Living
on the
water’s
edge
Flood risk and resilience
of coastal cities
in Sub-Saharan Africa
Suggested citation: The World Bank (2022) Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa
Credits: The World Bank Group
Cover Image Credits: The World Bank Group
This document is the property of the World Bank.
It is permissible to copy and use any of the material in this report provided that the source is appropriately acknowledged. Further
information is available from: © The World Bank 2022
Adaptations—If you create an adaptation of this work, please add the following disclaimer along with the attribution: This is an adaptation
of an original work by The World Bank. Views and opinions expressed in the adaptation are the sole responsibility of the author or authors of
the adaptation and are not endorsed by The World Bank.
Third-party content—The World Bank does not necessarily own each component of the content contained within the work. The World
Bank therefore does not warrant that the use of any thirdparty-owned individual component or part contained in the work will not infringe on
the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to re-use a component of
the work, it is your responsibility to determine whether permission is needed for that re-use and to obtain permission from the copyright
owner. Examples of components can include, but are not limited to, tables, figures, or images.
This work is a product of the staff of The World Bank and the Global Facility for Disaster Reduction and Recovery (GFDRR) with
external contributions. The findings, analysis and conclusions expressed in this document do not necessarily reflect the views of any
individual partner organization of The World Bank, its Board of Directors, or the governments they represent.
Although the World Bank and GFDRR make reasonable efforts to ensure all the information presented in this document is correct, its
accuracy and integrity cannot be guaranteed. Use of any data or information from this document is at the user’s own risk and under no
circumstances shall the World Bank, GFDRR or any of its partners be liable for any loss, damage, liability or expense incurred or suffered
which is claimed to result from reliance on the data contained in this document. The boundaries, colors, denomination, and other information
shown in any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the
endorsement or acceptance of such boundaries.
RIGHTS AND PERMISSIONS: The material in this work is subject to copyright. Because The World Bank encourages dissemination of its
knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given.
Cover: Aaerial view of Zanzibar, Tanzania. Photo: Gideon Ikigai
Living on the
water’s edge
Flood risk and resilience of coastal cities
in Sub-Saharan Africa
ii |
Contents
Abbreviations................................................................................................................................................................................................v
Acknowledgements.................................................................................................................................................................................. vii
Executive Summary........................................................................................................................................................................ viii
1. Introduction.....................................................................................................................................................................................1
A regional outlook of the coastal cities in Sub-Saharan Africa...............................................................................................1
The rapid growth of coastal cities of Sub-Saharan Africa...................................................................................................... 2
Coastal impacts are increasing....................................................................................................................................................... 3
Supporting coastal resilience in Sub-Saharan Africa............................................................................................................... 4
Understanding risk in coastal cities in Africa to boost resilience.......................................................................................... 5
2. Methods and Approach........................................................................................................................................................... 7
Exposure data...................................................................................................................................................................................... 7
Flood hazards.......................................................................................................................................................................................8
Sea Level Rise....................................................................................................................................................................................... 9
Calculation of risk: population affected and economic impacts............................................................................................ 9
3. Expansion of coastal cities in Sub-Saharan Africa............................................................................................. 11
Population growth.............................................................................................................................................................................. 11
Urban expansion in coastal areas................................................................................................................................................. 13
4. Flood risk in coastal cities is increasing.....................................................................................................................17
Increase in exposure to flood hazards......................................................................................................................................... 18
Economic Damages..........................................................................................................................................................................24
5. Effects of climate change and sea level rise.......................................................................................................... 30
A comparison of historic urban growth with future sea level rise....................................................................................... 31
6. From Risks to Resilience: Building resilient coastal cities in Sub-Saharan Africa.......................32
Leveling the field for understanding and managing coastal risk.........................................................................................32
Reducing and managing flood risk...............................................................................................................................................32
Compound risks, challenges and opportunities for resilience pathways ..........................................................................33
A climate action plan for Sub Saharan Africa...........................................................................................................................33
The World Bank is committed to help build resilience and sustainable growth..............................................................34
| iii
7. Future improvements to regional coastal resilience analyses...................................................................37
Appendix A. Extended results............................................................................................................................................................. 38
Exposure by countries by flood hazard type............................................................................................................................. 38
Flood damages in cities.................................................................................................................................................................. 40
Appendix B. Methods and Data........................................................................................................................................................... 41
Coastal flood hazard........................................................................................................................................................................ 41
River and rainwater flood hazard................................................................................................................................................. 41
Effects of climate change on flood hazards...............................................................................................................................42
Exposure data....................................................................................................................................................................................42
Damage vulnerability curves..........................................................................................................................................................43
Uncertainty in the exposure data analysis................................................................................................................................44
Appendix C. Sub-Saharan Africa’s cities included in the Atlas of Urban Expansion......................................................... 47
Comparison of built-up area change with the DLR WSF-Evo............................................................................................... 47
City population from the Atlas of Urban Expansion................................................................................................................49
Appendix D. Normalized values of Annual Expected Damages.................................................................................................55
List of Boxes
Box 2-1. Data availability...................................................................................................................................................................8
Box 4-1. Flood and storm impacts in coastal cities..................................................................................................................17
Box 4-2. Local risk analysis.............................................................................................................................................................26
Box 6-1. Investments in a resilient coast.....................................................................................................................................35
List of Figures
Figure ES-1. Growing exposure to flood risks in Sub-Saharan Africa.......................................................................................... ix
Figure ES-2. Projected dimensions of flood risk in Sub-Saharan Africa..................................................................................... ix
Figure 1-1. Coastal population projections in Africa....................................................................................................................... 3
Figure 2-1. Approach for assessing coastal risk in cities of Sub-Saharan Africa................................................................... 7
Figure 2-2. Infographic of flood risk calculation............................................................................................................................. 10
Figure 3-1. Population and built-up area growth in all the coastal cities of Sub-Saharan Africa..................................... 11
Figure 3-2. Population within one kilometer from the shoreline and change over time...................................................... 12
Figure 3-3. Urban growth in Lagos, Nigeria, the largest coastal city in the region.............................................................. 14
Figure 3-4. Urban growth in Accra, Ghana....................................................................................................................................... 14
Figure 3-5. Urban growth in Luanda, Angola................................................................................................................................... 15
Figure 3-6. Cities with over 1 million people over time.................................................................................................................. 15
Figure 3-7. Countries by built-up area (BU) and population in 2015........................................................................................ 16
Figure 4.1. Population and built-up areas in the 10-year flood hazard zone for coastal and rainwater sources......... 19
iv |
Figure 4.2. Flood map for Lagos, Nigeria..........................................................................................................................................22
Figure 4.3. Flood maps in Luanda, Angola.......................................................................................................................................23
Figure 4.4. Annual average damage from flood hazards.............................................................................................................24
Figure B-1. Transition from imagery to built-up areas extraction (GHS-BU),
population modeling (GHS-POP), and settlements classification (GHS-SMOD)..............................................42
Figure B-2. Normalized damage factor for Africa: Residential buildings and content.........................................................44
Figure C-1. Spatial comparison of urban footprint between WSF (left panels) and GHSL (right panels)...................... 48
Figure C-2. Urban extent and urban growth (per year) for the 18 cities from Sub-Saharan Africa................................ 50
Figure C-3. Comparison of the main cities in Africa with the other global cities in the Atlas of Urban Expansion...... 51
Figure C-4. Example of urban expansion in Accra, Ghana............................................................................................................52
Figure C-5 Example of urban change in Accra, Ghana.................................................................................................................53
List of Maps
Map 4-1. Annual average damage (AAD) for coastal cities in SSA. ......................................................................................25
Map 5-1. Increase in coastal flood risk (10-year return period) by sea level rise by the mid-century.
Note that the values only correspond to coastal flooding ................................................................................... 30
Map 5-2. Percent increase in exposure to flooding for coastal cities in SSA...................................................................... 31
Map C-1. The Atlas of Urban Expansion collects and analyzes data on the quantity and quality
of urban expansion in a stratified global sample of 200 cities. ...........................................................................49
List of Tables
Table 3-1. Coastal population and built-up area by country over time. ............................................................................... 12
Table 3-2. Top 10 cities by population in coastal cities in 2015............................................................................................... 13
Table 3-3. Top 10 cities by built-up area in 2015.......................................................................................................................... 13
Table 4-1. Top 10 cities by population exposed to flooding (10-year flood hazard zone). ............................................... 20
Table 4-2 Top 10 cities by built-up area exposed to flooding (10-year flood hazard zone)............................................. 20
Table 4-3. Critical infrastructure exposed to flooding (10-year flood hazard zone)
in the 20 most populated cities. ................................................................................................................................... 21
Table A-1. Population in the 10-yr flood hazard zone in each country, by type of flood hazard.................................... 38
Table A-2. Built-up area exposed to the 10-yr flood hazard in each country, by type of flood hazard......................... 39
Table A-3. Top 40 cities by Annual Average Damage from flood hazards............................................................................ 40
Table C-1. Changes in built-up area for the World Settlement Footprint Evolution (WSF)
and the Global Human Settlement Layer (GHSL)..................................................................................................... 47
Table D-1. Average annual flood damage for the top 40 cities in SSA. .................................................................................67
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | v
Abbreviations
AAD Annual Average Damage
CGSW Global Surface Water dataset
CityCORE City Coastal Resilience Africa
DEM Digital Elevation Model
DLR German Aerospace Center
DG REGIO Directorate General for Regional Development
GDP Gross Domestic Product
GHSL Global Human Settlement Layer
GSTR Global Tides and Surge Reanalysis
JRC Joint Research Centre
LECZ Low Elevation Coastal Zone
METEOR Modeling Exposure Through Earth Observation Routines
RCP Representative Concentration Pathway
SERRP Saint Louis Emergency Response and Resilience Project
SSA Sub-Saharan Africa
UNESCO United Nations Educational, Scientific and Cultural Organization
WACA West Africa Coastal Area Management Program
WSF-Evo World Settlement Footprint Evolution
vi | Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa
Cape Coast, Ghana. Photo: © Dallaskoby | Dreamstime.com
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | vii
Acknowledgements
Living on the Waters’ Edge was prepared by a core team This work was possible through the support of the ACPEU
composed by Borja G. Reguero, Fabio Cian, Enock Seth Natural Disaster Risk Reduction Program of the African,
Nyamador, Kristina Wienhoefer, and Lorenzo Carrera. Caribbean and Pacific Group, funded by the European
Union and the Multi Donor Trust Fund, both managed by
The team thanks the following individuals for their valu-
the Global Facility for Disaster Reduction and Recovery
able contributions, revisions, and advice: Ana Campos
(GFDRR) at the World Bank.
Garcia, Brenden Jongman, Joaquin Ignacio Munoz Diaz,
Mathijs Van Ledden, and Oscar Anil Ishizawa Escudero. The document was prepared by the World Bank’s Urban,
Rural, Land and Resilience (URL) Global Practice, under
The report was skillfully edited by Chitra Arcot, Miki Fer-
the guidance of Meskerem Brhane (Practice Manager,
nandez, and Erika Vargas. Data used for the analysis was
URL), Sylvie Debomy (Practice Manager, URL), Peter Ellis
kindly provided by Climate Central, Fathom Global, the
(Practice Manager, URL), and Niels Holm-Nielsen (Prac-
Joint Research Center from the European Commission,
tice Manager, GFDRR).
and Open Street maps.
viii | Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa
Executive Summary
S
ince the 1970s, Sub-Saharan Africa (SSA) has on a city basis. The study combines datasets, models
experienced vast urban development along and remote sensing to provide the first comprehensive
its coastal areas. Today, more than 86.3 regional overview of the rising risks in coastal cities of
million people live in 153 coastal cities with a SSA. The main findings of this report include:
population in excess of 50,000. Such urban development
● Historic information on population and urban trends
has occurred partly in flood-prone areas and has led to
combined with flood modeling at every coastal city
increased flood risk. Urban expansion and concentration
in SSA demonstrates that flood risk has increased
of people and assets in flood-prone areas has been one
significantly in recent decades.
of the main drivers of flood risk in SSA’s coastal cities,
which indicates the importance of taking proactive ● Coastal cities of SSA have experienced substantial
flood risk management, adaptation, and resilience growth in population and urban built-up area in
planning. Furthermore, climate change is already the last decades. The urban footprint of coastal
impacting the region and jeopardizes the achievement cities expanded by 58 percent from 1975 to 2015,
of SSA’s core development priorities. Sea level rise and whereas population increased by 269 percent, from
other effects of climate change are additional threats 23.4 million to more than 86 million. As a result of
to the growing coastal risks in SSA. A compelling body the demographic concentration in coastal cities,
of evidence cautions that the magnitude and scale population density has sharply grown.
of climate impacts on the region’s economies and on
● Demographic concentration in coastal cities has
the poorest populations can roll back hard-earned
driven an intense expansion of urban areas in coastal
development gains and have serious intergenerational
zones. Only five coastal cities in SSA in 1975 had
consequences. To address these present and future
populations in excess of one million people; however,
challenges, Sub-Saharan Africa must ramp up climate
by 2015, twenty cities exceeded such threshold, seven
adaptation and resilience as a cornerstone of poverty
of them had reached more than one million residents
eradication. Yet, the nature of climate impacts on low-
just between 2000 and 2015.
income economies needs to be better understood and
addressed proactively across time scales, as stated in ● Urban expansion, to a large extent, has occurred in
the World Bank’s “Climate Change Action Plan” of 2021. flood-prone areas. As a result, exposure to floods has
also been growing as cities became larger and more
Because understanding risk and its dynamic relationship
populated. In 2015, the population exposed to the 10-
with urban change is a first step to manage risks and
year flood was over 11 million, three times more the
build resilience, this report provides new understanding
people exposed in 1975. However, the built-up area
of how flood risk has changes in the coastal cities of SSA
exposed to flooding increased almost five times over
over time. By combining historic information on urban
the same period, as new development in cities took
development and population trends with coastal and
place in flood prone areas (figure ES-1).
rainwater flood hazards, the study provides new insight
into coastal and flood risk and their historic changes
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | ix
● Across SSA, almost US$200 million per year of ● Flood risk has increased significantly in SSA coastal
infrastructure are estimated to be at risk from cities in the past driven by natural hazards and fast-
flooding (coastal and rainwater). Angola, Ghana, growing urbanization. Population and infrastructure
Nigeria, and South Africa present the highest risk. concentrated in coastal cities is expected to continue
growing into the future. Responding to the present-
● Amongst the cities identified at highest risk are: Accra
day risk and the additional threats of climate change
in Ghana, Cape Town and Durban in South Africa, Lagos
requires proactive and urgent measures, targeted
in Nigeria, and Luanda in Angola, which exceed US$10
resilience planning and adaptation actions. With most
million per year at risk.
infrastructure in SSA yet to build, avoiding future arms
● Coastal and rainwater floods do not affect all cities is imperative for countries to build a more resilient and
equally and vary strongly across coastlines in SSA. Accra, sustainable future.
Cape Town, and Lagos present the highest flood risk from
The intensification and scale of climate impacts will
rainfall flooding; whereas Cotonou, Lagos, and Luanda
challenge the ability of many countries in SSA to reach
are the three cities at highest risk from coastal flooding.
their economic growth and sustainable development
● In 28 coastal cities, the people exposed in the 10-year ambitions. If they remain unaddressed, flood risk
flood will increase by at least a 10% (figure ES-1). The and climate change will continue to deepen existing
effect of sea level rise will be most severe in Mkpanak vulnerabilities and low capacities, leading to poverty,
and Okrika in Nigeria, and Tombua in Angola. fragility, conflict, and violence. This requires urgent action
Therefore, the cities with high exposure to coastal and adequate planning at the city level. Understanding
flooding should consider carefully the rising sea levels how risk has evolved in the past may help inform effective
in future urban plans. interventions in the future.
● Historic urban development in flood-prone areas has
been the main driver of risk in the past.
Growing exposure to flood risks
Figure ES-1 Projected dimensions of flood risk
Figure ES-2
in Sub-Saharan Africa. in Sub-Saharan Africa.
17.7 million people less than
153 Coastal Cities
1km from the sea
(with +50,000 people)
153 coastal cities with
+50,000 people
In eight countries, sea level rise will
increase the number of people exposed
in the 10-year flood zone by more Coastal Urban Annual
than 10% by the mid-century, but the Population built-up area Flood risk
increase in economic risk will be larger.
86.4 +5,520 $204
million sq. km million/yr
Population density increased from
6.7 to 15.7 persons/sq. km between
1975 and 2015, and has doubled x3.6 in x 1.6 in Rainwater = $141 m/yr
since 1990 40 years 40 years Coastal = $56 m/yr
Flooded Jangwani, Dar es Salaam, Tanzania. Photo: Moiz Husein
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | 1
Introduction
Coastal cities and the need to adapt and generate 56 percent of its GDP. Coastal erosion
also poses enormous risks. Environmental degradation
At the beginning of the 20th century, only 14 percent
on the coasts of Benin, Côte d’Ivoire, Senegal, and Togo,
of the world’s population lived in cities. Today, more
as key examples, cost US$3.8 billion, or 5.3 percent of
than half of the global population lives in urban areas
the four countries’ GDP in 2017. Beyond the economic
and generates 80 percent of global gross domestic
cost, floods and air and water pollution caused 13,000
product (GDP). This trend is expected to continue in
deaths a year. In Nigeria, the cost of coastal degradation
the future, and by 2050, cities are projected to host
amounted to US$9.7 billion or 8.1 percent of the GDP in
nearly 70 percent of the global population. Cities are
the states of Cross River, Delta, and Lagos in 2018.1
increasingly becoming world’s centers of innovation,
political and economic activity. Yet this concentration Floods are the most frequent and widespread natural
comes with significant challenges in the face of a disasters in Africa (Niang et al. 2014). Winds, storm
changing climate. Poorer populations are the most surges, large waves, and tsunamis threaten coastal
vulnerable in the global urban expansion as they often cities and communities and represent increasing
inhabit more hazard-prone places and lack the means challenges and costs. Storms impact national economic
to recover from economic or environmental shocks and productivity; threaten water and food security; increase
stresses. These challenges will only continue to grow diseases; and damage critical public infrastructure,
unless urgent action is taken. This trend is ever more basic services and value chains (Munich Re 2013). Many
intense in the coastal zones because of the economic governments and public utilities are overexposed and
benefits that accrue from access to ocean navigation, undermanage these risks (Munich Re 2013; The Geneva
fishing, tourism, and recreation. Coastal zones are Association 2014). When a disaster strikes, vulnerable
also some of the most vulnerable areas globally to the communities and individuals turn to governments and
impacts of climate change, unsustainable development, international agencies for recovery assistance. The
and environmental degradation. rising costs of storms manifest the urgent need to adapt
and manage these risks more effectively (Hallegatte et
Sub-Saharan Africa (SSA), home to 1 billion and more
al. 2013, 2019; Hallegatte 2019).
people, has contributed the least to global warming but,
without urgent action at scale, it will suffer some of the Proactive management of risks grant economies the
worst consequences of a changing climate including resilience to bounce back after disasters and shocks.
cyclones, flooding, and severe droughts—all of which Investments in risk reduction and adaptation can save
threaten economic growth. Significant climate impacts lives, costs, and continuous rebuilding many times over.
on economic growth projected for SSA indicate that GDP However, targeted response and planning need to be
could be reduced by 10 to 15 percent by 2050 (Kompas et informed by a characterization of hazards and risks
al. 2018). Coastal cities are particularly at risk given the that allows identifying present and future risks, their
concentration of economic activity and infrastructure drivers, and the potential solutions (GFDRR 2014). This
in low-lying areas. For example, West Africa’s coastal information is often lacking, especially in smaller cities,
areas host about one-third of the region’s population and is not yet available for all cities in SSA. Furthermore,
2 | Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa
coastal areas in SSA present additional challenges in has not kept pace with the rapid pace of urban growth.
comparison with other regions including: rapid urban Population in cities is growing at 4.1 percent per year,
growth, high level of informality and weak enforcement about twice the global average rate of 2.0 percent, and
of urban planning tools, including zoning and absence the tally in 2020 stands at 472 million people. At this
of localized information on coastal flooding, and the rate of growth, the urban population in SSA is projected
effects of rising sea levels. to double over the next 25 years as more migrants move
to cities from the countryside. However, many cities are
Climate change will especially impact coastal zones
expanding into areas at high risk from natural hazards
(Wong et al. 2014). Global sea level rise is expected to
while infrastructure planning and investments are
rise between 0.5 to 1 meter by the end of the century
critically lagging. Furthermore, unplanned urban growth
(IPCC, 2021- AR6). Even if global temperatures are kept
critically contributes to exacerbate flooding impacts,
below 2 degree Celsius warming, by 2050 at least 570
and as runoff increases in paved and devegetated areas,
cities and some 800 million people will be exposed to
drainage systems become insufficient, and previous
rising seas and storm surges globally.2 Coastal cities
natural water retention areas—wetlands, swamps
are already suffering the increasing impacts from
and low-lying ground—are built up and occupied by
hazards such as coastal and urban floods, erosion, land
buildings.
subsidence, and salt intrusion, which will drive greater
costs in the future. Critical infrastructure such as roads, Without adequate urban planning, many cities have
railways, ports, and other assets will also be affected. sprawled into river deltas, low elevation coastal zones,
A sizeable number of coastal cities have yet to prepare wetlands, or hill slopes—areas highly prone to floods,
adequately for the rising risks, as it is estimated that landslides, and cyclones. A global analysis of urban
approximately 90 percent of all coastal areas globally vulnerability to future flooding determined that of the
will be affected to varying degrees (World Economic top 25 cities by population vulnerability, ten are in the
Forum 2019). The need of coastal cities to adapt is SSA region, eight of which are in coastal West Africa
urgent. alone (Dasgupta et al. 2009). Furthermore, at least half
of the top ten cities were in only two countries of SSA:
In many coastal cities of SSA, flooding and erosion
Cote d’Ivoire with two cities and Mozambique with three.
impacts are already felt and impacting livelihoods
The risks are particularly severe in poor neighborhoods
and coast-dependent economies. The widespread
and slums, where infrastructure is often nonexistent
floods of 2020 in Africa demonstrated once again
or poorly designed and ill-maintained. Exposure is not
that unless hazard risks are managed, communities
balanced across countries.
will stay vulnerable and less resilient to new shocks.
Coastal hazards, combined with unsustainable fast- Demographic and coastal development critically
growing urbanization and increased vulnerability will interact with climate hazards and climate change.
have catastrophic consequences on the poorest urban This convergence of risk factors creates a unique set
communities. With impacts and costs on the rise, of development challenges. Coastal areas in Africa, like
people in coastal areas might be forced to relocate, elsewhere in the world, are increasingly concentrating
driving a pervasive loop (World Bank 2014). To mitigate more population because of the economic opportunities
such consequences, it is crucial to mainstream climate they offer. For example, in Nigeria’s low elevation
adaptation and disaster resilience as a top priority in coastal zones (LECZs)—areas located 10 meters or less
the development of African coastal areas. above mean sea level—the population density is 491
inhabitants per square kilometer compared with 134
inhabitants per square kilometer nationally. By a global
The rapid growth of coastal cities of Sub-
estimate, Africa will be the continent to experience
Saharan Africa, key to understanding risk
the highest rates of population density growth and
SSA is the fastest urbanizing region on the planet. urbanization in LECZs (Neumann et al. 2015). Coastal
Africa’s cities present similar population density of population of SSA, all of Africa except northern Africa,
other cities around the globe, but its economic growth represented 45 percent of the African nations’ LECZ
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | 3
population in 2000 and is expected to grow from the 24 to name a few. These factors make the urban poor more
million in 2000 to 66 million by 2030, and to 174 million vulnerable to the impacts of floods, erosion, extreme
by 2060, at an average growth rate of up to 3.3 percent. weather events, and climate change.
These rates are considerably higher than in Asia, where
annual rates of growth are expected to reach 1.4 percent
in the first three decades of 2000 to 2030, but are
Coastal impacts are increasing
estimated to drop to 1.2 percent after. Such growth will Rapidly growing urban agglomerations in low-lying
be more intense in western Africa (figure 1-1) with some coastal areas are increasingly exposed to flooding and
individual countries experiencing even more extreme erosion. In West Africa, for example, cities like Accra,
changes. For example, in Senegal, the share of the LECZ Dakar, Freetown, Lagos and Monrovia are already
population is projected to grow at a 50 percent rate by suffering the negative effects of climate hazards, which
2060, up from 20 percent in the early 2000s. will be exacerbated by sea level rise and other climate
Poverty and high levels of urban informality also trigger change effects, including changes in wave action (Morim
new low-income dwellers in areas at risk. Low-income et al 2021). In Lagos, which hosts 12 million people in its
groups often settle in highly risk-prone areas, such as main urban center and is the economic center of Nigeria,
flood plains and coastal low-lying zones, and in buildings sea level rise and coastal erosion have already led to a
that cannot resist hazard shocks. A high fraction of decline in water quality, the destruction of drainage
the population in African cities still live in informal infrastructure, and an increase in incidences of water-
settlements and hazard-vulnerable houses—Accra 38 and vector-borne disease. Coastal erosion has also
percent, Abidjan 56 percent, Dakar 39 percent, Dar es hurt indigenous communities that depend on coastal
Salaam 51 percent, Lomé 51 percent, Mogadishu 74 resources for survival. Flooding from storm surges
percent, Mombasa 56 percent, or Monrovia 66 percent has forced the relocation of resorts, businesses, and
Figure 1-1 Coastal population projections in Africa.
140
120
Population LECZ (million)
100
80
60
40
20
0
Southern Middle Eastern Western Northern
Africa Africa Africa Africa Africa
n Baseline 2000 n Scenario C 2030 n Scenario C 2060
Source: Neumann et al. 2015.
4 | Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa
artisans from some popular tourist destinations, like including their impacts (IPCC 2014).
the Victoria Island. In Dakar, Senegal, coastal erosion
Subsiding land is also a growing problem for coastal
not only threatens the city’s beaches, which have an
cities as its effects compound with those from sea
important touristic value, but are also displacing 12
level rise (Cian et al. 2019). Although Africa might not
of the city’s 19 communes by the Atlantic sea front.
experience similar levels of subsidence than megacities
Coastal erosion is also affecting infrastructure, the
in Asia, increasing land urbanization and water demand
harbor, and other industrial plants along the coast,
leads to groundwater overexploitation, which can
such as the Cap des Biches power plant, which is
rapidly evolve into subsiding land and increased flood
located one meter below sea level and has suffered
risk. For example, in Banjul, Gambia, the low-lying city
recurrent floods caused by strong seasonal swells.
is subsiding while sea levels continue to rise. These
Climate change presents an added challenge to the
two factors increasingly stress and compromise the
city, where 60 percent of beaches could disappear by
city flood defenses. Land subsidence has also been
2080 (World Bank 2013a).3 The city of Saint Louis, the
documented in Lagos and other urban agglomerations in
second largest city in Senegal, also suffers from coastal
the Niger Delta5 (Adeyinka et al. 2005), where individual
flooding, severe, decade-long erosion, river flooding, and households are increasingly using private boreholes to
saltwater intrusion (box 4-1). These natural calamities compensate public utilities’ deficiencies in water supply.
are severely impacting the livelihood of many. Cities of
eastern Africa are also at risk. In Tanzania, for example,
sea level rise could submerge large areas of land and Supporting coastal resilience
affect approximately 1.6 million people (Kebede et al. in Sub-Saharan Africa
2010). Without adaptation measures, climate change The World Bank is committed to help respond to the
can displace thousands of people, which might be challenges of improving city resilience in Africa across
forced to relocate to other urban areas, already under the board. The Plan of Action for Implementation of the
significant pressure (World Bank 2013b). Sendai Framework for Disaster Risk Reduction 2015–
Cyclones and tropical storms affect mostly the 2030 in Africa calls for urban resilient development
southeastern coast, primarily Madagascar, Mozambique, and climate change adaptation. The New Generation
and the Indian Ocean islands (World Bank 2008). Low- Africa Climate Business Plan (NG-ACBP) is a blueprint
lying countries along the African coast of the Indian for climate action to help address the key climate
Ocean are also susceptible to tsunamis (UNISDRR 2009). vulnerabilities and SSA’s core development priorities:
Storm events regularly cause severe damages and food, water, energy, and human and environmental
losses, particularly during the southwest Indian Ocean security. The NG-ACBP underscores the importance
cyclone season. In eastern and western Madagascar, of pursuing climate-smart urban transitions including
green mobility, while supporting the region’s ability to
the cyclone seasons can cause losses and damages
protect against climate shocks and pandemics. Other
to individual households, affecting 10–30 percent of
initiatives, such as the West Africa Coastal Areas
their average annual GDP per capita. For example, the
Management Program (WACA) P152518, are also
three cyclones that struck Madagascar in 2008 caused
contributing to sustainably manage coastal areas by
damages equivalent to 4 percent of their national GDP
accessing expertise and finance. To date, WACA has
and affected 342,000 people. More recently, Cyclones
provided technical assistance and national multisector
Idai and Kenneth in 2019 were two of the five worst
investment plans in Benin, Cote d’Ivoire, Ghana,
storms to ever hit Mozambique. Catastrophic flooding
Mauritania, São Tome and Principe, and Togo.
from the storms affected close to 2.2 million people
in Malawi, Mozambique, and Zimbabwe.4 They caused The World Bank Group also supports many other local
an unprecedented amount of damage. It is expected resilience initiatives, some examples include, among
that along the eastern coast of Africa, climate change others: (i) the Cities and Climate Change Project
will further exacerbate existing climate variability and P123201 and the Climate Technical Assistance P131195
increase the frequency and magnitude of extreme events, in Mozambique; (ii) the Dar es Salaam Urban Resilience
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | 5
Investment Planning in Tanzania P160397; (iii) the Risk Reduction, the Paris Agreement on Climate Change,
Coastal Region Water Security and Climate Resilience and the New Urban Agenda.
Program P145559 in Kenya; (iv) the Gambia Integrated
Urban and Coastal Resilience Program P172822; and
Understanding risk in coastal cities
the Saint Louis Emergency Recovery and Resilience
in Africa to boost resilience
Project P166538 also in Gambia.
As part of initiatives of the World Bank in Africa for
As a response to the unique set of pressures and challenges coastal resilience, this study analyzes the historic,
of coastal Africa, the World Bank and the Global Facility present, and future exposure to flooding of coastal cities
for Disaster Reduction and Recovery (GFDRR) launched in the SSA. Disaster risk reduction has been hindered
the City Coastal Resilience Africa (CityCORE). CityCORE’s in the region, among other factors, by limited risk
goal is to create knowledge and foster policy dialogue and identification and assessment and weak integration of
investments for urban resilience in coastal cities in Africa. disaster risk reduction in development plans (World
The aim is to catalyze a shift from a primarily siloed, Bank 2008). This study aims to address this gap and
single-stream, city level resilience approach into a longer improve the capacity of communities, institutions,
term, more comprehensive and multidisciplinary one businesses, and infrastructure systems to prepare,
that supports risk management, resilience planning, and adapt, and grow sustainably. By generating and
access to sources of financing. CityCORE also leverages providing city-based information on coastal risk across
global knowledge and partnerships to support Africa´s sectors and geographic areas of SSA. This information
resilience objectives. These initiatives are essential is now available to each city to help can help catalyze
as governments implement the 2030 Sustainable and direct coordinated investment to build climate
Development Agenda, the Sendai Framework for Disaster resilience and adaptation.
Notes Cian, F., Blasco, J.M.D., and Carrera, L. 2019. Sentinel-1 for
Monitoring Land Subsidence of Coastal Cities in Africa Using
1. World Bank. Supporting a Green, Resilient and Inclusive PSInSAR: A Methodology Based on the Integration of SNAP
Recovery on West Africa’s Coast. Blog. https://www.
and StaMPS. Geosciences, 9.
wacaprogram.org/article/supporting-green-resilient-and-
inclusive-recovery-west-africas-coast Church, J.A., Clark, P.U., Cazenave, AGregory, J.M., Jevrejeva,
S., Levermann, A., Merrifield, M. A., Milne, G., ANerem, R., Nunn,
2. C40 Cities. https://www.c40.org/other/the-future-we-don-
t-want-staying-afloat-the-urban-response-to-sea-level- P.D., Payne, A.J., Pfeffer, W.T., Stammer, D., and Unnikrishnan,
rise A. S. 2013. Sea level change. Clim. Chang. 2013 Phys. Sci.
Basis. Contrib. Work. Gr. I to Fifth Assess. Rep. Intergov. Panel
3. World Bank and Ministry of Environment and Nature Clim. Chang., 1137–1216.
Protection. Economic and spatial study of coastal zones’
vulnerability and adaptation to climate change. Dasgupta, S., Laplante, B., Murray, S., and Wheeler, D. 2009.
Climate Change and the Future Impacts of Storm-Surge
4. World Vision. https://www.worldvision.org/disaster-relief-
Disasters in Developing Countries. Working Paper 182.
news-stories/2019-cyclone-idai-facts#timeline
5. Environmental statistics: Situation in Federal Republic of Geneva Association 2014. The Geneva Association. 2014.
Nigeria; Being Country Report Presented at the Workshop The Global Insurance Protection Gap Assessment and
on Environment Statistics Held in Dakar, Senegal 28th Recommendations.
February – 4th March 2005 GFDRR. 2014. Understanding Risk in an evolving world.
Emerging best practices in Natural Disaster Risk Assessment.
References Washington, D.C.
Adeyinka M.A., P.O., Bankole., and O., Solomon. 2005. Hallegatte, S. 2019. Disasters’ impacts on supply chains.
Environmental statistics: Situation in Federal Republic Nature Sustainability, 2, 791–92.
of Nigeria. Country Report Presented at the Workshop
on Environment Statistics Held in Dakar, Senegal 28th Hallegatte, S., Green, C., Nicholls, R.J., and Corfee-Morlot, J.
February – 4th March 2005. Dakar. http://mdgs.un.org/unsd/ 2013. Future flood losses in major coastal cities. Nat. Clim.
environment/nigeria.pdf Chang. 3, 802–06.
6 | Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa
Hallegatte, S., Rentschler, J., and Rozenberg, J. 2019. Work. Gr. II to Fifth Assess. Rep. Intergov. Panel Clim. Chang.
From Resilient Assets to Resilient Infrastructure Services. eds. Barros, V.R., Field, C.B., Dokken, D.J., Mastrandrea, M.D.,
In: Lifelines Resilient Infrastruct. Oppor., Sustainable Mach, K.J., Bilir, T.E., Chatterjee, M., Ebi, K.L., Estrada, Y.O.,
Infrastructure Series. The World Bank, pp. 109–25. Genova, R.C., Girma, B., Kissel, E.S., Levy, A.N., MacCracken,
S., Mastrandrea, P.R. and White, L.L. Cambridge University
Intergovernmental Panel on Climate Change’s (IPCC) Sixth
Press, Cambridge, United Kingdom and New York, NY, USA,
Assessment Report (AR6) Climate Change 2021: The Physical
pp. 1199–1265.
Science Basis.
Reguero et al (2019). Financing coastal resilience by combining
IPCC. 2014. Climate Change 2014: Impacts, Adaptation,
nature-based risk reduction with insurance. Ecological
and Vulnerability. Part B: Regional Aspects. Contribution
Economics. https://www.sciencedirect.com/science/article/
of Working Group II to the Fifth Assessment Report of the
abs/pii/S0921800918315167.
Intergovernmental Panel on Climate Change. Barros, V.R., C.B.
Field, D.J. Dokken, M.D. Mastrandre. Cambridge University UN (2018) 2018 Revision of World Urbanization Prospects.
Press, Cambridge, United Kingdom and New York, NY, USA. Population Division of the UN Department of Economic and
Social Affairs (UN DESA) https://population.un.org/wup/
Kebede, A.S., Brown, S., and Nicholls, R.J. 2010. Synthesis Publications/Files/WUP2018-Report.pdf
Report: The Implications Of Climate Change And Sea-Level
Rise In Tanzania − The Coastal Zones. UNISDRR. 2009. Global assessment report on disaster risk
reduction.
Kompas, T., Pham, V.H., and Che, T.N. 2018. The Effects of
Climate Change on GDP by Country and the Global Economic (Wong et al 2014)
Gains From Complying With the Paris Climate Accord. Earth’s Wong, P.P., Losada, I.J., Gattuso, J.-P., Hinkel, J., Khattabi, A.,
Futur., 6, 1153–173. McInnes, K.L., Saito, Y., Sallenger, A., and IPCC, I.P. on C.C.
Morim, J.; Vitousek, S.; Hemer, M.; Reguero, B.; Erikson, L.; 2014. Coastal systems and low-lying areas. Climate Change.
Casas-Prat, M.; Wang, X.L.; Semedo, A.; Mori, N.; Shimura, 2014: Impacts, Adaptation and Vulnerability. Part A Global
T.; et al. Global-scale changes to extreme ocean wave events Sect. Asp. Contributing Working Group II to Fifth Assessment
due to anthropogenic warming. Environ. Res. Lett. 2021, 16, Report. Intergovernmental Panel Climate Change. 361–409.
74056, doi:10.1088/1748-9326/ac1013. World Bank. 2008. Report on the status of Disaster Risk
Munich Re. 2013. Economic consequences of natural Reduction in Sub-Saharan Africa.
catastrophes: Emerging and developing economies particularly World Bank. 2013a. Economic and Spatial Study of the
affected Insurance cover is essential. Position paper. Munich, Vulnerability and Adaptation to Climate Change of Coastal
Germany. Areas in Senegal. Washington, D.C.
Neumann, B., Vafeidis, A.T., Zimmermann, J., and Nicholls, R.J. World Bank. 2013b. Turn Down the Heat: Climate Extremes,
2015. Future Coastal Population Growth and Exposure to Sea- Regional Impacts, and the Case for Resilience. 1–34.
Level Rise and Coastal Flooding - A Global Assessment. PLoS Washington, D.C.
One, 10, e0118571.
World Bank Group. 2014. Turn Down the Heat : Confronting
Niang, I., Ruppel, O.C., Abdrabo, M.A., Essel, A., Lennard, C., the New Climate Normal. Washington, D.C.
Padgham, J., and Urquhart, P. 2014. Africa. In: Clim. Chang.
2014 Impacts, Adapt. Vulnerability. Part B Reg. Asp. Contrib. World Economic Forum. 2019. The Global Risks Report 2019.
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | 7
Methods and Approach
This report aims to inform policy dialogue around well as the critical infrastructure based on data available
challenges in coastal areas by providing a regional to date in Open Street Maps. The main data sources
overview of risk across the coastal cities in SSA (figure and methods are briefly described in this chapter, but
2-1). The analysis integrates spatial and socioeconomic a more detailed description can be found in Appendix A.
information pertinent to city-level decision making.
The study focuses on cities that are less than about 10
Exposure data
kilometers from the coastline, analyzing historic trends
of urban expansion, population, and infrastructure Growth of historic change in population and built-
assets that characterize changes in exposure and up areas in each city was calculated using the Joint
flood risk; and identify key challenges in each city and Research Centre’s global human settlement layer
country. The study aims to inform the policy dialogue (GHSL) (Pesaresi et al. 2015). GHSL uses global, fine-
between cities, World Bank teams, and development scale satellite image data streams, census data, and
partners engaged in supporting climate resilience in the crowd sources or volunteered geographic information
coastal systems and economies. sources (see detailed information in appendix A). Urban
centers in GHSL are defined from cut-off values on
The analysis combines information on population, built- resident population and built-up areas in a one-square
up area and critical infrastructure with exposure to kilometer global uniform grid, as of 2015. The built-up
flood hazards (figure 2-1). Flood zones were calculated areas data correspond to areas with buildings—enclosed
for each city for both rainwater flooding—pluvial and constructions above ground intended or used for shelter
fluvial—and coastal flooding. Historic data obtained or production of economic goods—and that refer to
from remote sensing were used to calculate population structures at a horizontal resolution of 30 meters.
and built-up area in the flood hazard zones over time, as Population data were derived from national population
$
8 | Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa
census data and global built-up areas extracted from effectively or urban drainage systems are overwhelmed
Earth Observation data analytics,1 with a horizontal by excessive water flow. Fluvial or riverine flooding
resolution of 250 meters. occurs when runoff from excessive rainfall causes a
river to exceed its capacity. Coastal flooding, however,
The GHSL urban centers data were used to identify the
occurs when normally dry, low-lying land is flooded by
cities as those urban areas less than about 10 kilometers
seawater, induced by storm surges, wave run-up, or
from shoreline. The city boundaries in GHSL were
by a combination of these factors with high tides and
expanded by one kilometer to include the coastal areas
elevated levels of river discharge (box 2-1).
and population in suburbs that were not included in the
original urban polygons in the GHSL dataset. The urban Flood hazard zones, for coastal and rainwater sources,
area and population were calculated on a city basis for were defined to calculate the people, built-up area, and
the years 1975, 1990, 2000, and 2015. This multiyear infrastructure exposed and at risk as follows:
snapshot provides unique insight into the temporal
● Coastal flooding was defined based on the Surging
changes in the urban built-up area and population in
Seas project from Climate Central2 using sea level
each coastal city, and its relation to flood risk.
model results and the recently released global coastal
Information on critical infrastructure, including roads, DEM. This DEM uses neural networks to reduce
railways, schools, and hospitals, was obtained from previous errors in global elevation models for coastal
Open Street Maps, a popular and well-supported dataset areas. The elevation model was combined with the
of volunteered geographic information. However, the extreme sea levels associated to the 1-, 10- and
completeness of information in Open Street Maps varies 100-year return periods from the global tides and
between countries and direct comparisons between surge reanalysis (Muis et al. 2016).3 The extent of
countries should be taken with caveats. the coastal flood maps was limited to five kilometers
from the shoreline to avoid large overestimations
of coastal flooding in low-lying areas inland that
Flood hazards
would not be affected by storm surge—inlets and
Flood hazards were determined using a combination of bays—based on visual inspection of the shoreline
global numerical models, remote sensing observations and city location. Cities inland from this threshold
for riverine and pluvial hazards, a new digital elevation were not considered under the influence of coastal
model (DEM) for coastal areas, and information on dynamics. The shoreline was defined based on the
extreme sea level data. Extreme rainfall can produce global self-consistent hierarchical high resolution
floods when the ground cannot absorb rainwater geography dataset. Coastal flooding maps for the
BOX 2-1 Data availability
This regional set of maps constitutes a first homogenous The main limitations of this approach are that: (i) the flood
overview of the coastal and rainwater flood hazard zones hazard mapping relies on specific hazard probabilities or
for the entire region. The resulting flood hazard maps are return periods; (ii) the models do not account for local effects
available for each coastal city in SSA. The main outputs such as flood protection measures; and (iii) they also rely on
area available upon request for each city, and include: regionally available information on elevation, flood hazard,
and city exposure (see appendix A for a more detailed
● Coastal and rainwater flood hazard maps
discussion of uncertainties). Therefore, the maps should be
● Historic information on built-up area change considered a first estimate of flood risk at the local scale.
Yet, local data on elevation, building exposure, population
● Historic shoreline change (not used in the analysis)
distribution, critical infrastructure, and flood management
These maps can be used in absence of more precise, local measures—flood walls, dunes ridge, flood proofing, coastal
information. However, the hazard zones are defined based on protection measures, drainage systems—as well as refined
global models and data, and the results should be taken with definition of the coastal and rainfall dynamics could be
caveats at local scales. included to improve the analysis at the local level.
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | 9
mid-century were also developed by adding the Sea Level Rise
effect of sea level rise to the 10-year flood zone to
Sea level rise estimates were based on the regional
estimate future changes in coastal flooding (chapter 3).
sea level rise projections in (Kopp et al. 2017) that use
● Rainwater flood hazard maps were calculated Antarctic ice sheet physics to create a probabilistic
combining the flood modeling results with remote ensemble for the representative concentration
sensing data. The model results were obtained from pathway (RCP) 8.5. The values of sea level rise by the
Fathom-Global (Sampson et al. 2015) that provide mid-century were added to the 10-year flood zone to
flood depths at 90 meters resolution through compare changes in coastal flooding from sea level
modeling of fluvial and pluvial flooding. The precision rise. A comparison is provided in the results between
is adequate for regional areas, but its horizontal the total number of people exposed to the provided for
resolution has limitations for local and very flat areas. the 10-year coastal flooding in 2015 and the projected
To address this limitation, the numerical modeling flooded area in the year 2050.
results were combined with the Joint Research
Centre Global Surface Water dataset (GSW) (Pekel
et al. 2016) that provides seasonal water maps— Calculation of risk: population affected and
water surface that is not permanent—from 1984 economic impacts
to 2018 at 30 meters resolution based on optical The calculation of number of people affected by floods
satellite observations (Landsat). This remote sensing and the economic damages were estimated for each
source provides real observations of flooded areas. flood hazard zone. Economic damages were calculated
However, the dataset presents limitations: (i) it does using flood damage curves and maximum damage
not distinguish between river, coastal, and pluvial values for each country. The flood water depths of
flooding; (ii) the repeating cycle of the satellites (16 coastal and rainwater flooding were combined with
days) may not completely capture flood events, which damage curves to estimate damage degree on buildings.
may last days to weeks; (iii) and the cloud coverage, In the absence of local data, the analysis uses damage
especially relevant in tropical areas can prevent the functions developed by the Joint Research Centre for
observation of specific single events, which results in Africa, based on literature and local normalized values
an incomplete time series of flood events. As a result, (Huizinga et al. 2017) (appendix B, see note #1, and
not all the floods are detected by CGSW, while those figure B-2). An estimated mean damage degree function
captured may be underestimated for the peak of was obtained using the residential buildings land use,
the flood event. Therefore, statistics, such as return assuming this was the most widespread building type in
period, based on these data are conservative. These
the built-up urban footprint. Water depths in the flood
two datasets are complementary and were combined
zones were used to obtain damage degrees to buildings
to determine the flood hazard zones for each city:
for every pixel of built-up area in each city. For both
(i) the satellite-based maps underestimate flood
flood hazards, the event associated with the 1-year
extent but capture areas that have been flooded in
return period were assumed to produce no damage
the historic record, which represents a lower bound of
such as nuisance flooding, and were discounted from
flooded areas; (ii) on the other hand, the model results
the calculations for the other return periods.
provide areas that could be flooded by extreme
events in the future. The combined flood hazard maps A maximum damage value per unit of built-up area was
were calculated by selecting the maximum pixel obtained for each country applying conversion factors
value between the two datasets. The return periods to consider depreciated value for the construction cost
of satellite-based floods maps were empirically at 0.6 and undamageable fraction of the buildings at
calculated based on the number of observations and 0.4, (Huizinga et al. 2017). Correcting for the fraction of
occurrences. These return periods were combined with buildings per land use was not necessary because this
those with the same probability from the computer analysis used the most recent version of GHSL built-
model flood outputs. up area, which provides information of the percentage
10 | Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa
of buildings per pixel. The resulting maximum damage Notes
values per country varied between a maximum of 1. An urban or rural classification model determines a “Degree
US$34.3 per square meter in Equatorial Guinea to a of urbanization” and discriminates between (i) cities,
minimum of US$7.2 per square meter in Somalia. These (ii) towns and suburbs and (iii) rural areas, based on the
population density. The methodology for the delineation
values are in the range of local estimates for local
of urban and rural areas was developed by the European
analysis developed by the World Bank in the region. Commission, the Organization for Economic Co-operation
and Development (OECD), the Food and Agriculture
The results of flood damages associated to each return Organization of the United Nations (FAO), UN-Habitat and
period, per type of hazard were integrated into an the World Bank.
annual average damage, which was calculated as the 2. Climate Central is a science and news organization that
expected value of the damage–probability distribution bridges the scientific community and the public, providing
(figure 2-2 and appendix B, note #2). clear information to help people make sound decisions
about the climate. https://www.climatecentral.org
3. Environmental statistics: Situation in Federal Republic of
Nigeria; Being Country Report Presented at the Workshop
Figure 2-2 Infographic of flood risk calculation. on Environment Statistics Held in Dakar, Senegal 28th
February – 4th March 2005.
References
Urban centers
Huizinga, J., Moel, H. de, and Szewczyk, W. 2017. Global flood
depth-damage functions.
Flood maps Kopp, R.E., DeConto, R.M., Bader, D.A., Hay, C.C., Horton,
(coastal and R.M., Kulp, S., Oppenheimer, M., Pollard, D., and Strauss, B.H.
rainwater)
2017. Evolving Understanding of Antarctic Ice-Sheet Physics
and Ambiguity in Probabilistic Sea-Level Projections. Earth’s
Future. 5, 1217–233.
Population data
Muis, S., Verlaan, M., Winsemius, H.C., Aerts, J.C.J.H., and
Ward, P.J. 2016. A global reanalysis of storm surges and
extreme sea levels. Nat. Commun., 7, 11969.
Built-up area Pekel, J.-F., Cottam, A., Gorelick, N., and Belward, A.S. 2016.
High-resolution mapping of global surface water and its long-
term changes. Nature, 540, 418–22.
Pesaresi, M., Ehrilch, D., Florczyk, A.J.., Freire, S., Julea, A.,
Flood damages Kemper, T., Soille, P., and Syrris, V. 2015. GHS built-up grid,
per city (AAD) derived from Landsat, multitemporal (1975, 1990, 2000,
2014). European Commission, Joint Research Centre. http://
data.europa.eu/89h/jrc-ghsl-ghs_built_ldsmt_globe_r2015b
Sampson, C.C., Smith, A.M., Bates, P.D., Neal, J.C., Alfieri, L.,
and Freer, J.E. 2015. A high-resolution global flood hazard
Source : Infographic modified from City Scan, World Bank. model. Water Resour. Res., 51, 7358–381.
Note: Flood damages are calculated by integrating flood maps,
population and built-up data, within coastal cities polygons. The
assessment provides a regional perspective of coastal risk, and
information on the critical resilience challenges that coastal cities
face, using the best publicly available global geospatial datasets and
open source tools. AAD stands for Average Annual Damage.
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | 11
Expansion of coastal cities
in Sub-Saharan Africa
The 153 cities with more than 50,000 people in coastal Population growth
SSA have grown significantly since the 1970s (figure
Approximately 23.4 million people were living in coastal
3-1). In 1975, the urban footprint in SSA constituted
cities of SSA in 1975; by 2015 the coastal population
approximately 3,477 square kilometers of built-up area,
had risen to 86.4 million (table 3-1). This represents a
but it expanded to about 5,520 square kilometers by
267 percent increase over 40 years and is considerably
2015, an increase of 1.6 times or averaging 1.1 percent
larger than the increase in urban built-up area of about
growth per year. Meanwhile, the population in coastal
65 percent over the same period. Furthermore, just five
cities experienced an increase of 3.7 times from 23.4
countries, in order of magnitude—Nigeria with 18.8
million people in 1975 to 86.4 million people by 2015
million; Angola, with 9 million; South Africa with 8.4
at an average rate of 3.3 percent per year (figure 3-1).
million; and Tanzania with 6.5 million—accrue half of
These trends clearly demonstrate the concentration
the coastal people in the entire region. Dar es Salaam,
of people in coastal areas that has experienced rapid
Lagos, and Luanda are the coastal cities with the most
development and urbanization.
population (table 3-2).
Figure 3-1 Population and built-up area growth in all the coastal cities of Sub-Saharan Africa.
Population (million people)
86.4
51.4
+267%
38.0
23.6
Built up area (sq. km)
6,393
5,911
5,047
+65%
3,863
1975 1990 2000 2015
12 | Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa
Table 3-1 Coastal population and built-up area by country over time.
Population (thousands) Built up Area (square km)
Country 1975 1990 2000 2015 1975 1990 2000 2015
Nigeria 6,163 9,260 11,791 18,755 545.5 845.0 1,047.8 1,126.2
Angola 194 668 1,595 9,014 366.5 423.7 503.0 552.9
South Africa 3,842 6,080 7,803 8,361 629.5 799.0 893.3 927.1
Tanzania 734 1,681 2,882 6,483 217.6 224.9 303.9 320.3
Ghana 1,181 2,211 3,215 5,552 379.9 458.1 521.2 550.5
Mozambique 2,216 2,815 3,601 5,010 187.6 244.9 301.8 320.3
Côte d’Ivoire 1,318 2,551 3,476 5,005 217.3 218.4 240.5 257.6
Senegal 1,542 2,332 3,047 4,461 156.0 171.8 200.3 224.6
Cameroon 572 1,175 1,752 3,477 49.6 78.3 91.1 95.7
Guinea 695 1,015 1,541 2,627 79.4 108.8 143.4 165.5
Benin 1,021 1,273 1,572 2,517 124.3 178.4 208.9 219.1
Somalia 748 1,187 1,283 2,276 43.2 57.1 63.1 63.8
Togo 537 846 1,140 2,146 105.4 124.2 156.3 161.3
Liberia 336 523 804 1,528 95.7 97.6 102.1 105.0
Sierra Leone 397 665 752 1,361 39.0 45.6 52.7 61.0
Kenya 314 605 805 1,357 19.0 28.8 44.7 49.1
Mauritania 259 465 711 1,254 64.1 71.9 80.2 89.2
Madagascar 175 355 576 1,018 21.5 37.2 39.6 54.3
The Gambia 206 391 556 1,006 54.8 68.2 81.3 82.3
Gabon 322 440 506 731 32.9 46.7 49.5 50.5
Republic of the Congo 31 317 624 620 25.5 25.5 31.6 34.4
Djibouti 182 423 470 552 0.8 1.3 1.4 1.7
Guinea-Bissau 129 218 309 509 17.9 19.5 25.0 26.3
Equatorial Guinea 59 108 182 399 16.7 19.3 22.7 24.2
DRC 360 346 317 307 7.7 7.7 10.1 10.8
Namibia 10 23 39 66 0.0 0.1 0.1 0.1
Eritrea 33 34 32 48 0.9 1.0 1.0 1.0
Total 23,576 38,008 51,383 86,443 3,498 4,403 5,216 5,575
Note: Values represent population (in thousands) and built-up area (square kilometers), as of 2015. Countries are ordered by decreasing
population values in 2015.
Furthermore, 17.8 million people lived less than one therefore, may be the most directly impacted by storms
kilometer from the sea in 2015, compared with 7.2 million and climate change, including the effects of sea level
in 1975. Although this represents a lower increase than rise. Angola, Ghana Mozambique, Nigeria, and Senegal
the total population in coastal cities, these people can be present the greatest number of people close to the sea
considered directly dependent on marine resources, and and half of the people in the entire region (figure 3-2).
Figure 3-2 Population within one kilometer from the shoreline and change over time.
18,000,000
16,000,000
14,000,000
Rest of countries
12,000,000
10,000,000
8,000,000 Angola
6,000,000 Madagascar
4,000,000 Senegal
Nigeria
2,000,000
Mozambique
-
1975 1990 2000 2015
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | 13
Table 3-2 Top 10 cities by population in coastal cities in 2015.
Increase over the last
15 years
City Country 1975 1990 2000 2015 (% per year)
Lagos Nigeria 3,925,586 6,011,414 7,773,122 11,688,074 3%
Luanda Angola 585,202 (*) 1,440,046 (*) 2,763,833 (*) 6,861,222 9.4% (**)
Dar es Salaam Tanzania 475,635 1,231,368 2,253,419 5,439,560 9%
Abidjan Ivory Coast 1,197,950 2,327,089 3,222,655 4,597,538 3%
Accra Ghana 1,028,120 1,892,016 2,705,562 4,510,140 4%
Cape Town South Africa 1,207,753 2,179,521 3,033,535 3,569,263 1%
Dakar Senegal 1,159,437 1,752,765 2,272,241 3,374,761 3%
Douala Cameroon 479,497 1,013,329 1,572,219 2,988,707 6%
Durban South Africa 1,723,893 2,494,341 2,992,371 2,959,472 0%
Conakry Guinea 670,969 979,674 1,494,181 2,504,251 5%
Note: The values indicate the population (counts) for each year. (*) historic values in Luanda are considered incorrect based on the GHSL database
and were corrected using information from United Nations - World Population Prospects and scaled based on the spatial distribution obtained for
the year 2015. The 2015 values are close to United Nations’ World Population Prospects (**) Rate of increase calculated based on the UN World
Population prospects.
Urban expansion in coastal areas The number of coastal cities with more than one
million people has been increasing at each reference
Coastal cities have significantly grown in urban density
year since 1975; only seven cities in 1975 compared
in the last four decades (table 3-3). Accra, Durban and
with 20 cities by 2015. Furthermore, in the last 15
Lagos are the largest coastal cities by built-up area.
years with data, eight new cities that entered such
They have been growing at rates of over 5 percent per
ranking. Figure 3-6 shows the urban expansion
year since the year 2000. These patterns of urban
pattern clearly increasing over time for the largest cities.
expansion are represented in figures 3-3 to 3-5 for
these three cities.
Table 3-3 Top 10 cities by built-up area in 2015.
Increase percent
City Country 1975 1990 2000 2015 per year
Lagos Nigeria 320 684 889 1,170 7%
Accra Ghana 264 485 555 876 6%
Durban South Africa 129 391 448 785 13%
Luanda Angola 184 221 281 771 8%
Cape Town South Africa 323 567 606 743 3%
Dar es Salaam Tanzania 196 196 282 654 6%
Maputo Mozambique 30 174 212 418 32%
Abidjan Ivory Coast 305 305 328 394 1%
Cotonou Benin 88 150 194 300 6%
Conakry Guinea 63 127 147 273 8%
Note: The values indicate the built-up area (square kilometers) for each year.
14 | Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa
Figure 3-3 Urban growth in Lagos, Nigeria, the largest coastal city in the region.
Note: Colors indicate urban built-up area over years. The color scale indicates the year when the
land was built up. The red indicates built-up area in 1975, while purple represents the most recent
period, in year 2015.
Figure 3-4 Urban growth in Accra, Ghana.
Note: Colors indicate urban built-up area over years. The color scale indicates the year when the
land was built up. The red indicates built-up area in 1975, while purple represents the most recent
period, in year 2015.
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | 15
Figure 3-5 Urban growth in Luanda, Angola.
Note: Colors indicate urban built-up area over years. The color scale indicates the year when the
land was built up. The red indicates built-up area in 1975, while purple represents the most recent
period, in year 2015.
Figure 3-6 Cities with over 1 million people over time.
1975 1990 2000 2015
Lagos
Luanda
Dar es Salaam
Abidjan
Accra
Cape Town
Dakar
Douala
Durban
Lagos Conakry
Maputo
Lome
Catonou
Mogadishu
Durban Monrovia
Cape Town Freetown
Abidjan Port Harcourt
Dakar Nouakchott
Maputo Mombasa
Accra Massawa
Note: The values represent the population (million) for each year. Only cities at each reference year with more than 1 million people are shown. In
the last 15 years with data, 8 new cities that entered the ranking, including Luanda, the second most populated city. (*) Population in Luanda for
the first three time is incorrect in the GHSL. The historic growth in population has been corrected using information from United Nations - World
Population Prospects and scaled based on the spatial distribution obtained for the year 2015.
16 | Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa
The rates of coastal growth are also evident across all Lagos, Accra and Luanda (figures 3-3, 3-4 and 3-5).
countries (figure 3-7). Coastal cities in Namibia, Djibouti, However, cities have also expanded inland, close to
Madagascar, Kenya, Nigeria and Guinea have expanded waterways, as coastal zones became more urbanized.
to the coastal zones more than any other country. These Historic patterns can provide important insight into risk
patterns of urban growth indicate that the coastline, to inform future pathways. The next section analyzes
seaward city waterfront, has been historically highly how much of this urban expansion has taken place in
urbanized, with development occurring in areas not flood hazard zones.
previously occupied, as seen in historic patterns for
Figure 3-7 Countries by built-up area (BU) and population in 2015.
BU 2015 (sq km)
Coastal City Growth from 1975 to 2015
2 Nigeria
3
4
900 South Africa
5
6
600
Ghana Angola
300 Tanzania
Mozambique
Benin Senegal
Guinea
Togo
Mauritania
The Gambia Cameroon
Leone Somalia
Sierra Kenya
Madagascar
Gabon
Republic of the Congo
Guinea-Bissau
Equatorial Guinea
0 Namibia DRC Djibouti
100 1000 10000
Population 2015 (#)
Note: The size of the circles represents the historic growth in population (ratio of change) between 1975 to 2015. Angola has increased its coastal
population the most.
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | 17
Flood risk in coastal cities is increasing
Flooding is the highest hazard risk in many countries of city level, this chapter characterizes how the historic
SSA (box 4-1). To help understand existing risks in SSA urban growth in coastal cities has shaped the prevailing
and inform future climate-resilient development at the flood risk profile.
BOX 4-1 Flood and storm impacts in coastal cities.
Flood and storm impacts man-made factors, such as poorly planned infrastructure
and urban development. The government is trying to plan
Flood and storm impacts in coastal cities such as Accra, for longer term adaptation options to save the city and its
Benin, Dar Es Salaam, or Saint Louis during recent years have communities, while facing limited resources and increasing
demonstrated the significant cost and need of preventive anthropogenic pressures.
action against the flood hazards. Fiscal constraints and
poverty render both governments and households limited The 2020 floods
capacity to prepare, cope, and adapt to changing conditions
and climate extremes. In 2020, record-breaking sea temperature difference
between the western and eastern Indian Ocean caused
Saint Louis, for example, is one of the most important fishing extreme flooding in many countries in Africa. Between
communities in West Africa and a UNESCO site. The city has August and September of 2020 alone, more than 1.21 million
experienced regular ocean storms the last five years since people in 12 different countries were affected by floods with
2015 that have swelled and knocked rows of houses off the many other countries experiencing widespread rainfall,
coast along the Langue de Barbarie. This sandy peninsula, leading to transboundary flooding. Some of the countries
densely populated, protects the old city from the ocean, but most affected were Cameroon, Chad, Ethiopia, Kenya, Mali,
has suffered severe beach retreat that now allows flooding Niger, Nigeria, Somalia, South Sudan, Sudan, and Uganda.
on the city.1 As a result of storm events, 800 families from In West Africa, Ghana, Ivory Coast, and Togo were also
the city’s seashore had to be evacuated to more protected severely impacted. The floods of 2020 also affected coastal
zones, while the city is facing the challenges of a rapid cities. For example, in October 2020, floods in Accra,2
shoreline retreat caused by a combination of climatic and Benin,3 or Dar Es Salaam4 produced damages, fatalities, and
Box photo 4-1.1. Damaged houses from wave action in Box photo 4-1.2. Tents housing people who lost their homes to rising sea levels and
Saint Louis, Senegal. Source of image: World Bank. coastal erosion in 2018. Source of image: World Bank.
18 | Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa
BOX 4-1 Flood and storm impacts in coastal cities (cont.).
thousands of displaced people. A few months before, Lagos maintenance of existing infrastructure. Similar impacts
was swamped and houses damaged after heavy rain in June could be exacerbated in the future by climate change through
2020.5 Extreme precipitation across Africa showed once its effects on seasonal rainfall in Africa. Understanding
again how cities, coastal or not, in SSA were poorly equipped historic and present flood hazard risk at the city level is an
to cope with flooding, due to informal or unplanned urban initial step to manage the existing risks and inform more
expansion, dysfunctional or unsuitable drainage systems, climate-resilient urban development in the future.
poor waste management clogging drains, and overall limited
Box photo 4-1.3. Floods in Accra, Ghana. Box photo 4-1.4. Floods in October 2020.
Source of images: Shutterstock.com
Notes
1
http://floodlist.com/africa/senegal-city-races-to-move-families-as-sea-swallows-homes
2
http://floodlist.com/africa/ghana-flash-floods-cause-traffic-chaos-in-accra
3
http://floodlist.com/africa/benin-floods-september-2020
4
http://floodlist.com/africa/floods-dar-es-salaam-tanzania-10-dead
5
http://floodlist.com/africa/nigeria-floods-lagos-june-2020
To this scope, population and built-up data are analyzed increased the number of people and urban area exposed
in combination with present-day coastal and rainwater to flooding over the years. More than 10.5 million people
flood hazard zones obtained for each city in SSA. The lived in the 10-year flood zone in 2015, compared to less
results depict how and why risk has been increasing than 4 million in 1975 (figure 4-1). The significance of
across the region, which provides critical input to shape this figure can be grasped considering that the people
a more climate-resilient future. living in the 10-year flood zone in the 159 coastal cities
represent 83 percent of the total people living less than
one kilometer from the sea for the entire SSA. Such
Increase in exposure to flood hazards comparison speaks of the concentration of people in
The historic evolution of exposure to floods in SSA coastal cities and also of how many people are exposed
to flood hazards in them.
demonstrates that part of the rapid urban growth has
occurred in flood-prone areas. New urban development The total built-up area in the 10-year flood zone has also
has taken place in areas exposed to flooding, which has increased steadily since 1975, and grew even faster after
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | 19
2000 (figure 4-1). The total coastal built-up area across Abidjan, Dakar and Lagos are the cities where most
SSA was 1,257 square kilometers in 2015, which is almost people are exposed to the 10-year flood (table 4-1). Accra,
five times the 258 square kilometers exposed in 1975 Dar es Salam, and Lagos are the top three cities with
in the 10-year flood zone. Built-up areas in flood-prone most built-up area in the same flood zone (table 4-2).
zones also expanded faster than population growth. Lagos (flood map areas in figure 4-1) has also in excess of
Flood urban exposure increased by 126 percent between 54 kilometers of roads exposed to coastal flooding and
2000 and 2015, but the population increased less by 47 kilometers to rainwater floods (table 4-3). Meanwhile,
58 percent during the same period. Therefore, built up Luanda has 41 kilometers exposed to rainwater flooding
area has expanded in relative terms more built-up area but only 7 kilometers to coastal hazards. Cape Town
has expanded more rapidly in flood prone areas than and Durban are, however, the two cities with the largest
population growth in the same areas, which indicates number of kilometers of transportation network in
higher socio-economic impacts from floods over time. the rainwater 10-year flood hazard zone. Cities and
Figure 4-1 Population and built-up areas in the 10-year flood hazard zone for coastal and rainwater sources.
(a) Population (#) exposed in the 10-year (b) Population (#) exposed in the 10-year
(a) Population (#) exposed in the 10-year (b) Population (#) exposed in the 10-year
zone
flood (a) over the(#)
Population years
exposed in the 10-year flood zone by countries
(b) Population (#) exposed in the 10-year
flood zone over the years flood zone by countries
flood zone over the years flood zone by countries
12 4.0
12 4.0
Millions
Millions
Millions
3.5
Millions
10 3.5
10
3.0
3.0
8 2.5
8 2.5
6 2.0
6 2.0
1.5
4 1.5
4 1.0
1.0
2 0.5
2 0.5
0.0
- 0.0
-
1975 1990 2000 2015
1975 1990 2000 2015
(d) Built up area (sq km) exposed in the 10-year
(c) Built up
(c) area (sq
Built-up km)
area exposed
(sq in the
km) exposed in10-year
the (d) Built up (d)
area (sq km)
Built-up exposed
area in
(sq km) the 10-year
exposed in the
(c) Built up area (sq km) exposed in the 10-year flood zone by countries
flood zone over the
10-year years
flood zone over the years flood zone by countries
10-year flood zone by countries
flood zone over the years
1,400 300
1,400 300
1,200 250
1,200 250
1,000 200
1,000 200
800 150
800 150
600
600 100
100
400
400 50
50
200
200 -
- -
-
1975 1990 2000 2015
1975 1990 2000 2015
Total rainwater Total coastal
Total
Note: The left panels provide the regional estimates; the panels provide aTotal
rainwater
right coastal
breakdown by country. The values for each country are provided in
the appendixes.
20 | Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa
countries exhibit large differences in exposure to each As examples of the new flood risk information now avail-
type of flood hazard (figure 4-1 and tables 4-1 and 4-2). able for coastal cities in SSA, the flood maps for Lagos and
For example, Accra and Dar Es Salaam present higher Luanda are represented in figures 4-2 and 4-3. Lagos is the
exposure to rainwater floods, whereas coastal flood city with largest exposure to floods. Luanda is the second
hazards are dominant in Abidjan or Dakar. most populated coastal city, and also second by exposure
of infrastructure to coastal and rainwater flooding.1
Table 4-1 Top 10 cities by population exposed to flooding (10-year flood hazard zone).
Population in the 10 year flood zone
Rainwater
City Country floods City Country Coastal floods City Country Total floods
Dar es Salaam Tanzania 314,433 Lagos Nigeria 1,104,240 Lagos Nigeria 1,357,971
Accra Ghana 294,568 Dakar Senegal 621,926 Dakar Senegal 719,195
Lagos Nigeria 253,731 Abidjan Côte d’Ivoire 506,211 Abidjan Côte d’Ivoire 684,838
Monrovia Liberia 250,567 Cotonou Benin 432,272 Dar es Salaam Tanzania 619,770
Benguela Angola 193,940 Beira Mozambique 344,635 Conakry Guinea 475,412
Freetown Sierra Leone 188,877 Dar es Salaam Tanzania 305,338 Cotonou Benin 474,192
Abidjan Côte d’Ivoire 178,627 Conakry Guinea 301,008 Freetown Sierra Leone 395,187
Conakry Guinea 174,404 Nouakchott Mauritani 277,326 Monrovia Liberia 379,356
Luanda Angola 168,758 Djibouti Djibouti 234,533 Beira Mozambique 363,177
Cape Town South Africa 164,684 Freetown Sierra Leone 206,310 Accra Ghana 351,285
Note: (*) According to the original GHSL data, Massawa in Eritrea ranked second in population exposed, but this was considered an error in the
database and was, therefore, not considered in the analysis, as the population in Massawa is ~53,090.
Table 4-2 Top 10 cities by built-up area exposed to flooding (10-year flood hazard zone).
Built up area in the 10 year flood zone
Rainwater
City Country floods City Country Coastal floods City Country Total floods
Accra Ghana 68.5 Lagos Nigeria 64.8 Lagos Nigeria 128.1
Lagos Nigeria 63.3 Conakry Guinea 36.6 Dar es Salaam Tanzania 79.4
Cape Town South Africa 57.4 Abidjan Côte d’Ivoire 34.4 Accra Ghana 76.8
Monrovia Liberia 56.4 Dar es Salaam Tanzania 33.5 Monrovia Liberia 71.4
Dar es Salaam Tanzania 45.9 Beira Mozambique 30.2 Conakry Guinea 66.8
Luanda Angola 37.0 Colonou Benin 30.1 Cape Town South Africa 65.7
Durban South Africa 33.9 Nouakchott Mauritania 28.3 Luanda Angola 59.8
Conakry Guinea 30.1 Dakar Senegal 24.3 Abidjan Côte d’Ivoire 56.1
Somerset W. South Africa 21.9 Luanda Angola 22.7 Durban South Africa 49.0
Abidjan Côte d’Ivoire 21.7 Maputo Mozambique 19.2 Colonou Benin 41.7
Note: (*) According to the original GHSL, Somerset West refers to Somerset, Strand and Gordon Bay, under one unique city boundary.
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | 21
Table 4-3 Critical infrastructure exposed to flooding
(10-year flood hazard zone) in the 20 most populated cities.
Top 10 cities with more people and infrastructure in the 10 year flood zone
*present infrastructure in OSM
Coastal flooding
City Country Km of roads Km of railroads Schools Hospitals
Lagos Nigeria 54 2 12 3
Luanda Angola 7 0 1 –
Dar es Salaam Tanzania 37 3 48 8
Abidjan Côte d’Ivoire 5 0 4 1
Accra Ghana 0 – – –
Cape Town South Africa 0 0 – –
Dakar Senegal 43 11 54 4
Doula Cameroon 1 1 4 –
Durban South Africa 10 49 – –
Conakry Guinea 49 11 76 7
Maputo Mozambique 18 5 – –
Lomé Togo 35 – 26 2
Cotonou Benin 50 12 32 6
Mogadishu Somalia 8 – – 1
Monrovia Liberia 5 – 3 1
Freetown Sierra Leone 6 – 5 1
Port Harcourt Nigeria – – – –
Nouakchott Mauritania 21 – 9 2
Mombassa Kenya 3 13 1 1
Massawa Eritrea 0 0 – 1
Top 10 cities with infrastructure in the 10 year flood zone
*present infrastructure in OSM
Coastal flooding
City Country Km of roads Km of railroads Schools Hospitals
Lagos Nigeria 47 – 6 2
Luanda Angola 41 3 8 2
Dar es Salaam Tanzania 28 2 48 9
Abidjan Côte d’Ivoire 24 4 17 –
Accra Ghana 63 9 23 3
Cape Town South Africa 157 66 89 5
Dakar Senegal 6 6 7 –
Doula Cameroon 13 6 2 4
Durban South Africa 43 186 3 1
Conakry Guinea 26 9 23 3
Maputo Mozambique 8 3 – 1
Lomé Togo 12 4 6 –
Cotonou Benin 8 1 7 –
Mogadishu Somalia 2 – 1 –
Monrovia Liberia 20 2 7 2
Freetown Sierra Leone 13 – 12 2
Port Harcourt Nigeria 19 – 3 1
Nouakchott Mauritania – – – –
Mombassa Kenya 10 27 – 2
Massawa Eritrea 2 1 1 –
22 | Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa
Figure 4-2 Flood map for Lagos, Nigeria.
Rainwater flood hazard zone (10-year return period) and key infrastructure assets.
Lagos, Nigeria
Coastal flood hazard zone (10-year return period) and key infrastructure assets.
Lagos, Nigeria
Note: Flood hazard maps are available for each city from the World Bank.
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | 23
Figure 4-3 Flood maps in Luanda, Angola.
Rainwater flood hazard zone (10-year return period) and key infrastructure assets.
Luanda, Angola
Coastal flood hazard zone (10-year return period) and key infrastructure assets.
Luanda, Angola
Note: Flood hazard maps are available for each city from the World Bank.
24 | Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa
Economic Damages city (box 4-2). The average proportion of risk across
SSA from coastal flooding compared to rainwater flood
Across SSA, US$204 million per year of urban damages is 40 percent. However, countries and cities
infrastructure are estimated to be at risk from costal
have different risks to coastal and rainwater flood
and rainwater floods.2 Flood risks in Angola, Ghana,
hazards (figure 4-4). For example, Nigeria is the country
Nigeria, and South Africa exceed US$20 million per year
with largest annual average damage (AAD) to coastal
(figure 4-4).
flooding, and second to rainwater flooding. South
Most of the flood risk is concentrated in large cities. Africa’s AAD, however, is driven mainly by rainwater
The cities at highest risk are, in order of magnitude: flooding. At the city level, differences in risk also present
Lagos in Nigeria, Accra in Ghana, Cape Town in South by type of flood hazard. Accra in Ghana and Cape
Africa, Luanda in Angola, and Durban in South Africa Town and Durban in South Africa experience rainwater
(figure 4-4, map 4-1, and appendix). These five cities flooding, which represents the largest fraction in the
exceed US$10 million per year, considering costs of both total flood risk, but in other cities like Beira, Dar es
hazards. Salaam, or Luanda, coastal flooding is more dominant.
Understanding the relative contribution of each hazard The contribution of each hazard and the total AAD for
is important to identify effective solutions, as the the 40 cities most at risk in SSA is outlined in appendix
interventions to address each risk may differ in each A, table A-3.
Figure 4-4 Annual average damage from flood hazards.
50
50
Annual
Annual Average
Average Damage
Damage
Millions
45
45
n Coastal n Rainwater
40
40 Coastal Rainwater
35
35
30
30
Millions
25
25
20
20
15
15
10
10
55
0
0–
r
ria
a
a
re
na
a
e
ea
l
rs
ga
ca
qu
ic
ol
ni
he
oi
in
a
ge
r
g
ne
za
as
bi
Iv
Af
Gh
Gu
An
Ot
Ni
n
am
d’
Se
ag
h
Ta
te
ut
ad
oz
Cô
So
M
M
Note: Values represent flood damages in US$ million (2015). The top 10 countries present US$170
million at risk per year. Other countries in Sub-Saharan Africa are aggregated together under
‘Others’ and represent US$29 million per year at risk.
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | 25
Average Annual Damage (US$)
<1M
1M - 1.5M
1.5M - 4M
4M - 7M
7M - 10M
>10M
Note:
26 | Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa
BOX 4-2 Local risk analysis.
The need for local risk analyses Example 1: Integrated urban and coastal
The maps and results presented in this report are available
resilience in Gambia
for each coastal city in SSA. A regional overview of flood The World Bank is developing local risk analysis in many
risk in annual damages can be a valuable metric to compare cities of SSA. For example, in Banjul, Gambia, the Gambia
and benchmark different countries, but it becomes difficult Integrated Urban and Coastal Resilience Program (P172822)
to assess the uncertainty in the flood damage estimates has developed of flood and coastal hazards to characterize
at the local level and against past events. The flood hazard present-day and future risk and identify priority risk
mapping relies on regional data and does not factor detailed mitigation measures.
local exposure data (e.g., building footprint) or other local
features such as flood management measures (e.g., drainage The local risk analysis uses historic and future climate
systems) that influence flood damages. For these reasons, conditions and a local elevation model to model flooding,
the flood risk values should be considered representative validated and corrected with field surveys. Spatial land use
across the region and used to compare countries and data and distribution of population and buildings are used
cities, but local city values can be considered as a first to infer flood damages and map risk (box figure 4-2.1). The
approximation that should be improved with local higher results identify the priority areas at highest risk of pluvial and
precision risk assessments. coastal flooding. The approach also includes historic analysis
of shoreline changes from remote sensing that is used to
derive shoreline retreat projections (see box figure.4-2.2).
Box figure 4-2.1. Flood risk in Banjul city, Gambia, accounting for the effects of climate change.
Legend
Urban risk 2040
[$m2/year]
n <1
n1–5
n 5 – 20
n >20
Critical Facilities
n City Councel
n Fire Station
n Hospital
n Market
n Police/Armed Forces
n Power Substation
n School
n Water Supply
Essential access routes paved
Essential access routes unpaved
Source: Flood and Coastal Risk Assessment and Priority Investment Planning for Greater Banjul, World Bank.
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | 27
BOX 4-2 Local risk analysis (cont.).
Box figure 4-2.2. Erosion hazard map for a coastal section in Banjul, Gambia.
Legend
Reference coastline
position (January 2020)
Erosion zone 20 years
(permanent erosion)
Erosion zone 20 years
(permanent erosion +
Temporary storm erosion)
Erosion zone 50 years
(Permanent erosion)
Erosion zone 50 years
(Permanent erosion +
Temporary storm erosion)
Source: Flood and Coastal Risk Assessment and Priority Investment Planning for Greater Banjul,
World Bank.
Example 2: Emergency Response and Resilience Box photo 4.21. Coastal erosion along the coast
in Saint Louis, Senegal of Saint Louis, Senegal
Another coastal city of SSA at risk is Saint Louis in Senegal.
The historical city of Saint-Louis—registered as a World
Heritage Site by UNESCO in 2000—is located on the
northwest coast of Senegal at the mouth of the Senegal
River. With more than 230,000 residents in 2017, the city
has experienced rapid urban growth over the last 50 years,
and population trends continue to increase due to high
urbanization rates. The economy of Saint Louis is mainly
driven by tourism, fishing, agriculture, and other commercial
and industrial activities such as sugar production.
Often referred to as the “Venice of Africa”, Saint-Louis is
one of the areas in Africa most threatened by rising sea
levels. While its geophysical characteristics of low lying
topography and the wind and wave climate render this
shoreline susceptible to flooding and erosion, the city’s
progressive encroachment upon the Langue de Barbarie
as well as climate change impacts have significantly
exacerbated these hazards. In recent years, coastal erosion
has accelerated between 5 and 6 meters of beach loss per
year. Moreover, unplanned settlement has contributed to
the degradation of the coastal system (box photo 4-2.1). The
city is also vulnerable to increasingly frequent urban floods
aggravated by poor land use planning, inadequate urban
governance, lack of sufficient drainage infrastructure, and
poor waste management. Source: World Bank.
28 | Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa
BOX 4-2 Local risk analysis (cont.).
In the Government of Senegal’s priority development plan of permanent dwelling units—and support for livelihood
and vision for 2035 (namely the Plan Senegal Emergent), restoration; and (iii) support for long-term development
Saint Louis is expected to play an important role in driving planning tools, including detailed flood and coastal modeling
the territorial development of the Senegal North region. studies, coastal protection works, and urban and drainage
However, this potential growth is threatened by the economic masterplans.
impacts of climate change combined with unsustainable
Using local elevation models and information on wave and
development practices. A World Bank study (2013) assessed
sea level conditions, local coastal flood risk models quantify
the main damage from future climate change in Saint Louis
flooding from storms enabling to simulate the overtopping
as having a net present cost of US$1.66 billion by the 2080s.
of the encroaching shoreline and the impact on large areas
The Saint Louis Emergency Response and Resilience Project of the city’s seafront (box figure 4-2.3). Similarly, river flood
(SERRP) aims to respond to the immediate, medium- and models are able to replicate the river flooding during historic
long-term needs of the city. Specifically, this entails: (i) events (box figure 4-2.4). These local results are used to
providing temporary accommodation to families displaced by determine the magnitude and location of economic damages
coastal erosion; (ii) the planned relocation of approximately and people affected. The risk reduction investments can
15,000 people in fishing communities living on the edge of be included in the models to determine the reduction in
the coastline in the high risk area—including development of damages and identify priority interventions, such as cost
a new relocation site situated inland in a safe area, provision benefit analyses.
Box figure 4-2.3. Analysis of extreme coastal flooding (10-year return period) for the city of Saint Louis, Senegal.
0.5
1775 1774
1776 0.45
1774.5
1773.5 0.4
1775 1774
0.35
Maximum inundation depth [m]
1773.5
1774 1773 0.3
y [km] in UTM 28N
y [km] in UTM 28N
y [km] in UTM 28N
1773
0.25
1773
1772.5 1772.5 0.2
1772 1772
0.15
1772
1771.5 0.1
1771
1771 0.05
1770 1771.5
0
338 338.5 339 338 338.5 339 338.4 338.6 338.8
x [km] in UTM 28N x [km] in UTM 28N x [km] in UTM 28N
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | 29
BOX 4-2 Local risk analysis (cont.).
Box figure 4-2.4. River flooding in Saint Louis for a historic flood event, modeled (left) and satellite imagery (right)
Source: Emergency Response and Resilience in Saint Louis, Senegal, World Bank.
Reference
World Bank. 2013a. Economic and Spatial Study of the
Vulnerability and Adaptation to Climate Change of Coastal
Areas in Senegal. Washington, D.C.
Notes
1. Although this analysis does not make distinction between
private (residential) or public infrastructure, it is likely that
ownership would also be an important factor in how risk
management actions are prioritized
2. Flood risk was calculated as Average Annual Damage
(AAD) by integrating flood damages associated to different
return periods (see chapter 2 on Methods). This method
relies on flood hazard zone defined for certain probabilities,
or return periods, and may differ from a stochastic
simulation of flood damages.
30 | Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa
Effects of climate change
and sea level rise
Climate change will likely increase the frequency rise include: Mkpanak in Nigeria, Tombua in Angola, and
and magnitude of extreme precipitation events and Okirka in Nigeria. However, when accounting for cities
sea levels in SSA. These effects, associated with that have at present day more than 100,000 people in
continued rapid urban expansion in flood-prone the 10-year flood zone—Cotonou in Benin, Djibouti in
areas will exacerbate further flood risk.1 In addition Djibouti, Douala in Cameroon, Maputo in Mozambique,
to demographic concentration and coastal urban and Saint Louis in Senegal—will see the largest increases
development, a comparison of the people exposed to the in populations flooded from the effect of sea level rise.
10-year coastal flooding in the years 2015 and 2050 These results indicate that these cities and also others
shows that sea level rise would represent a multiplier of with high exposure to coastal flooding should carefully
present-day flood risk. factor in future urban development how flood-hazard
zones will change with local sea level rise.
Sea level rise will expose more people to coastal flooding
in the absence of any adaptation measure. Eight
countries will see increases of over 10% in the population
exposed to the 10-year flood zone. It is estimated that
the countries with most people affected by sea level rise
will be (map 5-1): Equatorial Guinea and Eritrea with
more than 17 percent people flooded; and Benin and
Cameroon with greater than 15 and 14 percent of the
population affected respectively. However, in economic
terms, it is estimated that Cote D’Ivoire Liberia, and
Mauritania will increase flood damages on a 10-year
return period by more than 50% from the effect of sea
level rise. These countries are followed by Senegal at
more than 73 percent and Sierra Leone at 55 percent
or more. The rest of the SSA countries will increase the
10-year flood damages in the range of 10 to 20 percent
in general by mid-century from the effect of rising sea
levels. 50%
Regionally, the contribution of sea level rise will increase
Increase in Risk
coastal flooding, although the magnitude varies
between cities and countries. In 28 cities, sea level rise
2%
will increase coastal flood damages by more than 10
percent by 2050. Cities with the largest proportional
increase of people exposed to flooding from sea level
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | 31
200% 17%
1990–2015 (up to 2050)
-6% 0%
Note:
A comparison of population exposed to many low-lying areas, in contrast with episodic flooding
flooding from historic urban growth with effects. Preparing for sea level rise requires a call for
future sea level rise action to plan comprehensively around urban planning
and not only present-day hazards but also future flood
Historic urban expansion has substantially increased hazards. Climate change could turn previously suitable
risk in many coastal cities of Sub-Saharan Africa. In areas into new areas at risk, hence authorities need to
comparable timelines, population exposed in the 10-year carefully consider sea level rise in future urban planning.
flood zone increased from 1990 to 2015 at annual rates
greater of 10 percent in many countries and in many
countries more than double the initial number of people Notes
exposed to flooding (5.2 - left). However, the effect of
1. In this study, the exposure of urban population and assets
sea-level rise by the year 2050 will represent an increase
was calculated for 10-year coastal flooding events in 2015
of up to 17% in the people exposed in the future 10-year and 2050 using sea level rise projections (Representative
flood zone (map 5.2-right) (note: assuming only the Concentration Pathway 8.5), but no demographic growth
effect of sea-level rise and no increases in population). and increase of built-up area (Kopp et al. 2017). Therefore,
the results provided in this study should be considered
The only exception is Benin, being one of the countries
conservative and a lower end of the range of impacts of
most affected by sea level rise and where population sea level rise.
growth has increased only by four percent than what
References
the sea level rise will impact at 17 percent.
Kopp, R.E., DeConto, R.M., Bader, D.A., Hay, C.C., Horton,
Therefore, data show that historic urban expansion in R.M., Kulp, S., Oppenheimer, M., Pollard, D., and Strauss, B.H.
flood-prone areas presents a significant driver of risk,
2017. Evolving Understanding of Antarctic Ice-Sheet Physics
even when compared to sea level rise. However, sea and Ambiguity in Probabilistic Sea Level Projections. Earth’s
level rise would represent permanent submergence of Future. 5, 1217–233.
32 | Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa
From assessing risks to building
resilience: Resilient coastal cities
in Sub-Saharan Africa
Leveling the field for understanding and and infrastructure, jeopardizing sustainable economic
managing coastal risk growth. The historic view of past trends helps explain
the rising risks of floods in the coastal cities and
The past has shown that a resilient city is one that economies, but it should also help inform future action.
assesses, then plans, and acts to prepare for and This knowledge can now be used to raise awareness,
respond to hazards of any kind. This report levels the drive targeted action to address these factors and
field for understanding risk—in absence of better local catalyze risk-informed development plan investments
information—across all coastal cities in the region and to build climate resilience and adaptation.
sets a foundation to start addressing the risks. Unless
risks are assessed, managed, and planned for, and The data and analysis developed for this report are
governments and other stakeholders engage together available for countries and stakeholders at the city level
to address such risks, communities will stay vulnerable so that flooding can be factored in plans. The analysis
and less resilient to new shocks of any kind—natural also relies on open source and cloud computing to allow
disasters and pandemics may always come, but cities future advances as better information and data become
can be prepared. available. The approach can also help develop dynamic
assessments of risk and solutions at the national and
An understanding of urban flood risk is an initial step city scales.
to mainstream resilience into development. This report
provides a better understanding of risk in the coastal
cities of SSA. Comprehensive flood information for Reducing and managing flood risk
coastal urban areas is often lacking given the difficulties
In areas at risk, strategies for adaptation could include,
to model low-lying areas at regional scales. The SSA
among others: restrained development in flood-prone
region is particularly scarce and heterogenous in data,
areas, risk reduction projects, sustainable shoreline
which has remained as one of challenges hindering
management, planned retreat and/or realignment or
efforts in disaster risk management. The urban risk
both, and diversification from activities vulnerable to
profiles in this study provide a baseline view of the risk
climate change. In view of the rapid urbanization in the
of flood hazards and demonstrate how flooding has
region, solutions should also focus on risk-based plan-
represented a growing cost to communities, cities, and
ning to develop outside flood-prone zones. An under-
governments across SSA. Therefore, the next steps
standing of the areas at highest risk or where most peo-
should be to translate this risk information into action
ple can be protected, can further support government
and partnerships to charter the path for sustainable
decision making by allowing prioritization and adequate
coastal cities.
targeting of investments in risk mitigation. Overall, a
This report pinpoints the singular challenges driving risk-informed baseline is essential for governments to
flood risk in coastal SSA: rapid urban development, flood make better policies and for institutions to make sup-
hazards, and population increasingly concentrated in port more effective.
coastal zones. These factors compound risk to people
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | 33
Compound risks, challenges and and economic goals1, which could pay off across an
opportunities for resilience pathways economy.
COVID-19 has brought an additional layer of challenges ● Fourth, supporting the blue economy is also a
to consider, in addition to climate hazards in Africa. promising opportunity that coastal cities in SSA can
While natural disasters represent growing costs, the leverage. Through integrated management of the
pandemic has taken a large toll on the economic growth coastal zones and a sustainable use of resources,
of SSA, putting a decade of hard-won economic progress cities and countries can not only build resilience
at risk (World Bank 2020a). The pandemic is pushing against hazard risks but also contribute to the
the region into its first recession in 25 years. In 2020, economic recovery and long-term growth, by creating
GDP per capita was expected to contract by 6.5 percent economic and job opportunities.
in SSA and by the end of 2021, it is likely to regress to its
2007 level. The region will rebound, but growth will vary
A plan for climate action in Sub Saharan
across countries. The road to recovery will be long and
Africa
steep and must be paved with sound economic policies
and targeted planning and use of resources. Compared to other regions, and with few exceptions,
cities in SSA remain poorly equipped to adapt to the
This will occur as the threat of natural hazards, climate
changing climate given the levels of poverty, informality,
change, and increasing urban exposure continues to
rapid demographic growth, poor connectivity, and
loom. The rising sea levels and other effects of climate
financial capacity, among other factors. Yet, adaptation
change such as changes in rainfall events will also
and climate resilience must be the cornerstone of climate
threaten cities that are increasingly more populated
action for SSA countries to achieve their own sustainable
and that continue expanding into flood-prone areas.
development goals.
Targeted, decisive, and risk-informed action for
resilience building is urgent and necessary so that risk Overall, as coastal communities in SSA are increasingly
is not added as countries developed, and future shocks threat to lose ground to the oceans and livelihoods
will not roll back development gains made in the past to coastal erosions, flooding and other shocks,
decades. understanding and managing current and future risks
Building resilience in African cities is more critical than can provide the much-needed climate adaptation and
ever as the region moves forward during and post the resilience.
COVID-19 pandemic. The prevailing context increases
the need to discover and implement innovative The World Bank is committed to help build
opportunities to build back better, which could include resilience and sustainable growth
several actions.
The Climate Change Action Plan (CCAP) 2021–2025
● First, African cities will need to be strategic about how
aims to advance the climate change aspects of the
to align efforts for being resilient at all fronts. This will
WBG’s Green, Resilient, and Inclusive Development
also require reconciling short-term needs with long-
(GRID) approach, which pursues poverty eradication
term resilience and sustainability goals.
and shared prosperity with a sustainability lens. In the
● Second, making sustainable progress also demands Action Plan, the WBG support countries and private
pooling and combining efforts from governments, the sector clients to maximize the impact of climate finance,
private sector, and development partners. aiming for measurable improvements in adaptation and
resilience and measurable reductions in GHG emissions.
● Third, aligning the climate resilience, environmental,
The new Action Plan represents a shift from efforts to
and recovery agendas will be paramount, and can
“green” projects, to greening entire economies, and from
present unique opportunities for investing in measures
focusing on inputs, to focusing on impacts.
such as nature-based approaches that can not only
build resilience but also offer other environmental
34 | Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa
As part of this new approach, the World Bank is fully World Bank, Washington, DC. © World Bank. https://
engage in supporting SSA countries to build more openknowledge.worldbank.org/handle/10986/35754.
License: CC BY 3.0 IGO.
green, resilient, inclusive and productive cities, including
in coastal areas, though integrated and multisectoral World Bank. 2020a. Proposed Sustainability Checklist
for Assessing Economic Recovery Interventions April
investments (box 6-1). The first Africa Climate Business 2020. Washington D.C. https://pubdocs.worldbank.org/
Plan delivered 346 projects between 2015 and June en/223671586803837686/Sustainability-Checklist-for-
2020 for a total of US$33 billion. The Next Generation Assessing-Economic-Recovery-Investments-April-2020.pdf
Africa Climate Business Plan (World Bank, 2020c) World Bank, 2020b. Proposed Sustainability Checklist for
underscores the importance of pursuing climate-smart Assessing Economic Recovery Interventions. Washington D.C.
https://pubdocs.worldbank.org/en/223671586803837686/
urban transitions and provides a blueprint to help SSA
Sustainability-Checklist-for-Assessing-Economic-Recovery-
economies achieve low carbon and climate-resilient Investments-April-2020.pdf
development. World Bank 2020c. The Next Generation Africa Climate
Business Plan. Ramping up development-centered
This means ramping up support on adaptation to
climate action. Washington D.C. https://openknowledge.
help countries and private sector clients prepare for worldbank.org/bitstream/handle/10986/34098/34098ov.
and adapt to climate change while pursuing broader pdf?sequence=26&isAllowed=y
development objectives through a Green, Resilient and World Bank Group. 2021. World Bank Group Climate Change
Inclusive Development (GRID)approach. Action Plan 2021–2025 : Supporting Green, Resilient,
and Inclusive Development. World Bank, Washington, DC.
Notes © World Bank. https://openknowledge.worldbank.org/
handle/10986/35799 License: CC BY 3.0 IGO. https://
1. https://naturebasedsolutions.org/ openknowledge.worldbank.org/handle/10986/35799
References
Browder, Greg; Nunez Sanchez, Ana; Jongman, Brenden;
Engle, Nathan; van Beek, Eelco; Castera Errea, Melissa;
Hodgson, Stephen. 2021. An EPIC Response : Innovative
Governance for Flood and Drought Risk Management.
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | 35
BOX 6-1 Investments in a resilient coast.
Enabling resilient West African Coasts: also nature-based solutions such as restoring mangroves
The West Africa Coastal Areas Management or replenishing beaches, allowing for sustainable support of
Program (WACA) biodiversity with local sector benefits and lower costs. WACA
also supports income-generating activities and job creation,
West Africa’s coastal areas host about one-third of the involving not just public but also private partners. Helping
region’s population and generate 56 percent of its GDP.
develop public–private partnerships across various sectors,
The region is, however, exposed to enormous risks. Poor
including tourism or waste management, is an anchor for job
infrastructure planning, limited environmental governance,
creation and spillover effects and inclusion.
and human-induced pollution linked with population increases
and an overexploitation of coastal natural resources have Building Coastal Resilience: The case of Beira,
led to rapid degradation of the region’s coastal ecosystems. Mozambique
Climate impacts in coastal West Africa have already
Mozambique is one of the African countries most exposed
resulted in large financial costs and loss of human lives in the
to coastal and river flooding and home to some of the cities
past. In Gbekon, Benin, for example, the usual summer floods
from the local Mono River have been made worse over time most at risk within the continent. Its largest coastal city,
through coastal erosion and increasingly more unpredictable Beira, with a half a million habitants, is considered the city
rainfall, causing livelihoods and lives to be at risk. most threatened by climate change in SSA due to its coastal
location and its vulnerable infrastructure and population.
If unaddressed, these challenges can roll back West Africa’s Across the city, many neighborhoods are characterized by high
development gains and hinder its future growth aspirations. population densities, insufficient infrastructure, poorly planned
Interventions are needed to decrease the vulnerability of the settlements, high poverty rates, and lack of waste and storm
coastal communities, but preventing erosion in one country water drainage systems among others, enhancing the city’s
alone can lead to erosion in another, and the financing and vulnerability. For many years, storms and recurrent flooding
technical needs are too great for just one country to manage have hit and devastated the city, leaving residents vulnerable
effectively. to climate-related disasters. In 2020, two major cyclones, Idai
and Kenneth, had a catastrophic impact in Beira.
It is within this context that the West Africa Coastal Areas
Management Program (WACA), established in 2015 by the In the past, the World Bank has been actively engaged in
World Bank, aims to strengthen the resilience of communities emergency recovery and response work in Mozambique.
in coastal West Africa through a regional approach, The World Bank has provided support on preventive
supporting countries’ efforts in improving the management operations in the city and regional levels to work to address
of shared coastal resources and reducing natural and man- the climate change issues and future risks. The US$120
made risks. The WACA Program is a multi-country and million Mozambique Cities and Climate Change Project
regional response to support the strengthening of resilience has provided the city with the financial and technical
of coastal communities and assets in 17 western African assistance to strengthen its floodwater management
countries particularly vulnerable to erosion, flooding, and holistically and strengthen its resilience to weather-related
pollution. Activities include national resilience investment hazards. This included among others, the rehabilitation
projects, regional integration and support activities as well of the storm water drainage system, reducing flood risk
as a WACA platform to scale up and coordinate political
by 70 percent, and nonstructural interventions such as a
dialogue, mobilize public and private financing, and foster
waste collection program within the river’s neighboring
knowledge creation and sharing. In Benin, the WACA program
informal settlement. The project also includes the support
in 2020, for example, supported works to build dikes and
of nature-based solutions in combination with more
added measures to manage river flows and prevent flooding.
traditional infrastructure and includes the city’s investment
As a result, more than 3,600 households were less exposed
in a 17-hectare, multifunctional urban green park along the
to coastal erosion and flooding.
Chiveve river that represents one of the first nature-based
The WACA program demonstrates that building a resilient, urban flood management projects in Africa. While restoring
inclusive, and green future in African cities is possible. the river’s ability to mitigate floods, the project provides
WACA‘s approach to coastal resilience is multidimensional: additional recreational and economic opportunities for the
a mix of technical assistance and investments to preserve city’s residents. Lessons learned from this project bringing
the natural coastal resources, simultaneously spur together grey and green interventions for coastal cities
economic development, and enhance welfare and growth in supporting resilience have already been used to scale
of key sectors. For example, the program not only supports up support in other cities in Mozambique and can serve as
traditional infrastructure works for coastal resilience but guidance for cities across Africa.
Cape Twon, South Africa. Photo: © Hbh | Dreamstime.com
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | 37
Future advancements in regional
coastal resilience analyses
Remote sensing and new technologies, including of results, while also offering: (i) flexible automatization
cloud computing, have opened new opportunities to and (ii) future updates that could incorporate improved
understand and characterize hazard risks at regional datasets or methods. The approach also enables
scales in consistent and comprehensive ways. In this replicability of the analysis in different regions of the
analysis, consistent and homogenous datasets produced world. With the aim to support regional understanding
from remote sensing information and numerical flood of risks, this analysis was designed with the ambition to
and elevation models made this study possible. The allow improvements from the regional to the local scale.
flood maps and risk outputs are now available to
In the future, recent initiatives—some under
countries and cities with the best available, regionwide
development (see appendix B)—will permit addressing
data to date. However, such regional scope may present
some of the uncertainties at the local level, including:
instances of inaccurate results and limitations at local
(i) higher resolution estimates of urban expansion
scale, for example, derived from differences between
and population estimates for example, Atlas of Urban
city boundaries in GHSL compared with administrative
Expansion or World Settlement Footprint 2015; (ii)
boundaries at the local level. A section in Appendix B
coastal erosion; (iii) future changes in extreme rainfall
dedicated to the Methods discusses the main caveats
and their effects on flood hazard zones; (iv) local flood
and limitations found in the study.
modeling, accounting for local features, such as flood
This study was also built on technology that allows water management measures in drainage systems; (v)
efficient replication based on new datasets and future local and regional data on critical infrastructure; (vi)
improvements as they become available. Technological effects on flooding of the loss of ecosystems and the
advances to improve the assessment of coastal risk natural capital in urban areas; and (vii) assessment
dynamically in cities and landscapes are rapidly changing of adaptation strategies for coastal or rainwater
and allow for new insights to inform development. This flooding and their cost-effectiveness, such as using
analysis was developed in the Google Earth Engine representative effectiveness and cost estimates based
environment to exploit the power of cloud computing, on implemented projects.
which allows fast computation and easy visualization
38 | Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa
Appendix A. Extended results
Exposure by countries by flood hazard type
Table A-1 Population in the 10-year flood hazard zone in each country, by type of flood hazard.
Population in the rainwater Population in the coastal
10-yr flood zone 10-yr flood zone Total flood exposure
Country 1975 1990 2000 2015 1975 1990 2000 2015 1975 1990 2000 2015
Nigeria 279,544 531,585 755,758 1,089,372 263,708 398,509 472,456 679,967 543,253 930,094 1,228,213 1,769,339
Senegal 75,729 105,252 151,282 144,382 357,691 473,830 621,657 772,307 433,421 579,082 772,940 916,689
Mozambique 75,119 143,467 168,858 181,831 480,540 611,729 688,748 691,744 555,658 755,195 857,606 873,575
Tanzania 58,948 122,495 213,321 438,296 47,082 92,661 161,348 315,772 106,030 215,155 374,669 754,068
Somalia 22,579 70,100 84,517 199,437 161,119 226,655 217,751 480,770 183,698 296,755 302,268 680,207
Angola 41,946 109,542 183,098 417,693 16,190 41,046 83,722 239,833 58,136 150,588 266,820 657,526
Madagascar 24,782 70,901 125,765 168,614 67,366 124,760 201,676 319,449 92,148 195,660 327,442 488,063
Ghana 74,026 160,248 275,752 369,226 28,240 53,125 86,689 92,042 102,267 213,372 362,441 461,268
Benin 24,447 34,214 45,570 47,740 372,168 370,469 383,533 385,791 396,615 404,683 429,103 433,531
Guinea 29,886 48,667 90,309 188,685 106,355 121,871 160,636 235,108 136,241 170,538 250,944 423,793
South Africa 173,005 279,933 369,455 377,794 22,359 34,868 44,179 32,646 195,364 314,802 413,634 410,440
Sierra Leone 62,636 103,060 117,832 197,558 65,831 104,461 119,472 208,647 128,467 207,521 237,304 406,205
Liberia 70,372 112,029 177,860 257,686 24,053 37,673 58,778 133,444 94,424 149,701 236,638 391,129
Côte d’Ivoire 93,083 184,389 247,560 260,324 25,393 48,905 65,443 125,573 118,476 233,294 313,002 385,897
Djibouti 13,750 32,050 43,056 31,783 71,190 159,567 194,868 265,965 84,940 191,617 237,924 297,748
Cameroon 3,030 9,928 15,370 34,590 23,141 48,944 77,837 188,977 26,171 58,872 93,207 223,567
Kenya 29,671 50,469 65,721 60,667 47,778 74,351 87,705 140,385 77,449 124,821 153,426 201,052
Togo 19,535 25,900 32,901 54,885 63,450 92,003 107,053 125,023 82,985 117,903 139,954 179,908
Gabon 27,620 36,169 37,517 45,002 60,742 54,238 54,015 122,276 88,361 90,408 91,533 167,278
Mauritania 1,986 4,005 6,088 3,378 22,624 52,793 84,586 138,052 24,610 56,798 90,674 141,430
Guinea-Bissau 671 1,001 3,234 1,891 8,701 12,405 25,840 52,301 9,372 13,407 29,075 54,192
Equatorial Guinea 2,545 4,165 9,619 32,687 7,279 8,379 13,583 20,168 9,824 12,544 23,202 52,855
The Gambia 1,858 8,613 17,636 28,300 8,024 18,120 27,574 23,044 9,883 26,733 45,209 51,344
Eritrea 13,325 14,038 13,470 20,838 23,427 23,544 22,130 30,096 36,752 37,583 35,600 50,934
Republic of the Congo 746 7,596 18,063 24,922 407 4,143 14,757 24,191 1,152 11,740 32,819 49,114
DRC 38,874 37,384 37,168 36,762 699 672 511 1,449 39,573 38,056 37,679 38,211
Namibia 573 1,309 2,363 4,854 316 522 830 3,398 889 1,831 3,193 8,252
Total Built-up area 1,260,286 2,308,510 3,309,142 4,719,198 2,375,872 3,290,244 4,077,377 5,848,417 3,636,158 5,598,754 7,386,519 10,567,615
Note: The countries are listed in decreasing order by total flood exposure. The blue bars represent the relative ranking for each flood hazard (in decreasing order
of total exposure).
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | 39
Table A-2 Built-up area exposed to the 10-year flood hazard in each country, by type of flood hazard.
Built-up area in the rainwater Built-up area in the coastal Total flood exposure
10-yr flood zone 10-yr flood zone
Country 1975 1990 2000 2015 1975 1990 2000 2015 1975 1990 2000 2015
Nigeria 13 48 76 145 7 18 25 57 20 65 102 202
South Africa 40 78 85 136 2 5 6 31 42 83 91 167
Angola 4 11 17 66 3 4 7 34 7 15 24 100
Ghana 15 33 47 85 1 2 3 14 16 34 50 100
Tanzania 9 10 16 59 14 14 18 38 23 25 34 97
Mozambique 1 2 4 29 4 19 23 62 5 21 27 90
Liberia 39 42 48 62 6 7 8 17 45 49 56 79
Guinea 3 6 8 33 14 16 19 36 17 22 27 69
Senegal 2 4 4 17 10 15 17 38 11 19 21 55
Côte d'Ivoire 22 23 26 38 3 3 4 11 26 27 30 50
Benin 2 4 6 14 12 14 15 23 14 19 21 37
Somalia 1 3 4 13 4 7 8 21 4 10 12 34
Sierra Leone 3 4 5 21 1 2 2 12 4 6 8 33
Togo 2 4 5 10 9 10 11 16 12 14 16 26
Gabon 1 3 4 8 1 3 4 10 1 6 8 19
Kenya 0 1 2 8 0 2 2 11 0 3 4 19
Cameroon 1 2 2 5 1 2 3 9 2 4 5 14
Mauritania 0 0 0 1 4 5 6 13 4 6 6 14
Equatorial Guinea 0 1 3 8 0 1 2 4 1 2 4 12
The Gambia 0 1 2 6 0 1 1 4 1 2 4 10
Republic of the Congo 1 1 1 5 0 0 1 3 1 1 2 8
Djibouti 0 0 0 1 0 0 0 5 0 0 0 7
Guinea-Bissau 0 0 0 1 0 1 2 5 0 1 2 6
DRC 1 1 1 4 0 0 0 0 1 1 1 4
Eritrea 0 0 0 2 0 0 0 2 1 1 1 4
Namibia 0 0 0 1 - 0 0 0 0 0 0 2
Madagascar 3 3 4 - 6 8 9 - 8 12 13 -
Total 161 284 369 779 97 152 187 478 258 437 556 1,257
Note: The countries are ordered in decreasing total flood exposure. The blue bars represent the relative ranking for each flood hazard in decreasing order of total
exposure.
40 | Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa
Flood damages in cities
Table A-3 Top 40 cities by annual average damage from flood hazards.
AAD AAD
City Name
City Country
Country coastal
AAD flood
Coastal flood rainwater
AAD flood
Rainwater flood Total
Total flood
flood
Lagos Nigeria $ 10,456,858 $ 19,887,405 $ 30,344,262
Accra Ghana $ 90,316 $ 17,238,834 $ 17,329,150
Cape Town South Africa $ 371,306 $ 15,991,367 $ 16,362,672
Luanda Angola $ 5,597,586 $ 9,826,826 $ 15,424,412
Durban South Africa $ 1,076,419 $ 12,788,724 $ 13,865,142
Dar es Salaam Tanzania $ 3,306,051 $ 3,846,259 $ 7,152,310
Conakry Guinea $ 3,917,017 $ 2,566,371 $ 6,483,388
Cotonou Benin $ 4,517,498 $ 1,167,380 $ 5,684,878
Dakar Senegal $ 3,635,476 $ 1,550,417 $ 5,185,893
Beira Mozambique $ 4,544,287 $ 140,132 $ 4,684,420
Abidjan Côte d'Ivoire $ 516,041 $ 3,644,449 $ 4,160,490
San-Pedro Côte d'Ivoire $ 6,799 $ 3,896,970 $ 3,903,768
Monrovia Liberia $ 484,470 $ 3,415,795 $ 3,900,265
Bata Equatorial Guinea $ 479,045 $ 3,109,733 $ 3,588,779
Port Elizabeth South Africa $ 454,499 $ 2,798,216 $ 3,252,715
Ikorodu Nigeria $ 111,104 $ 3,131,063 $ 3,242,167
Lomé Togo $ 1,856,193 $ 1,330,265 $ 3,186,458
Somerset West South Africa $ 42,195 $ 2,758,747 $ 2,800,942
Nouakchott Mauritania $ 2,066,285 $ 253,927 $ 2,320,213
Benguela Angola $ 70,478 $ 2,150,567 $ 2,221,046
Warri Nigeria $ 32,548 $ 2,171,815 $ 2,204,363
Freetown Sierra Leone $ 736,817 $ 1,451,316 $ 2,188,133
Libreville Gabon $ 961,436 $ 1,207,073 $ 2,168,508
Lobito Angola $ 168,458 $ 1,903,179 $ 2,071,637
Maputo Mozambique $ 985,616 $ 1,049,181 $ 2,034,797
Douala Cameroon $ 706,582 $ 1,077,070 $ 1,783,652
Kamsar Guinea $ 1,279,208 $ 140,904 $ 1,420,112
Saint-Louis Senegal $ 700,208 $ 635,919 $ 1,336,127
Pointe-Noire Republic of the Congo $ 119,863 $ 1,173,719 $ 1,293,582
Takoradi [Sekondi-Takoradi] Ghana $ 115,614 $ 1,150,069 $ 1,265,683
Bonny Nigeria $ 1,026,618 $ 218,045 $ 1,244,663
Port-Gentil Gabon $ 630,819 $ 565,675 $ 1,196,494
Mombasa Kenya $ 397,068 $ 609,161 $ 1,006,229
Zanzibar City Tanzania $ - $ 946,297 $ 946,297
Lakuwe Nigeria $ - $ 854,596 $ 854,596
Port Harcourt Nigeria $ 12,038 $ 823,569 $ 835,607
Badagry Nigeria $ 8,045 $ 806,531 $ 814,575
Aboisso Côte d'Ivoire $ - $ 806,964 $ 806,964
Calabar Nigeria $ - $ 794,031 $ 794,031
Wells Estate South Africa $ - $ 763,109 $ 763,109
Note: Values of AAD normalized by the total exposure of the city are provided in Appendix D. All $ values expressed are in US dollars.
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | 41
Appendix B. Methods and Data
Coastal flood hazard ● Joint Research Centre Global Surface Water
(CGSW): provides seasonal water maps from 1984
Coastal flooding results were obtained based on the to 2018 at 30-meter horizontal resolution based on
Surging Seas project from Climate Central1 using sea optical satellite observations (Landsat). Seasonal
level model results and a new global digital elevation water refers to each type of water surface that is not
model (DEM) for coastal zones or coastal DEM. This permanent, which can be associated to flooding. The
new coastal DEM uses neural networks to reduce dataset presents some limitations, however, such as
previous errors in other models. The elevation data not allowing the distinction between river, coastal, or
were used to calculate global flooding from extreme pluvial flooding; the repeating cycle of the satellite
sea levels and sea level rise. (Kulp and Strauss 2019). at 16 days does not allow for a complete observation
The elevation data were combined with the extreme of flood events; and the cloud coverage, especially
sea levels associated to the 1-, 10-, and 100-year return relevant in tropical areas, can prevent the observation
periods from the global tides and surge reanalysis (Muis of individual events. As a result, not all the floods
et al. 2016). Whereas the lower return periods were are detected by the CGSW and the ones that are
provided by Climate Central, the 100-year extreme sea captured may be underestimating the peak of the
levels were calculated from GSTR and combined with floods. For these reasons, a statistic, such as return
elevation maps. The flood population and built-up areas period, based on these data should be considered
were interpolated using the extreme sea level and the conservative (Pekel et al. 2016).
lower and upper elevation zones because the vertical
resolution of the model is one meter. The results for the These two datasets are complementary and were
1-year flood zone were discounted from the rest of the combined to determine the flood hazard zones for
calculations assuming that areas flooded on an annual each city. The satellite-based maps capture areas that
basis do not produce flood damages. have been flooded and, therefore, represent minimum
and historic flood extents or lower bound estimates
of floods. Meanwhile, the modeling results represent
River and rainwater flood hazard areas that could be flooded by more extreme events
or that were not captured in the historic period in the
The rainwater flood hazard maps were calculated
remote sensing data. Numerical modeling results are
combining global flood modeling results with remote
therefore more adequate to capture the maximum flood
sensing data. Each dataset is explained below.
extents associated with higher return periods or upper
● Fathom-Global flood maps: these global maps bound of flooded areas. The return periods of satellite-
provide flood depths at a horizontal resolution of based floods maps—pixel based for the whole area of
90 meters by modeling fluvial and pluvial flooding interest—were empirically calculated based on the
for several return periods. Although the precision of number of observations and occurrences such as: return
the maps is adequate for regional areas, its medium period (satellite) = years of observation per number of
resolution does not allow specific insights into local or occurrences.
flat areas (Sampson et al. 2015).
42 | Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa
The years of observation are 35 (from 1984 to 2018) disasters, including flash floods and landslides, have
and the number of occurrences is the number of times severely impacted cities across Africa, most recently
that water was detected per image pixel. The maximum in 2020. Understanding how these changes will modify
return period is therefore 35 years—water observed the existing rainwater floodplains, will be critical to
at least once in the period of observation—while the characterize flood risk better in coastal cities of Sub-
minimum return period can be one year, that is yearly Saharan Africa.
flooding. When occurrences equal zero, the return period
is not defined. The flood maps associated with the return
Exposure data
periods 1-, 10- and 35-years were combined with the
Fathom-Global model-based flood results for the same Population and built-up area: The built-up areas were
return periods and merged in extent. Because the global calculated based on the global human settlement layer
models do not provide the return period of 35 years, (GHSL), a global, multitemporal evolution of built-up
the 50-year results were used instead, and the flood surfaces derived from Landsat satellite data collections
map from satellites was assumed to have the higher organized in four epochs 1975, 1990, 2000, and 2014. This
return period, which can be considered a conservative dataset was selected because it provides information
approach. Flood maps are binary maps with value 1 from both urban built-up area as well as population,
wherever there is flood and zero elsewhere. The final using a consistent procedure. GHSL is the result of the
maps were calculated by selecting the maximum pixel reprocessing 33,202 Landsat images organized in four
value between the two datasets. data collections. The dataset was developed jointly
by the European Joint Research Centre (JRC) and the
Directorate General for Regional Development (DG
Effects of climate change on flood hazards
REGIO) to provide global evidence-based analytics and
In addition to sea level rise, changes in rainfall patterns knowledge describing the human presence in the planet
and the frequency and intensity of extreme rainfall (see figure B-1). The GHSL methods rely on automatic
events will be key factors to understand and plan spatial data mining technologies from large amount of
against future flood risk in the region. Rain-triggered data that include global, fine scale satellite imagery,
Figure B-1 Transition from imagery to built-up areas extraction (GHS-BU),
population modeling (GHS-POP), and settlements classification (GHS-SMOD).
Source: Joint Research Centre.
Note: Examples shown are in Bangkok, Thailand.
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | 43
census data, and volunteering geographic information well-supported examples of Volunteered Geographic
sources. The GHSL consists of three main information Information. However, the completeness of the
components hierarchically placed at three different information varies between countries and comparisons
levels of abstraction: global human settlement built-up between countries should be taken with caution.
areas (GHS-BU), the GHS population grids (GHS-POP)
and the GHS urban or rural classification model (GHS-
Damage vulnerability curves
SMOD).
A flood disaster can have two main effects on a society
The GHSL built-up areas correspond to areas where
and economy (GWP-WMO 2013): the total or partial
buildings can be found (Pesaresi et al. 2015). The
destruction of physical assets, resulting in subsequent
concept of buildings formalized by the GHSL are
changes or losses to economic flows in the affected
enclosed constructions above ground, which are
area, including the interruption of services. This report
intended or used for the shelter to produce economic
calculates the damages of coastal and rainwater
goods or referred to as constructed structures. The
flooding based on direct tangible losses, which are
GHSL also includes population in a grid at 250 meters
defined as those induced by the physical contact of
resolution. This information layer is derived from the
combination of global collections of national population flood waters with humans, property, or any other
census data and global built-up areas as extracted objects. Other losses induced by the direct impacts and
from Earth Observation data analytics. The GHSL occurring outside the flood event were not included in
dataset also includes an urban or rural classification the analysis such as transport disruption, business
model (GHS-SMOD), with less spatial detail (1 km), by losses, or loss of family income.1
combining the built-up and population information The damages expected to result at a specified depth
and following the degree of urbanization (DEGURBA) of flood water were calculated using a normalized
model that discriminates three settlement classes: (i) damage ratio curve, typically used for houses and
cities, (ii) towns and suburbs and (iii) rural areas—the other buildings. A review of damage functions—or
discrimination is based on the population density in the vulnerability functions—in Africa were developed by
square kilometer grid, total settlement population, and the Joint Research Centre based on literature and local
other spatial generalization parameters. The method values (Huizinga et al. 2017), which provides normalized
for delineation of urban and rural areas was made for damage functions (figure B-2). An estimated mean
international statistical comparison purposes, and was damage degree function was obtained based on the
developed by the European Commission, the Organisation residential buildings land use, assuming is the most
for Economic Co-operation and Development (OECD), widespread building type in the built-up urban area
the Food and Agriculture Organization of the United (figure B-2). Economic damages were calculated using
Nations (FAO), UN-Habitat and the World Bank. the damage ratios applied to maximum damage values
Urban centers: Urban centers are defined by specific per pixel. The maximum damage values were defined
cut-off values on resident population and built-up based on a review developed by the European Joint
surface share in a 1x1 kilometer global uniform grid. Research Centre. The maximum damage values per
The reference epoch for the spatial delineation of the built-up area were obtained by country, taking the
urban centers is 2015. The coastal cities were selected residential building types as a reference, and applying
as those urban centers, as defined in the GHSL-SMOD, conversion factors to consider the depreciated value
falling within 10 kilometers from the coastline. for the construction cost (0.6), and an undamageable
fraction of the buildings (0.4), following guidelines
Infrastructure data: Data from roads, railways, schools, for flood damage calculation (Huizinga et al. 2017).
and hospitals were obtained from OpenStreetMap. Correcting for the fraction of buildings per land use
This data source is heterogeneous across the various was not necessary because this analysis uses the most
cities and hence coverage and quality are variable. The recent version of the GHSL built-up area, which provides
OpenStreetMap project is one of the most popular and information of the percentage of buildings per pixel—
44 | Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa
this built-up percentage was not included in previous US$7.2 per square meter in Somalia. These values are in
versions. The resulting maximum damage values per the range of local estimates from analysis developed by
country varied between a maximum of US$34.3 per the World Bank for coastal cities in the region.2
square meter in Equatorial Guinea to a minimum of
Figure B-2 Normalized damage factor for Africa: Residential buildings and content.
1.20
Mozambique, urban house
Normalised damage factor
1.00
Mozambique, rural house
0.80
South Africa, small house
0.60 South Africa, medium house
0.40 South Africa, large house
Europe
0.20
Africa
0.00
0 2 4 6 8
Depth (m)
Source: Huizinga et al. 2017.
The annual average damage (AAD) was calculated by Population data:
integrating the flood damages across their probability
● The Urban Centre Database (R2019A) was used as
of occurrence as given in this equation:
a reference to analyze the evolution of cities in time
������������
1 1 1 from 1975 to 2015, regarding population and built-
AAD = ∑ ( − ) (������������������������ + ������������������������+1 ) up areas. In some instances, nonreliable values were
2 ������������������������ ������������������������+1
������������=1 found. For example, the city of Luanda had nearly 6.8
where i refers to the number of return periods (n), Ti is million people in 2015, but it appears in the dataset
the return period, Di represents the damages for the as 1,777 inhabitants in 1975; 58,125 in 1990; and
probability of 1/Ti for example, the flooding associated 488,036 in 2000. This indicates an underestimation
with a return period of 100-year has a probability of of historic values. Similarly, the city of Massawa,
occurrence of one percent in a given year. For rainwater a small town with a reported population of about
flooding, the return periods of 5-, 10-, 20-, and 50-years 50,000, appears in the database with more than
were used, whereas for coastal flood risk, the annual 400,000 people in 1975 and with a population greater
average damages were determined based on the 1-, 10-, than one million in 2015. Some values in the dataset
and 100-year flood hazards. These return periods were greatly underestimate or overestimate the population.
limited by the data available for each flood hazard. For this reason, the dataset was corrected for these
instances, as indicated in the tables and figures when
appropriate.
Uncertainty in the exposure data analysis
● The spatial distribution of population from GHSL-POP
This analysis was limited by the availability of consistent, was used to assess people affected by floods in the
robust, and homogeneous data on coastal exposure and years 1975, 1990, 2000 and 2015. The year 2015 has
flood hazard modeling. Based on the data used in the two versions of the dataset: an older version published
analysis, the main uncertainty factors identified in the in 2016 (https://data.jrc.ec.europa.eu/dataset/jrc-
analysis include: ghsl-ghs_pop_gpw4_globe_r2015a) and a newer
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | 45
version published in 2019 (GHS_POP_E2015_GLOBE_ boundaries. As a result, the reference and association
R2019A_54009_250_V1_0). The newest version was of these results by name to administrative boundaries
used because it is considered more reliable, although of cities should be taken with caveats, and are only
it presented some inconsistencies. For example, in representative of the city boundaries defined in the
the city of Massawa, the newer version provided, as dataset. For example, as noted in the results, the
occurred in the Urban Centre Database, more than polygon representing the city of Somerset West in
1 million people, whereas in the older version the South Africa also includes the settlements of Strand
population was nearly 47,000. In these cases, the and Gordon Bay, which correspond to different cities,
older version of GHSL_POP was used to correct the but are included as one single urban polygon in this
incorrect values in the most recent dataset. analysis. This presents large values for the city
boundary, but the small settlements should be taken
Built-up areas: individually, like Somerset West.
● The GHS-BUILT dataset includes built-up areas for the Future advances in hazard and exposure data can allow
years 1975, 1990, 2000 and 2015. For 2015, a newer improvements in the approach and could provide better
version of the dataset (GHS_BUILT_LDS2014_GLOBE_ or more precise estimates for understanding historic
R2018A_54009_250_V2_0) provides information and future changes in risk. Some ongoing initiatives
about the percentage of built-up per pixel, instead of include the modeling of exposure through new advances
a binary mask (0 or 1), as in the previous version. In the in Earth observation Routines For example, the United
most recent version, a value of 100 represents a total Kingdom Space Agency is developing Earth observation
built-up. This information allowed a more precise technologies (METEOR) to improve understanding of
estimation of the built-up areas at risk. However, this exposure with a specific focus on the countries of Nepal
can cause inconsistencies with historical data, where and Tanzania, as a validation set. METEOR will deliver
built-up areas could be overestimated compared to countrywide, openly available exposure data for the 48
the most recent version of the dataset. least developed Official Development Assistance (ODA)
recipient countries. National and local analyses can also
use “Digitize Africa” that contains building footprints
Urban centers:
and roads in many African countries. Similarly, the
● The cities selected for the study are based on the German Aerospace Center has recently developed the
urban centers in the GHS_SMOD dataset, which World Settlement Footprint Evaluation dataset, which
provides data on urban centers defined based on characterizes settlement growth from 1985 to 2015.
the DEGURBA methodology. Therefore, these urban A comparison of the preliminary results with the GHSL
centers do not coincide with real administrative used here can be found in Appendix C.
46 | Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa
Notes References
1. The damages calculated represent direct tangible losses, Kulp and Strauss 2019. Kulp, S.A., and Strauss, B.H. 2019.
that is, those induced by the physical contact of flood New elevation data triple estimates of global vulnerability to
waters with humans, property, or any other objects. The sea-level rise and coastal flooding. Nat. Commun. 10, 4844.
term “flood losses” on the other hand, refers to temporary
Muis, S., Verlaan, M., Winsemius, H.C., Aerts, J.C.J.H., and
changes in economic flows from the time of the flood
Ward, P.J. 2016. A global reanalysis of storm surges and
disaster until full economic recovery and reconstruction.
extreme sea levels. Nat. Commun., 7, 11969.
This report only considers direct flood damages to built-up
area. Other losses occurring (in space or time) outside the Huizinga, J., Moel, H. de, and Szewczyk, W. 2017. Global flood
flood event, such as transport disruption, business losses, depth-damage functions.
or loss of family income were not included in the analysis.
Pekel, J.-F., Cottam, A., Gorelick, N., and Belward, A.S. 2016.
Conceptually, it is also important to note the difference
High-resolution mapping of global surface water and its long-
between flood damages and flood losses. The term “flood
term changes. Nature, 540, 418–22.
damage” refers to the physical damage caused to public and
private assets such as infrastructure, houses, and vehicles Pesaresi, M., Ehrilch, D., Florczyk, A.J.., Freire, S., Julea, A.,
as a result of contact with floods. See: GFDRR, 2010: Kemper, T., Soille, P., and Syrris, V. 2015. GHS built-up grid,
Damage, Loss and Needs Assessment (DaLA): Guidance derived from Landsat, multitemporal (1975, 1990, 2000,
Notes. Global Facility for Disaster Risk Reduction, World 2014). European Commission, Joint Research Centre. http://
Bank. www.gfdrr.org/gfdrr/DaLA_Guidance_Notes data.europa.eu/89h/jrc-ghsl-ghs_built_ldsmt_globe_r2015b
2. Climate Central is a science and news organization that Sampson, C.C., Smith, A.M., Bates, P.D., Neal, J.C., Alfieri, L.,
bridges the scientific community and the public, providing and Freer, J.E. 2015. A high-resolution global flood hazard
clear information to help people make sound decisions model. Water Resour. Res., 51, 7358–381.
about the climate. https://www.climatecentral.org GWP-WMO. 2013. Integrated Flood Management Tools Series
3. The approach to calculate annual expected damages relies Conducting Flood Loss Assessments.
on flood hazard zones and their associated return periods
and, therefore, it could provide different results than a
stochastic simulation of flood events and damages at each
city. For example, the 10-year flood damage corresponds to
all assets in the flood zone with probability of occurrence
of 10% in a given year, which may be different than the
damage with probability 10% from all possible flood events.
Furthermore, the 10-year flood will not occur simultaneously
across the region, so the spatially aggregated estimates
should be taken as a statistic of flood risk.
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | 47
Appendix C. Sub-Saharan Africa’s cities
included in the Atlas of Urban Expansion
The present analysis relies on the global human in partnership with Google Earth Engine. The dataset
settlement layer as a consistent and updated dataset includes an excess of six million satellite images from
for both urban area and population change, at the 1985 to 2015—almost the complete archive of the
preparation of this study. However, the results were US Landsat missions—to characterize the worldwide
compared with two other datasets that characterize growth of human settlement on a year-by-year basis.
city population and urban area change: The DLR kindly gave access to this dataset to compare
and validate the analysis in this study, which relies on the
1. The World Settlement Footprint Evolution, developed
GHSL. Using the same city boundaries for all the cities
by The German Aerospace Center (DLR) (Marconcini
in this study, we computed the extent of built-up area
et al. 2020)their location and extent is still under
at each reference year in the common period between
debate. We provide here a new 10 m resolution (0.32
datasets: 1990, 2000, and 2015. The comparison for
arc sec.
selected cities of different extent and population is
2. The Atlas of Urban Expansion that provides historic provided in table C-1. The WSF dataset is more precise
information on urban growth and maps of urban area than the GHSL as it includes more remote sensing data
and density and blocks and roads for 200 cities.1 in its estimates. In the selected cities, the GHSL tends to
underestimate historic built-up area and overestimate
the most recent urban extent, especially in the smallest
Comparison of built-up area change with the
cities. However, the GHSL provides population to
DLR WSF-Evo
these urban areas using an equivalent process, so the
The German Aerospace Center (DLR) recently released comparison between population and urban extent rates
the “World Settlement Footprint Evolution” (WSF-Evo) is considered consistent.
Table C-1 Changes in built-up area for the World Settlement Footprint Evolution (WSF)
and the Global Human Settlement Layer (GHSL).
WSF built-up area [km2] GHSL built-up area [km2]
City Country 1990 2000 2015 1990 2000 2015
Accra Ghana 670 684 784 485 555 876
Dar Es Salaam Tanzania 285 385 470 196 282 654
Durban South Africa 405 470 540 391 448 785
Lagos Nigeria 575 680 930 684 889 1170
Luanda Angola 405 475 490 221 281 771
Maputo Mozambique 205 270 355 174 212 418
Massawa Eritrea 1.1 4 4.9 0.7 0.7 5.8
48 | Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa
A visual comparison between WSF-Evo and GHSL for the outskirts and sprawl areas, where the lack of detail
Accra, Lagos, and Luanda are represented spatially to or resolution can lead to over- or underestimation of
show that the two datasets are qualitative very similar results. However, the analysis is also sensitive to what
(figure C-1). However, they present slightly different total is considered built-up areas.
urban extent results. Spatially, the differences occur at
Figure C-1 Spatial comparison of urban footprint between WSF (left panels) and GHSL (right panels).
Accra, Ghana
a) WSF Evolution, 1985–2015 GHSL Built, 2015 b)
Lagos, Nigeria
c) WSF Evolution, 1985–2015 GHSL Built, 2015 d)
Luanda, Angola
e) WSF Evolution, 1985–2015 GHSL Built, 2015 f)
Note: (a, c, e): Spatial pattern of built-up area change from WSF. (b, d, f): Spatial pattern of built-up area change from GHSL.
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | 49
As a caveat to the historical information, the WSF City population from the Atlas of Urban
provides an analysis of the quality of the estimate Expansion
based on the availability of images, and indicates that
from the 1980s to the 1990s, the availability of images Based on the Atlas of Urban Expansion as of 2010, the
was not optimal. Therefore, the IDC score before year world contained 4,245 cities with 100,000 or more
2000 is low and consequently the detection of built-up people (map C-1). The NYU Urban Expansion Program
areas is more difficult and less precise. at the Marron Institute of Urban Management and the
Stern School of Business of New York University, in
In summary, the new DLR dataset seems to be more partnership with UN-Habitat and the Lincoln Institute
precise than that of the JRC’s. However, as a new dataset, of Land Policy, has initiated a multiphase research effort
it has not been extensively used yet. Furthermore, a to monitor the quantitative and qualitative aspects of
population map derived from this built-up area mask global urban expansion. The Atlas of Urban Expansion
has not been produced yet. Therefore, this study relies (Lincoln Institute of Land Policy 2012) has now
on the GHSL as it has been extensively used in the completed the data collection in a global representative
literature and keeps consistency with the population sample of 200 global cities with 100,000 people or
data. However, future improvements in the analysis more. The project focuses on entire metropolitan
should consider DLR’s WSF-Evo data, including updates areas—contiguous urban areas that may contain many
of the present analytical work that can leverage the municipalities are considered a single city.
cloud computing and implementation in Google Earth
Engine used in this work.
Note:
50 | Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa
Cities are defined by the extent of their built-up area, Expansion. The qualitative ranking agrees with the GHSL
rather than by their administrative or its jurisdictional results. However, this global atlas provides important
boundaries. The project has identified 4,245 cities on insight into the coastal cities in SSA compared with other
our planet with more than 100,000 people in 2010, regions (figure C-3). This comparison demonstrates
which totals 2.5 billion people, or 70 percent of the that coastal cities in SSA are not the most populated
world’s 2010 urban population of 3.6 billion. in the world, which would indicate the need to consider
smaller cities too, as in the GHSL (figures C-4 and C-5).
Figure C-2 shows the annual growth of the 18 SSA’s
cities that are currently included in the Atlas of Urban
Figure C-2 Urban extent and urban growth (per year) for the 18 cities from Sub-Saharan Africa.
Urban Extent Population (#) Annual Growth (%)
Nakuru 326,160 Nakuru
Arusha 377,169 Arusha
Beira 382,575 Beira
Gombe 416,327 Gombe
Ndola 443,327 Ndola
Oyo 452,477 Oyo
Kigali 821,881 Kigali
Port Elizabeth 952,747 Port Elizabeth
Lubumbashi 1,746,415 Lubumbashi
Bamako 2,358,106 Bamako
Ibadan 2,954,967 Ibadan
Addis Ababa 3,009,130 Addis Ababa
Kampala 3,017,000 Kampala
Accra 4,429,649 Accra
Luanda 5,555,024 Luanda
Johannesburg 8,000,159 Johannesburg
Kinshasa 10,226,183 Kinshasa
Lagos 11,008,357 Lagos
0.0% 2.0% 4.0% 6.0% 8.0% 10.0% 12.0%
Note: included in the Atlas of Urban Expansion n Urban Extent Population n Built-up Area Total
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | 51
Figure C-3 Comparison of the main cities in Africa with the other global cities
in the Atlas of Urban Expansion.
200 Global Cities
10.0%
9.0%
8.0% East Asia and the
Pacific, 194,515,894
Built up in coastal cities 2015 (sq km)
7.0%
South and Central Asia, Southeast Asia,
6.0% 124,277,306 59,478,200
5.0%
Western Africa and Sub-Saharan Africa,
North Africa, 56,478,200
4.0% 961,907,907
Europe and Japan,
123,535,504
3.0% Latin America and the
Caribbean, 94,346,684
2.0%
1.0% Land-Rich Developed
Countries,
72,232,511
0.0%
0.0% 1.0% 2.0% 3.0% 4.0% 5.0% 6.0%
Urban Extent Population (mean % annual growth)
Note: Sub-Saharan cities show the third largest mean annual growth in urban extent and in population growth.
52 | Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa
Figure C-4 Example of urban expansion in Accra, Ghana.
Study area Rural open space
Accra, Ghana Urban extent Exurban built-up area
1991–2014 Urban built-up area Exurban open space
km Suburban built-up area Water
0 5 10 15 20 Rural built-up No Data
Urbanized open space CBD
Note: Based on the Atlas of urban expansion. The built-up area expansion pattern is consistent with the GHSL data.
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | 53
Figure C-5 Example of urban change in Accra, Ghana.
Accra, Ghana (Sub-Saharan Africa 1903–2014)
1996 1929 1966
1991 2000 2014
CBD
Accra, Ghana 1903 1991
1903–2014 1929 2000 Study area
1966 2014 Water
0 4.25 8.5 17 km Arterial roads
No data
Note: Based on the Atlas of urban expansion. The built-up area expansion pattern is consistent with the GHSL data.
54 | Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa
Notes
1. List Available at: http://www.atlasofurbanexpansion.org/
data
References
Lincoln Institute of Land Policy. 2012. Atlas of Urban Expansion.
Cambridge, MA. http://www.atlasofurbanexpansion.org/
Marconcini, M., Metz-Marconcini, A., Üreyen, S., Palacios-
Lopez, D., Hanke, W., Bachofer, F., Zeidler, J., Esch, T., Gorelick,
N., Kakarla, A., Paganini, M., and Strano, E. 2020. Outlining
where humans live, the World Settlement Footprint 2015. Sci.
Data, 7, 242.
Living on the water’s edge. Flood risk and resilience of coastal cities in Sub-Saharan Africa | 55
Appendix D. Normalized values of Annual
Expected Damages
Table D-1 Average annual flood damage for the top 40 cities in SSA.
Exposure Max damage Reference exposure AAD acoastal AAD rainwater AAD total
City in city (km2) unitary ($) value in city ($) (Normalized) (Normalized) (Normalized)
Lagos 1,170 15.9 18,596,578,301 0.0562% 0.1069% 0.1632%
Accra 876 12.8 11,199,296,317 0.0008% 0.1539% 0.1547%
Cape Town 743 24.6 18,292,167,621 0.0020% 0.0874% 0.0895%
Luanda 771 20.3 15,616,395,031 0.9356% 0.06295 0.0988%
Durban 785 24.6 19,336,956,448 0.0056% 0.0661% 0.0717%
Dar es Salaam 654 9.0 5,907,247,542 0.0560% 0.0651% 0.1211%
Conakry 273 8.4 2,299,814,401 0.1703% 0.1116% 0.2819%
Cotonou 300 10.0 2,994,249,783 0.1509% 0.0390% 0.1899%
Dakar 246 11.5 2,836,815,338 0.1282% 0.0547% 0.1828%
Beira 30 8.4 255,901,570 1.7758% 0.0548% 1.8306%
Abidjan 394 12.8 5,034,370,387 0.0103% 0.0724% 0.0826%
San Pedro 37 12.8 470,809,856 0.0014% 0.8277% 0.8292%
Monrovia 244 7.5 1,825,314,482 0.0265% 0.1871% 0.2137
Bata 49 34.3 1,671,571,336 0.0287% 0.1860% 0.2147%
Port Elizabeth 239 24.6 5,876,141,248 0.0077% 0.0476% 0.0554%
Ikorodu 265 15.9 4,204,756,899 0.0026% 0.0745% 0.0771%
Lomé 265 9.0 2,399,476,765 0.0774% 0.0554% 0.1328%
Somerset West 54 24.6 1,332,018,821 0.0032% 0.2071% 0.2103%
Nouakchott 136 11.5 1,564,503,144 0.1321% 0.0162% 0.1483%
Benguela 50 20.3 1,011,212,633 0.0070% 0.2127% 0.2196%
Warri 136 15.9 2,156,353,009 0.0015% 0.1007% 0.1022%
Freetown 115 8.4 970,279,831 0.0759% 0.1496% 0.2255%
Libreville 113 27.4 3,106,647,207 0.0309% 0.0389% 0.698%
Lobito 69 20.3 1,388,412,644 0.0121% 0.1371% 0.1492%
Maputo 418 8.4 3,514,600,020 0.0280% 0.0299% 0.0579%
Douala 189 12.2 2,301,598,066 0.0307% 0.0468% 0.0775%
Kamsar 23 8.4 192,906,197 0.6631% 0.0730% 0.7362%
Saint-Louis 17 11.5 199,526,440 0.3509% 0.3187% 0.6696%
Pointe Noire 74 17.5 1,286,458,660 0.0093% 0.0912% 0.1006%
Takoradi [Sekondi-Takoradi] 105 12.8 1,347,612,842 0.0086% 0.0853% 0.0939%
Bonny 19 15.9 298,291,195 0.3442 0.0731% 0.4173%
Port Gentil 22 27.4 602,685,677 0.1047% 0.0939% 0.1985%
Mombasa 104 11.5 1,205,071,156 0.0329% 0.0505% 0.0835%
Zanzibar City 80 9.0 723,494,563 0.0000% 0.1308% 0.1308%
Lakuwe 34 15.9 542,267,295 0.0000% 0.1576% 0.1576%
Port Hartcourt 93 15.9 1,474,537,309 0.0008% 0.0559% 0.0567%
Badagry 43 15.9 686,156,087 0.0012% 0.1175% 0.1187%
Aboisso 8 12.8 100,910,757 0.0000% 0.7997% 0.7997%
Calabar 78 15.9 1,235,232,672 0.0000% 0.0643% 0.0643%
Wells 38 24.6 942,377,754 0.0000% 0.0810% 0.0810%
Note: Normalized by the total maximum damage per city. $ values reflect USD.
Photo: Sarah Farhat/World bank