and the Gendarmerie are also involved in the disaster response system. 178. For the classified settlements (at risk) there are the Internal Operational Plans (POI) supported by the ORSEC Plan, depending on the scope of the disaster. But, altogether this mechanism lacks - suffers of an insufficiency of - resources and a prevention mechanism. The population civic commitment is usually called upon when relief interventions are organized. 179. However, this disaster response mechanism is not often tested and there are hardly any simulations. The legal arrangement has to be improved by identifying and locating resources for simulation exercises. The mechanism is not updated on a regular basis. 5.3.7 Peri-urban Areas 180. The institutions that exist in the urban communes are 19 arrondissement communes in the Department of Dakar, 3 arrondissement communes and 2 communes in the department of Rufisque. In the department of Rufisque, there are 2 rural communities. Rufisque is the only administrative department in the Dakar Metropolitan Area that includes in its local authorities two rural jurisdictions: Sangalkam and Yène. We classify as peri-urban communes five arrondissement communes in Guédiawaye and sixteen in Pikine. 181. In peri-urban communes, there is not a specific, formal and well organized disaster response system. However at departmental level, in addition of the fire department, there are punctual actions of arrondissement communes and also in case of disasters, there is, what we call, community dynamics that is a solidarity dash coming from associations, and individual citizens at the quarter level. Unfortunately, there is no formal interactions between communes and administrative departments on disaster management activities, even though communal authorities, often, use to ask help of some institutions (to have for example motor-pumps in case of flooding). 182. A number of public (or para-public) agencies are responsible for infrastructure investments. The public institutions are essentially the Government and Local Collectivities. The Para-public institutions are the Municipal Development Agency (ADM), Public Works and Employment Agency (AGETIP), Autonomous Agency of Roads Works (AATR Agence Autonome des Travaux Routiers), Programme of Construction and Rehabilitation of State Build Patrimony (PCRPE Programme de Construction et de Réhabilitation du Patrimoine Bâti de l’Etat), National Office of Sewage System in Senegal (ONAS Office National de l’Assainissement du Sénégal), and National Programme of Local Development (PNDL Programme National de Développement Local). 81 5.4 Knowledge and Capacity Gaps for disaster management and climate change impacts 183. The survey helped point out that apart from the interviewed technicians, most of the people interviewed, even though they are not aware of disaster risks which are threatening, can remember serious accidents such as the ammonia explosion in SONACOS factory in 1992, which claimed several casualties and important material damages, although they know nothing about climate change and its negative impacts on their life environment. Apart from their participation in seminars and other meetings dealing with climate change and other initiatives taken within the Rufisque Municipal Council which includes an Environment Commission steered by an expert, and in Pikine where a reflection has been initiated for three years, local authorities are not knowledgeable about the issue of the climate change impacts. 184. Policies addressing the natural hazards and climate change risks exist at the national level. Risk management mechanisms also exist, for instance, in risky industries. However, no mechanism is set up to deal with the impact of climate change, and local authorities are not prepared to face them either. 185. The capacity of institutions to face climate change was found to face limitations, but more in terms of resources and planning capacity than in terms of awareness: • In general, technicians from the ministries involved in the climate change issues and those from organizations such as ADM (Municipal Development Agency), research institutes, universities and NGOs are well aware of the issue. In some institutions, actions are taken but much is yet to be done; and • The issue of resources is often referred to, to explain the fact that almost nothing is done in terms of preparedness to climate change. 186. ADM is aware of the scope of natural disasters, which are a recurrent phenomenon in the country (flooding, coastal erosion, atmospheric pollution as a result of mobility conditions). ADM also developed, in partnership with the Division of Civil Protection, the Fire Department, and other consultants, training modules for capacity building of municipal technicians, territorial administration staff, elected representatives, and other policy makers. For the cities, in partnership with environment associations, ADM considers developing a sharing, exchange, and popularization framework on effective risk prevention. 187. As for SAR (African Oil Refinery Company), an operational risk control mechanism is in place. The risks and disasters are managed by the Control and Method Service under the supervision of the Safety and Environment Sections in compliance with the arrangements of the International Security Evaluation System. At internal level, the Chief on duty is responsible for it. When a problem arises, the Internal Operational Plan (POI) is triggered in relation to the Prefet, the Fire Department, the Gendarmerie, and the populations. This mechanism is being simulated on an annual basis. SAR has equipments for sea leaking with a floating dam which helps absorb the leaked wastes. This equipment is unique in West Africa and is stored with Dakar Port’s Authority (PAD). It is also worth mentioning that SAR is involved in the ORSEC Plan. However, there is no mechanism to mitigate climate change impacts. 188. During the interviews conducted for this study, urban planners and land use specialists have confirmed that, in all the administrative departments, the risk zones are spotted by the 82 technical services, but there is a lack of specific structures in charge of hazards and disaster management and preparedness to facing climate change impacts. Municipal technicians point out the absence of a specific service devoted to hazards and disaster management at the commune level and of a preparedness mechanism to mitigate the climate change impacts. 83 6. Moving Forward: Lessons Learned and Action Plans to Ramp up Natural Hazard and Climate Change Risk Management Practices in Dakar, Senegal 189. The current pilot study focuses on applying spatial and institutional analysis at the regional/metropolitan scale to the identification of hazard risks facing the peri-urban expansion areas of Dakar. As a pilot for a broader work program on hazard risk management in peri-urban expansion areas, this study is helping in determining how the analytical methodologies can be improved, and what aspects of this analysis work best and can be most useful for awareness- raising and decision making. In addressing the specific problems of Dakar, the most appropriate use of this level of analysis is in awareness-raising itself. This pilot study can be and is being effectively used to improve overall awareness of governments at different levels and civil society stakeholders of the magnitude and nature of these exposures and vulnerabilities, to motivate further discussion, analysis and action. Next steps on both fronts – actions for Dakar and improvements and replication of the methodology – are discussed in this final section, against the broader backdrop of the Hyogo framework for action on disaster reduction. 6.1 Guiding principles: Hyogo Framework for Action 190. A recent global initiative to promote a strategic and systematic approach to reducing vulnerabilities and hazard risks is the Hyogo Framework for Action 2005-2015.42 It was adopted at the World Conference on Disaster Reduction in January 2005 in Kobe, Hyogo, Japan. The Conference underscored the need for, and identified ways of, building the resilience of nations and communities to disasters. 191. The Hyogo Framework for Action draws on the conclusions from the review of the 1994 Yokohama Strategy43, and underpins (i) more pro-active approach to informing, motivating and involving people in all aspects of disaster risk reduction in their own local communities, and (ii) the scarcity of resources allocated specifically from development budgets for the realization of risk reduction objectives while noting the significant potential to better exploit existing resources and established practices for more effective disaster risk reduction. 192. In order to achieve strategic goals, the Hyogo Framework for Action has selected five priorities for action. States, regional and international organizations and other actors concerned need to take into consideration the key activities under each of five priorities and implement them to their own circumstances and capacities. 42 The scope of this Framework for Action encompasses disasters caused by hazards of natural origin and related environmental and technological hazards and risks. It thus reflects a holistic and multi-hazard approach to disaster risk management and the relationship, between them which can have a significant impact on social, economic, cultural and environmental systems, as stressed in the Yokohama Strategy (section I, part B, letter I, p. 8). 43 The Yokohama Strategy for a Safer World: Guidelines for Natural Disaster Prevention, Preparedness and Mitigation and its Plan of Action (“Yokohama Strategy�) adopted in 1994, provides a landmark guidance on reducing disaster risks and the impacts of disasters. 85 Ensure that disaster risk reduction is a national and a local priority with a strong institutional basis for implementation. 193. Countries that develop policy, legislative and institutional frameworks for disaster risk reduction and that are able to develop and track progress through specific and measurable indicators have greater capacity to manage risks and to achieve widespread consensus for, engagement in and compliance with disaster risk reduction measures across all sectors of society. Key activities include (i) National institutional and legislative frameworks, (ii) Resources, and (iii) Community participation. Identify, assess and monitor disaster risks and enhance early warning. 194. A starting point for reducing disaster risk and for promoting a culture of disaster resilience lies in the knowledge of the hazards and the physical, social, economic and environmental vulnerabilities to disasters that most societies face, and of the ways in which hazards and vulnerabilities are changing in the short and long term, followed by action taken on the basis of that knowledge. Key activities include (i) National and local risk assessments, (ii) Early warning, (iii) Capacity, and (iv) Regional and emerging risks. Use knowledge, innovation and education to build a culture of safety and resilience at all levels. 195. Disasters can be substantially reduced if people are well informed and motivated towards a culture of disaster prevention and resilience, which in turn requires the collection, compilation and dissemination of relevant knowledge and information on hazards, vulnerabilities and capacities. Key activities include (i) Information management and exchange, (ii) Education and training, (iii) Research, and (iv) Public awareness. Reduce the underlying risk factors. 196. Disaster risks related to changing social, economic, environmental conditions and land use, and the impact of hazards associated with geological events, weather, water, climate variability and climate change, are addressed in sector development planning and programs as well as in post-disaster situations. Key activities include (i) Environmental and natural resource management, (ii) Social and economic development practices, and (iii) Land-use planning and other technical measures. Strengthen disaster preparedness for effective response at all levels. 197. At times of disaster, impacts and losses can be substantially reduced if authorities, individuals and communities in hazard-prone areas are well prepared and ready to act and are equipped with the knowledge and capacities for effective disaster management. 6.2 More pro-active approach to informing, motivating, and involving people in their own local communities 198. Following the Hyogo Framework for Action guidelines, local agencies and local communities should play a pivotal role in disaster management practices. They should be better informed and provided more resources to implement effective disaster management practices. National and regional governments legislate for various land-use regulations and mobilize physical and administrative resources for various ex-ante and ex-post disaster management activities. However, the actual implementation to prevent and mitigate natural hazard and climate 86 change risks including enforcement and supervision should be done by local agencies and local communities. 199. Consistent with these principles, the recommendations emerging from this study include measures that improve awareness of disaster risks, as well as a frank assessment of institutional resources and capacity for disaster management and prevention in the Dakar Metropolitan Area. We identify three avenues for engagement: • First, informing, motivating, and involving people in their own local communities. . • Second, strengthening local institutional capacity and coordination. . • Third, policy reforms and investments for improved hazard resilience and preparedness at the local level. . 6.2.1 Informing, motivating, and involving people in their own local communities. 200. The most immediate recommendation from this pilot study is to develop a general awareness campaign – which has already started with the process of validation and dissemination of the findings of the study. Local agencies and local communities should play a pivotal role in disaster management practices, and develop demand for improved land use planning and disaster response. In turn, empowered local communities and agencies can play a key role in identifying areas and structures under risk, and monitoring and evaluating the implementation of the measures. Alignment with the Hyogo framework requires a consultative and participatory process that ensures sustainability and ownership of the measures proposed. Action plan: General Awareness Campaign 201. (Already initiated in the context of this pilot study) • Organize local knowledge and information dissemination activities, targeting local public agencies and local communities, on the seriousness of the natural hazards and climate change impacts on their own lives, with focus on the behaviors that the population can control and improve on. • Arrange collaboration and joint activities with various local agencies, academic and research institutions, non-profit organizations (NGOs) to pursue these campaigns of sensitization. 6.2.2 Strengthening local institutional capacity and coordination. 202. The study also reveals failures of capacity, accountability and coordination among local agencies and between them and agencies at other administrative levels. These are issues that require further examination and discussion among the relevant stakeholders. This study limits itself to pointing to some of the key institutional capacity and coordination issues identified. It suggests some that may be on the critical path for progress and deserve special attention, such as the identification and empowerment of an institutional champion for disaster risk management and prevention at the metropolitan level, and the development of a local database on hazards and the training of local agency staff to use it effectively. 87 Action plan: Strengthening Local Institutional Capacity and Inter-agency Coordination • Identify viable and well-recognized institutional champion at the Metropolitan Level. • Initiate discussion at the highest political level for local institutional strengthening and coordination and reforms: o Initial focus: (a) development of early warning and quick-response system, paying attention to currently under-served peri-urban areas; and (b) improve local organization and capacity to enforce urban zoning and regulations to reduce vulnerability to natural hazards, with special focus on currently under- served and fast-growing peri-urban areas. o Medium-term focus: (a) adequate resourcing of key local agencies; (b) policy reform; and (c) reallocation of public expenditure and investment to local disaster risk mitigation and prevention. (See further below.) • Develop a spatial database for local disaster management in the Dakar Metropolitan Area, and ensure broad access and hands-on training for local agency staff. • Promote local communities’ engagement and participation in disaster prevention measures. 6.2.3 Policy reforms and investments for improved hazard resilience and preparedness at the local level 203. Beyond the immediate findings and recommendations of this study, stakeholders in Dakar need to consider what substantive policy and investment measures may be considered over time. Ultimately, policy reforms that influence behavior to promote better risk management and investments that strengthen resilience at the local level would be needed to improve the situation of Dakar and its peri-urban expansion areas. The findings of this study suggest the importance to focus on better local land use planning and management, and infrastructure, and the last segment in the table below summarizes possible measures to be considered. This study does not make specific recommendations in these areas, however, as this would require a more detailed analysis. Selection of viable choices would also depend importantly on stakeholder consultation. Action plan: Policy Reform and Investment • Improve local land use planning and management: (a) improvement of land property right assignment and enforcement, with special focus on peri-urban areas; and (b) consultative development of metropolitan development plan, including identification of disaster hotspots and corridors for urban expansion, and potential land acquisition plans to support urban growth corridors. • Strengthening resource base for local authorities, including through proposed betterment taxes that take advantage of improved land management plans and corridor development. 88 • Invest in climate- and disaster-proofing infrastructure and housing stock: retrofitting existing infrastructure and housing in hazard-prone areas; improving infrastructure planning and monitor quality of investments. 6.3 Replication of the Pilot Study 204. As mentioned earlier, this study was intended as a pilot to test new methodologies, and identify how the approach could be enhanced in case of replication in other cities and regions of the developing world. Interest in replicating the approach exists for other African cities, as well as for a selection of Asian cities. Also, this pilot study may also provide the foundation for the development of a city vulnerability index to be applied to a large number of cities. In replications of this study, it is proposed that the following enhancements be considered: • Consideration of a broader range of natural hazards, beyond the three included in this study. • More robust definition of peri-urban areas. • More detailed and better documented analysis of the economic impact of hazards. • More detailed discussion of the methodology for population density imputation, possibly considering different relationships between building density and population density, depending on whether the area is formal or informal. • Addition of information (layering) of major infrastructure (roads, electricity, sanitation). • Utilization of more detailed GIS data sources, such as those captured in cadastres, to inform the more detailed economic and population analysis suggested above. 205. These extensions of the current pilot study may lead to slight increases in the cost of undertaking the study, at least in the initial replications. 89 BIBLIOGRAPHY Birkmann, B. (ed., 2006), Measuring Vulnerability to Natural Hazards: Towards Disaster Resilient Societies, United Nations University Press. Buxton et al. (2006), “Change and continuity in Peri-Urban Australia,� mimeo, RMIT University, Australia. Crichton, D. (1999), The Risk Triangle. In Ingleton, J. (ed.): Natural Disaster Management, London, Tudor Rose. Dennis, K.C., I. Niang-Diop and R.J. Nicholls (1995), “Sea-Level Rise and Senegal: Potential Impacts and Consequences,� Journal of Coastal Research, Special Issue, 14:243-261. Diagne, M. (2002), Restructuration de l’habitat spontané dans les quartiersurbains pauvres : Exemple des outils institutionnels expérimentés auSénégal. Programme d’Appui à la Décentralisation et la GouvernanceLocale en Afrique, Dakar : Institut Africain de Gestion Urbaine (IAGU). Diop, S. (2000), Contexte géologique et risques naturels : Déterminants structuraux de l’instabilité des corniches de la partie sud de Dakar, Mém.Ing. I.S.T., Dakar, n°009/IST/2000. Fall, M., and R. Azzam (1998), “Application de la géologie de l'ingénieur et de SIG à l'étude de la stabilité des versants côtiers, Dakar, Senegal,� Proceedings 8th International IAEG Congress, Vancouver, 21.-25. Sept. 1998, Rotterdam. Fall, M., R.Azzam, and C.Noubactep (2006), “A multi-method approach to study the stability of natural slopes and landslide susceptibility mapping,� Engineering Geology, 82: 241-263. IMF (2007), “Senegal: Poverty Reduction Strategy Paper - Joint Staff Advisory Note,� IMF Country Report No. 07/317. IPCC (2007), IPCC Report on Coastal Systems and Low-Lying Areas, IPCC 4th Report. http://www.global-greenhouse-warming.com/IPCC-4th-Report.html. Khatsu, P., and C. J. v.Westen (2005), “Urban multi-hazard risk analysis using GIS and Remote Sensing: A case study from Kohima Town, Nagaland, India,� ACRS Proceedings. http://www.aars-crs.org/acrs/proceeding/ACRS2005/Papers/URM1-2.pdf. Mbow, C., A. Diop, A. T. Diaw, and C. I. Niang (2008), “Urban sprawl development and flooding at Yeumbeul suburb (Dakar-Senegal),� African Journal of Environmental Science and Technology, Vol. 2 (4), 75-88. Poulter, B., and Halpin, P.N. (2007), “Raster Modelling of Coastal Flooding from Sea-Level Rise,� Int. Journal of Geographical Information Science, Vol. 22, No.2, 167-182. Pryor, R. J. (1968), “Defining the Rural-Urban Fringe,� Social Forces, 47. Rakodi, C. (1998), “Review of the Poverty Relevance of the Peri-Urban Interface Production System Research,� Report for the DFID Natural Resources Systems Research Programme. 91 Thywissen, K., (2006), “Components of Risk, A Comparative Glossary,� SOURCE – Studies of the University: Research, Counsel, Education – Publication Series of UNU-EHS, No. 2/2006. http://www.unisdr.org/eng/library/Literature/9985.pdf UNDP (United Nations Development Programme) Bureau for Crisis Prevention and Recovery (2004), Reducing Disaster Risk: A Challenge for Development. A Global Report. Pelling, M.; Maskrey, A.; Ruiz, P.; Hall, L. (Eds.). John S. Swift Co., USA. UN/ISDR (United Nations International Strategy for Disaster Reduction) (2004), Living with Risk. A Global Review of Disaster Reduction Initiatives, 2004 version, United Nations, Geneva. Webster, D. (2002), “On the Edge: Shaping the Future of Peri-Urban East-Asia,� Discussion Paper, The Urban Dynamics of East Asia Project, Stanford: Aisa/Pacific Research Center. World Bank (2007), “Sustainable Development in East Asia’s Urban Fringe,� mimeo, Washington, DC. World Bank (2008a), World Development Report: Reshaping Economic Geography, World Bank, Washington, DC. World Bank (2008b), Climate Resilient Cities: A Primer on Reducing Vulnerabilities to Climate Change Impacts and Strengthening Disaster Risk Management in East Asia Cities, World Bank, Washington, DC. 92