Content uploaded by Brian Blankespoor
Author content
All content in this area was uploaded by Brian Blankespoor on Nov 03, 2015
Content may be subject to copyright.
Appendixes
125
Objectives and Aims
Part of the assessment and diagnostic of integrated urban water manage-
ment (IUWM) in Africa consisted of a knowledge, attitudes, and practices
(KAP) survey on urban water management planning for decision makers
and managers of African municipalities and water operators. The survey
was inspired by previous work carried out by the World Bank Water
and Sanitation Program for the Water Operators Partnership. In order to
understand opportunities and constraints for IUWM in Africa, its princi-
pal aims were as follows:
• identify the perceptions of the current and future challenges of urban
water management by cities and water operators in Africa
• provide information on the planning practices of these organizations.
The specific aims of the survey were to understand the following:
• aspects of current urban water management practices in Africa
• attitudes of cities and water operators to urban water management
APPENDIX 1
Knowledge, Attitudes, and Practices
Survey—Methodology*
* For more details, see Closas, Schemie, and Jacobsen. “Knowledge, Attitudes, and Practices Survey of Urban
Water Management in Africa.” World Bank, Washington, DC. Available at http://water.worldbank.org/AfricaIUWM.
126 The Future of Water in African Cities
• constraints to urban water management as perceived by cities and
water operators
• similarities and differences of knowledge and attitudes between cities
and water operators.
Survey Design and Dissemination
KAP surveys are designed to assess the knowledge that a selected com-
munity possesses on a certain topic. They also aim at revealing the
attitudes and perceptions of that selected community toward the topic,
and illustrate what occurs at the individual and organizational level as a
consequence of such knowledge and attitudes (WHO, 2008).
A review of previously conducted qualitative and quantitative surveys
conducted by the World Bank Water and Sanitation Program for the
Water Operators Partnership on water utility services helped leverage
some of the material needed to construct the survey. Additionally, a gen-
eral examination of relevant literature on urban water management and
the related documents helped design the survey. Previous work on
IUWM such as the European Union’s Sustainable Water Management
Improves Tomorrow’s Cities Health (SWITCH) project and publications
by organizations like the Global Water Partnership (Rees, 2006) also
helped define the issues relevant to assess opportunities and constraints
for IUWM.
The target community identified for this survey consisted of chief
planners and other senior managers of water operators and municipalities.
A first version of the survey was tested as a pilot among five water utili-
ties and two cities in January 2012. After the reception of the responses,
the team modified some of the questions to clarify the scope and aims of
some of them and to avoid double-barreled questions. Some questions
had to be rewritten to include more specific instructions and wording.
Some questions were deleted and new questions were added to help with
comprehension of the questions and consistency in responses.
Survey Structure and Contents
The survey included a combination of multiple choice questions, rat-
ing scales with equal weights, and close- and open-ended questions.
The questionnaire was divided into four sections. The first section was
intended to identify the respondent: name of organization, title, and loca-
tion. Section two was aimed at reflecting on the knowledge of the respon-
dents to capture the main general aspects of the organization’s current
Appendix 1 127
planning processes: if the city or water utility had an approved master
plan, the planning horizon of the master plan, the stakeholders involved
in the planning process, and the inclusion of 14 selected features in the
current management plan.
Section three focused on the attitudes and opinions of the respondents
regarding 23 specific planning aspects that could be included in the man-
agement plans of their organization. The aim of this section was to grasp
the extent to which managers and decision makers in these organizations
believed that certain features should be included in the management
plans. It also sought to reveal the degree of belief by managers and deci-
sion makers in the value of a more integrated approach to urban water
management in their planning practices.
Section four listed a series of open-ended questions with the purpose
of extending the respondents’ reflection on the topics and encouraging
them to expand on the following aspects: what can be added to their
organization’s management plan, what problems their organization is
currently facing with regard to urban water management, what are the
constraints to achieve better urban water management, at what political
or operational governance level the identified constraints can be resolved,
and what kind of partnership can help solve the mentioned problems.
Dissemination and Response Rate
The team used the 16th African Water Association (AfWA) Congress in
Marrakech in March 2012 to inform the participants about the Africa
IUWM economic and sector work and the accompanying survey. The
survey was designed and uploaded onto an Internet platform. The survey
was also disseminated in paper form as an email attachment for use by
respondents who had difficulties accessing the online format.
The survey was sent to 80 utilities in Central, Western, Eastern, and
Southern Africa and 24 questionnaires were filled out, which represents
a response rate of 30 percent. For municipalities, the survey was dissemi-
nated to 39 municipalities of which 13 finally responded, which repre-
sents a response rate of 33.3 percent.
Survey Responses and Representativeness
The selection of the sample for the survey was based on a combina-
tion of purposeful and convenience sampling methods. The inclusion
of respondents to the sample was structured around their membership
to two different organizations: AfWA and the United Cities and Local
128 The Future of Water in African Cities
Governments of Africa (UCLGA) organization. The sample selection
also comprised some convenience aspects due to the fact that no proac-
tive methods of data extraction were used. As a result, we do not know
whether the sample is statistically representative of cities and utilities in
Africa.
The decision to use these two organizations, to count on the active
participation of their secretary generals, and to include the secretary gen-
erals as parties in this exercise was done to increase the survey response.
The local knowledge possessed by these organizations facilitated better
and more direct communication with the respondents and a higher level
of exposure and dissemination of the survey.
The sample selected included a substantial representation of large cit-
ies in Africa (22 out of 34 cities having more than 1,000,000 inhabitants),
but also a representation of small and medium cities (four cities having
less than 100,000 inhabitants and eight others with between 100,000 and
1,000,000 inhabitants) (see Tables A1.1 and A1.2). The sample also
reflected the different colonial administrative traditions of African coun-
tries by including an almost equal representation of Francophone and
Anglophone countries (15 and 17 countries, respectively) and three
Lusophone countries. Answers to the survey by water operators came
from a combination of national- and local-level water utilities and an
almost equal mixture of Francophone and Anglophone countries. Out of
the 24 water operators who responded to the questionnaire, 9 were
national-level water utilities and 15 were local. Among cities, Francophone
countries dominated the responses received.
Appendix 1 129
Table A1.1 Water Utilities That Responded to the KAP Survey
Country City Population
National/local
level water utility Language
Benin Cotonou 882,000 National Francophone
Burkina Faso Ouagadougou 1,911,000 National Francophone
Cameroon Douala 2,348,000 National Francophone
Côte d’Ivoire Abidjan 4,151,000 National Francophone
Ethiopia Addis Ababa 2,919,000 Local Anglophone
Dire Dawa 256,800* Local Anglophone
Ghana Accra 2,469,000 National Anglophone
Guinea Conakry 1,715,000 National Francophone
Kenya Nairobi 3,237,000 Local Anglophone
Liberia Monrovia 812,000 National Anglophone
Mozambique Maputo 1,132,000 Local Lusophone
Nigeria
Abuja 2,010,000 Local Anglophone
Kano 3,271,000 Local Anglophone
Makurdi 297,398** Local Anglophone
Kabusa n.a. Local Anglophone
Lagos 10,788,000 Local Anglophone
Senegal Dakar 2,926,000 Local Francophone
South Africa Johannesburg 3,763,000 Local Anglophone
Cape Town 3,492,000 Local Anglophone
Tanzania Dar es Salaam 3,415,000 Local Anglophone
Togo Lome 1,453,000 National Francophone
Uganda Kampala 1,594,000 National Anglophone
Zambia Lusaka 1,719,000 Local Anglophone
Ndola 455,194* Local Anglophone
Sources: UNDESA, 2012; *city population, http://www.citypopulation.de; **Nigeria National Census 2006.
Note: n.a. = not applicable.
130 The Future of Water in African Cities
Table A1.2 Municipalities That Responded to the KAP Survey
Country City Population Language
Benin Cotonou 882,000 Francophone
Cameroon Yaoundé 2,320,000 Francophone
Cape Verde Mosterios 3,598* Lusophone
Central African Republic Bangui 622,771* Francophone
Congo Brazzaville 1,557,000 Francophone
Côte d’Ivoire Tiassale 19,894* (1988) Francophone
Treichville 170,000** Francophone
Morocco Rabat 1,807,000 Francophone
Niger Diffa 48,005* Francophone
Tahoua 123,373* Francophone
São Tomé and Principe Sao Tome 49,957* Lusophone
Senegal Dakar 2,926,000 Francophone
Seychelles Victoria 26,450* Anglophone
Sources: UNDESA, 2012; *city population, http://www.citypopulation.de; **Reseau Ivoire 2007, http://www.
rezoivoire.net/cotedivoire/ville/12/la-commune-de-treichville.html (accessed June 2012).
Appendix 1 131
Copy of the Questionnaire Sent to Water Operators
1. General information
Q1.1 Professional title
Q1.2 Name of organization
Q1.3 City
Q1.4 District
Q1.5 Region
Q1.6 Country
2. Your organization’s planning
In this section we ask you to answer the following questions about your
organization’s current urban water management plan.
Q2.1 Does your organization have an approved water master plan?
1 = Yes 2 = No 3 = I do not know
Q2.2 What is the planning horizon of your organization? (Please
check the one that applies.)
1 year 2 years 5 years 10 years 20 years Other (specify)
I don’t know
Q2.3 Please specify which one of the following stakeholders are
consulted before the approval of your urban water management plan
(check all that apply).
• City’s planning department
• City council
• Utilities regulator
• Consumers-users associations
• Other (specify)
132 The Future of Water in African Cities
Q2.4 Does your current urban water management plan include:
Yes No Do not know
1 Drainage
2 Bulk water supply (dams, storage, reservoirs)
3 Water resources within your catchment
4 Drought contingency plans
5 Flood contingency plans
6 Rainwater harvesting
7 Consideration of competing users at the river
catchment level
8 Solid waste management
9 Urban land zoning
10 Spatial distribution of population
11 Future population growth
12 Informal settlements
13 Roads
14 The city’s urban plans
3. Your opinion on urban water management
Please indicate your level of agreement with the following statements.
Q3.1 The following should be included in the urban water manage-
ment plan:
Strongly
agree
Moderately
agree
Moderately
disagree
Strongly
disagree
1 Water network
maintenance
2 Sewerage network
maintenance
3 Water reuse
4 Rainwater harvesting
5 Water meters
6 Solid waste
management
7 Wastewater treatment
Appendix 1 133
Strongly
agree
Moderately
agree
Moderately
disagree
Strongly
disagree
8 Mechanisms for
purchasing upstream
water rights
9 Permits for water
abstraction
10 Flood risk areas
11 Drought contingency
plans
12 Water supply to informal
settlements
13 Sanitation to informal
settlements
14 Consultation
mechanisms for water
users and communities
15 Conflict resolution
mechanisms for water
users and communities
16 The city’s urban plans
17 Roads
18 Future population
growth
19 Climate change
20 Mechanisms to follow
pollution standards
21 Mechanisms to pay
pollution taxes
22 Measures to ensure
environmental flows in
rivers
23 Monitoring procedures
for drinking-water
quality
134 The Future of Water in African Cities
4. Open questions
Q4.1 List what else you think should be included in the urban water
management plan.
Q4.2 Rank the three major problems your organization is facing with
regard to urban water management (from the most to the least).
Q4.3 List the constraints to achieve better urban water management.
Q4.4 At what governance level do you think these problems can be
best resolved? (Check all that apply.)
• Water operator
• Water resource management authority
• City council
• Regional government
• National government
• Other (specify)
Q4.5 List the type of partnerships your organization is ready to
develop in order to speed up the resolution of the problems you have
identified.
Copy of the Questionnaire Sent to Municipalities
1. General information
Q1.1 Professional title
Q1.2 City
Q1.3 District
Q1.4 Region
Q1.5 Country
2. Your city’s planning
In this section we ask you to answer the following questions about your
city’s current urban plan.
Appendix 1 135
Q2.1 Does your city have an approved urban development strategy?
1 = Yes 2 = No 3 = I do not know
Q2.2 What is the planning horizon of your city’s urban plan? (Please
check the one that applies.)
1 year 2 years 5 years 10 years 20 years Other (specify)
I do not know
Q2.3 Please specify which one of the following stakeholders or groups
are consulted before the approval of your urban plan (check all that
apply).
• City’s planning department
• City council
• Private operator
• Public operator
• Utilities regulator
• Consumers-users associations
• Other (specify)
Q2.4 Does your current urban plan include:
Yes No Do not know
1 Drainage
2 Bulk water supply (dams, storage, reservoirs)
3 Water resources within your catchment
4 Drought contingency plans
5 Flood contingency plans
6 Rainwater harvesting
7 Consideration of competing water users at the
river catchment level
8 Solid waste management
9 Urban land zoning
10 Spatial distribution of population
11 Future population growth
12 Informal settlements
13 Roads
14 The operator’s urban water management plan
136 The Future of Water in African Cities
3. Your opinion on urban planning
Please indicate your level of agreement with the following statements.
Q3.1 The following should be included in the city urban plan:
Strongly
agree
Moderately
agree
Moderately
disagree
Strongly
disagree
1 Water network maintenance
2 Sewerage network
maintenance
3 Water reuse
4 Rainwater harvesting
5 Water meters
6 Solid waste management
7 Wastewater treatment
8 Mechanisms for purchasing
upstream water rights
9 Permits for water abstraction
10 Flood risk areas
11 Drought contingency plans
12 Water supply to informal
settlements
13 Sanitation to informal
settlements
14 Consultation mechanisms for
water users and communities
15 Conflict resolution
mechanisms for water users
and communities
16 The city’s urban water
management plan
17 Future population growth
18 Climate change
19 Mechanisms to set pollution
standards
20 Mechanisms to set pollution
taxes
Appendix 1 137
4. Open questions
Q4.1 List what else you think should be included in the city urban
plan.
Q4.2 Rank the three major problems your city is facing with regard to
urban water management (from the most to the least).
Q4.3 List the constraints to achieve better urban water management.
Q4.4 At what governance level do you think these problems can be
best resolved? (Check all that apply.)
• Water operator
• Water resource management authority
• City council
• Regional government
• National government
• Other (specify)
Q4.5 List the type of partnerships your city is ready to develop in
order to speed up the resolution of the problems you have identified.
139
APPENDIX 2
Objective
City-level data for African cities are scarce. City leaders and decision
makers within and outside cities need data to make informed planning
and investment decisions. As part of the economic and sector work on
integrated urban water management (IUWM), the World Bank under-
took a major data collection exercise.
The objective of this diagnostic of water management for 31 cities is
to contribute to better decision making by providing in one place an
internally consistent overview of the current and future water manage-
ment challenges facing selected African cities. The diagnostic also pro-
vides an indication of the institutional capacities that these cities have to
deal with the challenges.
The diagnostic is presented in three forms:
• a city dashboard that provides an overview for each city
• a comparative table that illustrates the challenges faced by the selected
African cities relevant to each other
• an IUWM capacities and challenges index.
Diagnostic of Water Management
for 31 Cities in Africa*
* For more details, see Closas, Jacobsen and Naughton. “Africa Integrated Urban Water Management Index.”
World Bank, Washington, DC. Available at http://water.worldbank.org/AfricaIUWM.
140 The Future of Water in African Cities
This appendix documents the methodology, data collection, validation,
and representation of the diagnostic. A full list of indicators and a sample
of data sources are provided in Appendix 3.
City leaders and decision makers may also find Appendix 4,
“Methodology for Urban Extent Maps,” to be of interest and of relevance
in making informed planning and investment decisions in relation to
water. For the 31 cities, Appendix 4 shows historical and possible future
spatial extent up to 2025.
Methodology of Data Collection, Validation, and Representation
The diagnostic was constructed through the following four steps:
1. Select cities to be included.
2. Select broad variables that are likely to have an impact on water man-
agement challenges.
3. Select indicators for each variable. In doing so, consider how the indica-
tor represents the variable, data availability and quality, and process of
data collection and validation.
4. Create tools for data representation.
Selection of Cities
The 31 cities (see Table A2.1) were selected based on whether they ful-
filled some or all of the following criteria:
• population growth rate (more than 3 percent growth rate)1
• size (more than 2,000,000 inhabitants)2
• World Bank presence.
Appendix 2 141
Table A2.1 Cities and Selection Criteria
No. Country City
Population
(‘000
inhabitants)
Population
growth rate
1995–2010
Selection
criteria*
1 Angola Luanda 4,775 5.87 P,G
2 Benin Cotonou 841 2.82 WB
3 Burkina Faso Ouagadougou 1,324 7.02 WB
4Cameroon Douala 2,108 4.56 P,G,WB
5 Yaoundé 1,787 5.45 G,WB
6Democratic
Republic of
Congo
Kinshasa 9,052 4.18 P,G,WB
7 Lubumbashi 1,544 4.06 G,WB
8 Mbuji-Mayi 1,489 4.47 G,WB
9 Republic of
Congo
Brazzaville 1,505 4.19 G,WB
10 Côte d’Ivoire Abidjan 4,175 3.29 P,G
11 Ethiopia Addis Ababa 3,453 2.06 P,WB
12 Ghana Accra 2,332 3.27 P,G,WB
13 Kumasi 1,826 5.04 G
14 Guinea Conakry 1,645 3.30 G,WB
15 Kenya Nairobi 3,363 4.08 P,G,WB
16 Malawi Blantyre 733 n.a. WB
17 Lilongwe 866 4.75 G,WB
18 Mozambique Maputo 1,655 1.37 P,WB
19
Nigeria
Lagos 10,572 3.93 P,G,WB
20 Abuja 1,994 8.93 P,G
21 Ibadan 2,835 2.39 P
22 Kano 3,393 2.23 P
23 Senegal Dakar 2,856 3.66 P,G
24
South Africa
Johannesburg 3,618 2.38 P
25 Cape Town 3,357 2.52 P
26 Durban 2,839 2.33 P
27 Sudan Khartoum 5,185 2.53 P
28 Tanzania Dar es Salaam 2,498 4.77 P,G,WB
29 Uganda Kampala 1,597 3.72 G
30 Zambia Lusaka 1,421 4.30 G,WB
31 Zimbabwe Harare 1,663 1.30 WB
Source: World Bank.
Note: Selection criteria: P = population size (> 2 million); G = growth rate (>3 percent annual growth);
WB = World Bank presence. n.a. = not available.
142 The Future of Water in African Cities
Selection of Variables
The selection of variables included in the 31 cities diagnostic is based on
the existing knowledge and practices of IUWM. As defined in the book,
IUWM adopts a holistic view of all components of the urban water cycle
in the context of the wider watershed to develop efficient and flexible
urban water systems. The variables chosen for the diagnostic focus on the
aspects of IUWM highlighted in the book and present the main capaci-
ties and challenges for IUWM faced by major urban areas in Sub-Saharan
Africa. For this exercise, seven different variables were identified that
would best represent the challenges and capacities of IUWM faced by cit-
ies in Sub-Saharan Africa: urbanization challenges, solid waste management,
water resources availability, water supply services, sanitation services, flood
hazards in river basins, and economic and institutional strength.
Urbanization Challenges
One of the major impacts on urban water management is urban growth.
Rapidly expanding cities will need sustainable solutions as they face an
increase in water demand and expansion of water coverage. Increasingly
dense urban areas will also need new planning tools to cope with the
future demand of urban services and infrastructure. IUWM presents the
opportunity to do things differently and to prepare for the future chal-
lenges faced by cities in Sub-Saharan Africa.
Solid Waste Management
By integrating urban planning with water supply and resource manage-
ment, IUWM focuses on the linkages between different urban services
and their impact on population and the urban waterscape. The lack of
solid waste collection and management in urban areas can increase the
risk of disease and health problems among the populations directly or
indirectly exposed to the waste. Additionally, the lack of proper contain-
ment and management of solid waste causes environmental degradation
as well as pollution of water resources through seepage and leakage from
dumping sites. Poor solid waste collection can also increase urban runoff by
blocking inadequately maintained drainage channels during storm events.
Water Resources Availability
The need to supply water to growing urban populations has to take into
account the availability of water resources within the catchment in which
the city is located. By considering the close link between water resources
in the watershed and urban water demand, IUWM integrates these com-
Appendix 2 143
ponents across spatial scales. Urban population growth will put more
pressure on existing water resources and will increase competition. In
this case, exploring, diversifying, and planning new water supply sources
under IUWM will reduce vulnerability and increase security in a scenario
of climate change and potential future scarcity due to climate variability.
Water Supply Services
Water services and infrastructure capacity need to be planned in view of
expanding demand due to population growth. IUWM incorporates tradi-
tional water supply technologies but also includes innovative approaches
to help respond to the challenge of servicing more people. Diversity of
water supply, new technologies such as wastewater reuse, and decentral-
ized systems can help accommodate and adapt to growing demands and
future challenges in urban areas. However, the current state of infrastruc-
ture must first be assessed to estimate future needs.
Sanitation Services
Lack of basic access to sanitation is a major cause of human disease and
contamination of water sources. Dumping of untreated sewage, lack of
collection, or poor wastewater treatment infrastructure can affect major
surface and groundwater bodies for which cities are dependent on for
their water supply. The challenges related to the lack of sanitation and
the impact on water resources supply and human health can be dealt
with by planning and articulating in a more integrated way the access to
basic urban services by the population. The use in IUWM of innovative
technologies for wastewater treatment can improve sanitation access and
reduce the environmental and social costs of the lack of sanitation among
urban populations.
Flood Hazards in River Basins
Urban populations living in coastal cities are exposed to storm surges and
floods. Additionally, coastal cities also face a rise in sea level as well as the
impacts of flash floods, storm damage, and coastal erosion. People living
in flood-prone areas in cities will be more vulnerable to these challenges,
and the infrastructure can suffer severe damages during these types of
events. Due to the lack of storm drainage and the nondisposal of solid
waste, flood events can increase the risks of sickness through the trans-
mission of water-borne diseases in flooded areas and the contamination
of water sources. Floods can also add stress to stormwater and sewerage
systems as well as disrupt water supply and treatment systems. By inte-
144 The Future of Water in African Cities
grating water resources, water supply, and city planning at both the city
level and the watershed level, IUWM can help cities adapt to different
intensities and frequencies of flood events by incorporating new tech-
nologies for flood control and stormwater management.
Economic and Institutional Strength
In addition to the previous six variables, IUWM also focuses on the poten-
tial of institutions to address the different challenges faced by urban areas.
The potential of institutions to develop IUWM is linked to the manage-
ment, and the institutional and legal capacities of cities and water utilities.
For this reason, IUWM envisages close cooperation of water utilities with
cities and key stakeholders to deliver services to larger sectors of the urban
population. Adequate institutional and governance frameworks put in place
by local and national governments as well as regulators and water utilities
will ensure economic and socially sustainable urban water management.
Selection of Indicators
For the seven variables defined previously as relevant to future African
cities in the context of IUWM, indicators were selected based on the fol-
lowing criteria:
• Indicators should be as representative as possible and cover all aspects
of the variable (in terms of completeness, causality, and complementa-
riness).
• City-level indicators were preferred so as to enable comparison between
cities, and to present a more accurate description of the city-level situ-
ation. However, national proxies had to be used in some cases due to
data constraints. Similarly, utility-level data varied depending on the
utility’s coverage; mostly, coverage was at city-level, but some utilities
are national (for example, in Senegal). Nevertheless, for the indicators
concerned (see Appendix 3), these proxies were assumed to be valid.
• Indicators that were available consistently for all or most of the 31 cit-
ies were preferred.
• Indicators were selected to be accessible and useful to the end user due
to the target audience being both internal to the World Bank and exter-
nal (city leaders);
• The indicator selection process was very much constrained by the avail-
ability, consistency, and reliability of the data for the 31 cities, which
highlights the need to systematize such data for monitoring and plan-
ning purposes.
Appendix 2 145
Urbanization Challenges
While population growth and urbanization are major drivers for change
in Africa, indicators had to be at city level to make sense for this particu-
lar variable, which restricted the selection of indicators. The six indicators
chosen (see Appendix 3) complement each other by illustrating both
quantity and quality of urbanization challenges. The first four indicators
quantify past, current, and future city growth in relation to national pop-
ulation growth, while indicators five and six quantify current city popula-
tion in terms of pressure on urban infrastructure by looking at density and
the share of population living in informal areas. Indicators for quantifying
current and projected city population growth, as well as city density,
came from relatively homogeneous sources (see Appendix 3). However,
the last indicator (percentage of city population living in informal areas)
came from a variety of sources and years, which leads to discrepancies in
the data in terms of validation, definition, and consistency.
Solid Waste Management
It is solid waste collection and disposal, or more precisely the lack thereof,
that impacts infrastructure efficiency by clogging drains, leads to health
issues, and aggravates the effects of extreme weather events in a city. We
looked for indicators that determined the lack of solid waste management
both at household and business levels (primary collection) and municipal
level (secondary collection). This corresponds to indicators seven and
eight respectively (see Appendix 3). From this, the proportion of solid
waste that does not enter the formal disposal chain can be derived and
utilized for integrated urban planning. The lack of a single and compre-
hensive global source for solid waste management at city level for all 31
cities presented a problem, which restricted the number of indicators
for this variable, and impeded efforts to quantify informal waste disposal
more rigorously. Furthermore, the data used still presents some inconsis-
tencies; for instance, data sources rarely indicated whether the percentage
collected is on a wet or dry basis.
Water Resources Availability
As IUWM goes beyond the limits of the city and includes the water
resources available for the city within a wider geographical context, water
resources indicators were selected to encompass the following:
• Past and current water supply from precipitation (indicators 9 and 10)
in the basin from which the city derives its water supply. These indica-
146 The Future of Water in African Cities
tors are sourced from two databases, the Climatic Research Unit (CRU)
of the University of East Anglia 3.0 database and WorldClim database
(Hijmans et al., 2005).
• A range of indicators describing physical water availability in the basin
(indicators 13, 14, 15, 16, and 17), which are sourced from the baseline
data of Strzepek et al., 2011, and represent the period 1961 to 1999. As this
baseline data is not historically observed but modeled data generated from
the CRU 3.0 database, it comes with limitations (see Strzepek et al., 2011).
• The projected range of impact from climate change on specific hydro-
logical indicators at basin level (indicator 11). This indicator is sourced
from the Climate Change Knowledge Portal and combines the results
of 23 Global Circulation Models for three emission scenarios. For spe-
cific limitations associated with the baseline data, the modeling process,
and the resolution of the models, see Strzepek et al., 2011.
Water Supply Services
The current situation of water supply services determines how much
more of a challenge it will be for the city and the utility to deal with an
increase in demand for water at residential and business levels. This set of
indicators was devoted to describing water supply services at city level in
terms of the following:
• Capacity of current water infrastructure (indicators 18 and 19), which indi-
cate how much of the lack of water access is due to poor infrastructure.
• Quality and quantity of water supply coverage (indicators 20, 21, and
22) for current population, which gives an indication of how future
demand might increase in terms of consumption per capita as well as
population growth.
• Financial sustainability of water utility (indicators 23, 24, 25, 26, and 27),
which gives an estimate of how much more of a financial and manage-
ment challenge it will be to expand coverage and supply for the utility.
Data from most of these indicators came from a variety of sources (see
Appendix 3), which leads to inconsistencies in definition and consistency
across all cities. The International Benchmarking Network for Water and
Sanitation Utilities (IBNET) provided most of the data used to estimate
the financial sustainability of utilities as well as water supply coverage and
infrastructure. In cases where the water utility was national, it was assumed
that it would service the main urban conglomerations and therefore was
used as a reliable proxy to evaluate water utility governance at city level.
Appendix 2 147
Sanitation Services
The lack of sanitation presents a challenge in that it impacts both human
health, and water and environmental quality in growing cities. We there-
fore sought some reliable indicators to measure the lack of sanitation at
household (primary) level as well as municipal (secondary) level in the
cities surveyed. This corresponds to indicators 28 and 29 respectively.
The first limitation is that these indicators come from a variety of sources,
which has implications on definition, reliability, and consistency across the
cities surveyed. Secondly, it must be emphasized that access to improved
sanitation does not in itself indicate that sewage is disposed of in a safe man-
ner. As we had difficulties isolating an indicator for the disposal of all sewage,
we decided to look at cholera prevalence as an indicator of poor sewage
disposal (indicator 30). Again, sources are disparate and not consistent, and
cholera is not only an indicator of extremely poor sewage disposal, but
also of other factors including the resilience and efficiency of the health
system, which we did not intend to capture here. Nevertheless, for the
cities in consideration, it is the most representative indicator available.
Flood Hazards in River Basins
Indicator 31 selected for this variable was calculated for this study. It
presents an estimate of flood frequency based on the United Nations
Environment Programme (UNEP)/Global Resource Information
Database-Europe (GRID) PREVIEW flood data set. The unit is the
expected average number of events per 100 years (hydrological model of
peak-flow magnitude). The methodology and sources used to create this
indicator come from a variety of studies. City basin summary statistics
are derived from the basin definition used by Strzepek et al., 2011. The
frequency of flood events was created through a three-step process:
1. Use of GIS modeling using a statistical estimation of peak-flow magni-
tude and a hydrological model using HydroSHEDS data set and the Man-
ning equation to estimate river stage for the calculated discharge value.
2. Observed flood events from 1999 to 2007, obtained from the Dart-
mouth Flood Observatory.
3. The frequency was set using the frequency from UNEP/GRID-Europe
PREVIEW flood data set. In areas where no information was available,
it was set to a 50-year returning period.
The data set was designed by UNEP/GRID-Europe for the Global
Assessment Report on Risk Reduction. It was modeled using global data.
148 The Future of Water in African Cities
UNEP/GRID-Europe is credited for GIS processing, with key support
from the USGS EROS Data Center, Dartmouth Flood Observatory 2008.
The source for this data subset is the Dartmouth Flood Observatory,
Dartmouth College. The data used to represent the level of flood fre-
quency in the cities’ river catchments is limited by the inputs of the
hydrological model used. The data used the definition of hydrological
catchments as defined by Strzepek et al., 2011, which presents limita-
tions related to the use of satellite imagery in establishing land elevation
and surface water runoff at the urban level that could include the pres-
ence of water infrastructure.
Economic and Institutional Strength
This variable includes institutional cooperation, transparency, efficiency,
and the wider economic context needed to enable integration of urban
services. Framing a useful set of indicators to illustrate this variable for
the 31 cities proved difficult. Relevant information on institutions is
fragmented and the realities of political decision making are also difficult
to represent with indicators. Under this variable, we looked for city-level
indicators that showed the levels of
• efficiency as indicated by the sophistication of current management of
water and wastewater services (indicators 32, 33, and 34) and planning
of current and future urban services (indicators 39, 40, 41, and 42)
• transparency of institutional oversight and scrutiny (indicators 36 and
38) for water resources management, and of country-level governance
(indicator 44)
• cooperation between government institutions and water services pro-
viders (indicators 35 and 37)
• wider economic context (indicators 43 and 45).
Due to data availability, some specific indicators (38, 43, 44, and 45)
are given at the national level as there were no equivalent data at the city
level, but are assumed to be viable in the large cities under consideration.
Sources were varied, ranging from World Bank data to the existence of
formal planning documents for each city; therefore, specific planning
documents might have been missed if they were not uploaded onto a
public site. Some specific cluster indices were used (indicators 44 and
45), which also have their own advantages and limitations (see Appendix
3 for references). Furthermore, the Water Operators Partnership database
is self-reported by the utilities, which makes it difficult to validate data.
Appendix 2 149
Data Collection
The data used to create the indicators in this study was obtained through
a process of exhaustive analysis of relevant published materials available
on the Internet and produced by a wide array of organizations. Data col-
lection was carried out in Washington, DC, between January and March
2012. The process of data collection followed two iterations. The first
one included the initial review of the main sources of data identified
by the research team as primary data sources. Once the initial sources
were exhausted, the second iteration incorporated additional material
from different secondary sources to complete the data gaps in the list of
indicators.
Sources of Data
Due to the lack of a single or centralized source containing the data for
the indicators, this study had to rely on multiple sources of informa-
tion. Although data at the national level for urban water management
are available (for example, WHO-UNICEF Joint Monitoring Program,
Table A2.2 Main Data Sources
World Bank
sources UN sources
Other
international
agencies
National public
sources
Private and open
sources
• PADs
• AICD
• IBNET
• CSO AMCOW
• Other World
Bank
publications
• UN-Habitat
• UNDP
• UNEP/GRID
• AfDB PADs
• European
Union
• OECD
• International
development
agencies
• Specific
government
branches
(censuses,
household
surveys,
ministries, and
regulators)
• Water operators
(public)
• Dartmouth
Flood
Observatory,
Dartmouth
College
• IWA Water Wiki
• Wikipedia
• Water operators
(private)
• Consultancies
• NGOs
• Research
institutes
• Academic
journals
Source: World Bank.
Note: PADs, project appraisal documents; AICD, Africa infrastructure country diagnostic; CSO, country status
overview; AMCOW, African Ministers’ Council on Water; UNDP, United Nations Development Programme, AfDB,
African Development Bank; OECD, Organisation for Economic Co-operation and Development; IWA,
International Water Association.; NGOs, nongovernmental organizations.
150 The Future of Water in African Cities
African Ministers’ Council on Water, country status overviews), the lack
of city-specific data on urban water management made the task of creat-
ing the indicators and populating the sample of 31 cities more challeng-
ing (see Appendix 3 for a sample of sources).
Reliability and Quality of Data
The data contained in this general study of 31 cities has to be considered
with a degree of caution due to general inconsistencies in definitions, mea-
surements, and data collection methodologies. The inherent complexities
of the sector, the difficulties in measuring institutional arrangements, and
the validation of the data found added limitations to the data set.
The reliability of data and sources also affects the quality of the data
used in this study and the different types of analyses that can be derived
from the data. Following is a list of several of the main limitations affect-
ing the data set:
• The different methodologies used by the different data sources add
uncertainty to the data set.
• Different metrics and different definitions used by the sources add pre-
cision problems, which make the homogenization and integration of
the indicators difficult.
• The use of different sources for the same indicator and different years
adds inconsistencies and complications when homogenizing and nor-
malizing the data to compare the different indicators.
• These limitations present a problem when trying to test the robustness
of the data with different statistical methods or trying to use the data
for more complex statistical analyses.
• In some instances, the data was self-reported, which limited its validity.
Validation of Data
One question that can arise from the discussion in this section is how can
we be sure of the data’s representativeness—in other words, does the data
used in this study depict a consistent, useful, and rigorous story. This con-
cern is even more pertinent in the context of Sub-Saharan Africa where
the lack of data makes any new study a focus of serious and deep scrutiny.
The team is cognizant of the difficulties of finding good quality data
for cities in Sub-Saharan Africa. The team also attempted to use global
and regional sources where possible (for consistency) as opposed to spe-
cific documents and other reports where differences in definitions might
Appendix 2 151
constitute an issue for the data’s consistency. In many cases, however,
sources were scarce or presented methodological problems, and in some
cases they were nonexistent. The team has maintained the highest level
of rigor possible with the different sources used and in the data presen-
tation and use in the study. However, due to these limitations different
opinions about the choice of sources might still arise. The lack of data
and their poor quality is also indicative of the necessity to explore and
generate more data for Sub-Saharan Africa. Additionally, the need to
generate more primary data on the state of water and sanitation at the
city level is coupled with the need, as reflected in this study, to pay more
attention to IUWM.
To overcome these problems, an exhaustive record of the original
sources of data has been kept.3 Descriptive and clarification notes con-
taining the different definitions used and the sources have been added to
the different outputs as well as to each one of the tables used. These
notes try to be as detailed and comprehensive as possible although a
degree of uncertainty still remains, due to the difficulty of tracking and
annotating each one of the different metrics and definitions used.
However, when the data presented some discrepancies, the information
found was crossed-referenced and compared to other sources containing
similar data to check whether the data found was within a reliable inter-
val of significance.
Due to the variety of sources, discrepancies of data measurement can
arise from this data set. During the collection of the data, when a problem
of definition or measurement appeared, standardization was sought using
generally and widely used definitions (for example, for improved drinking
water we used data from the WHO-UNICEF Joint Monitoring Program).
However, it has to be acknowledged that some differences and discrepan-
cies might still remain within the data set due to the selection of the data
sources.
Data Representation
Dashboard
The dashboard represents a subset of the data collected for the 31 cities
in a format that would give the target audiences a snapshot of the current
and future situation in the six areas identified for IUWM in Sub-Saharan
African cities.
The dashboard has been constructed as follows:
152 The Future of Water in African Cities
i. Select indicators for the dashboard from the 31 cities database for each
variable.
For each of the seven variables, indicators were selected so as to be as
consistent and relevant for all the cities reviewed. The aim was to give
a concise idea of the situation in each city as well as to show how each
city fared in relation to some internationally established standard. Second,
they were chosen so as to lend themselves easily to visual representation and
enhance user friendliness. For instance, inter- and intra-annual precipitation
data over a period of time were selected specifically for the dashboard.
ii. Add international benchmarks for comparison.
To give an estimate of the relative situation of water utilities in Sub-
Saharan Africa, the dashboard also included a comparison of the indi-
vidual city indicators with a fixed benchmark and the average of values
for all 31 cities. The international benchmark for the operating ratio
(operating costs to revenue ratio) for water utilities was set at 130 percent
(based on Banerjee and Morella, 2011); for the collection ratio, it was set
at 95 percent, as it was considered a realistic benchmark for bill collection
(though a 100 percent collection ratio is considered ideal for best practice
but rarely attainable). A benchmark of 25 percent of nonrevenue water
was also considered acceptable. Finally, the benchmark for the future per-
formance indicator was evaluated at 100 percent—meaning that the prob-
ability of the water utility facing financial difficulties in the next two years
is nil, which is a strong indicator of best practice (Moffitt et al., 2012).
iii. Create visual tools to enhance end user accessibility.
To improve accessibility and comprehension of data, the dashboard
incorporates a selection of visual tools to better represent the selected
indicators and to monitor the information provided by the indicators at
a glance. Graphic visualizations including different types of charts and
graphs have been produced and customized with the purpose to better
present to end users the data gathered. The objective of these visual tools
is also to present different pieces of information together to identify the
relevant challenges and capacities for each one of the cities in the study.
By using colors, sizes, and shapes of figures, the dashboard also has the
aim to increase the understanding of the information presented. The fig-
ures also show the trends and the progress for some selected indicators
compared to the general average of the study sample.
Appendix 2 153
Comparative Table
The illustration of the 31 cities diagnostic also includes a comparison
of the relative position of a city in relation to the other 30 cities for a
subset of the indicators, designed to reflect the areas that constitute cur-
rent challenges and capacities. The comparative table enables each city to
pinpoint its relative strengths and weaknesses compared to other African
cities. The target audience is similar to that of the dashboard, but with
added emphasis on local policy and management audiences. A subset of
indicators from the original set of indicators used to study the 31 cities
was selected (see Appendix 4); again, this subset was thought to repre-
sent six variables identified as central to IUWM, as well as reliable data
that would be accessible to the target audience. The seventh variable on
economic and institutional strength was excluded due to the fact that it
included most of the national proxies used. Since this was a comparative
exercise, city-level data was preferred and national-level data that had
been used for the dashboard was deliberately omitted.
Methodology
A simple methodology based on the one used by the Economist
Intelligence Unit4 in their study of African cities was devised. This
methodology was chosen for its simplicity and also due to the fact that
it limited the level of normalization and aggregation of the indicators
by allowing a comparison of the data indicator by indicator. The choice
of this methodology also avoided ranking the cities or their comparison
against an established benchmark, for it simply compares the values for
each indicator for each city between themselves.
The data from the selected subset of indicators was then homoge-
nized, and the mean and standard deviation for each of the indicators
was calculated (see Table A2.3). The cities and their corresponding indi-
vidual values for every subindicator have been assigned to one of five
intervals depending on how much each of the individual values differed
from the mean, plus or minus x times the standard deviation. Each city
value has been normalized then aggregated into one single indicator, giv-
ing equal weight to each of the subindicators. The values were then clas-
sified on a scale of 0 to 4 and matched with the interval they belong to
according to their aggregated values. The groups were classified based on
different intervals calculated with the mean score and standard deviation
as follows:
154 The Future of Water in African Cities
• 0 = below mean minus 1.5 times standard deviation
• 1 = between mean minus 1.5 times standard deviation and mean minus
0.5 times standard deviation
• 2 = between mean minus 0.5 times standard deviation and mean plus
0.5 times standard deviation
• 3 = between mean plus 0.5 times standard deviation and mean plus 1.5
times standard deviation
• 4 = above mean plus 1.5 times standard deviation.
Table A2.3 Calculation, Definition, and Codification of Intervals
Calculation of
intervals
Below
mean
−1.5 × SD
Between
mean
− 0.5 × SD
and mean
− 1.5 × SD
Between
mean
− 0.5 × SD
and mean
+ 0.5 × SD
Between
mean
+ 0.5 × SD
and mean
+ 1.5 × SD
Above
mean
+ 1.5 × SD
Codification for
normalization of
intervals
01234
Values for
intervals
Between 0
and 0.99
Between 1
and 1.99
2 Between
2.01 and
2.99
Between 3
and 4
Definition of
intervals
Well below
average
Below
average
Average Above
average
Well above
average
Source: World Bank.
Note: SD = standard deviation.
IUWM Capacities and Challenges Index
There are many dimensions to consider when thinking of a city within
the framework of IUWM: that of water supply as well as wastewater
infrastructure, but also solid waste management, catchment condition,
urbanization, governance structures, and so on. The variety of dimensions
at the core of this approach make it difficult to compare cities to each
other to estimate where the most pressing needs in terms of water-related
management exist. The purpose of this index is therefore to describe the
capacity and challenges of the 31 cities investigated in this report in rela-
tion to the multidimensional aspect of IUWM.
This index follows in the footsteps of approaches that have attempted
to boil down multidimensional concepts into composite indicators, par-
ticularly in the sector of development, governance, and environment. A
composite indicator is formed when individual indicators are compiled
Appendix 2 155
into a single index, which should ideally measure multidimensional con-
cepts that cannot be captured by a single indicator (OECD, 2008). There
is considerable literature on the merits and criticisms of composite indica-
tors; for a review of the main composite indicators used to compare
countries see Bandura (2008). Booysen (2002) gives an evaluation and
critique of composite indicators in the field of development, which does
point out that indices remain invaluable in terms of their ability to sim-
plify complex measurement constructs. In the case of IUWM, the added
value of such an approach is to get a holistic picture of the water-related
challenges and capacities of the cities, which can then be easily broken
down into its component parts, or water-related variables.
Methodology
In this index, cities were thought of as having two dimensions: water-
related capacities and challenges. For each of these dimensions, the vari-
ables defined earlier as relevant to IUWM were categorized. However,
not all of the indicators included in each variable could be thought of
as belonging to the same dimension. For instance, the indicator for rev-
enue water was thought of as illustrating the utility’s capacity to deal
with infrastructure maintenance and was therefore separated from other
indicators for the water supply services variable, which remained in the
challenges dimension. This explains why there are more variables in the
index, though the indicators used all come from the same database (see
Appendix 3 for the full list of indicators used).
Table A2.4 Variables and Indicators Representing the Challenges and Capacities
of Cities
Challenges Capacities
Urbanization challenges (1, 6) Country policies and institutions (44)
Solid waste management (7, 8) Economic strength (43)
Water supply services (20, 22) Water-related institutions
(35, 36, 37, 38, 39, 40, 41, 42)
Sanitation services (28, 29, 30) Water utility governance (23, 25)
Flood hazards in river basin (31)
Water resource availability (13)
Source: World Bank.
Note: Indicators assigned for each variable in the index differ from the main database (see Appendix 3) and are
represented here by the numbers in parentheses. For indicator numbers, see Appendix 3.
156 The Future of Water in African Cities
Data for both challenges and capacities were normalized from zero to
maximum so as to facilitate aggregation. When needed, data was inverted
prior to normalization, so that results were consistent with the overall
message of the category (for example, percentage of the population with
improved sanitation was converted to percentage of the population with
no improved sanitation, then normalized, because a higher indicator
value means a higher challenge for the city). When this was not possible
(due to the unit of the indicator), data was normalized from zero to
maximum first, then substracted from a total of 100 percent (for exam-
ple, for water consumption, a high level of water consumption should be
represented by a low value in our index as it represents a lower chal-
lenge).
Indicators were then aggregated as follows: each water-related indica-
tor was assigned even weighting within each index dimension (challenges
and capacity). The practice of even weighting for indices can be subject
to debate but is corroborated by expert opinion (Chowdhury and Squire,
2006).
The limitations faced during the data collection process for the 31 cit-
ies general data set, and outlined in detail in the methodology for 31
cities database, also apply to the capacity/challenge matrix. Gaps in the
data collected meant that for some cities, fewer indicators were available
than for others, which affected the city’s score.
This index is a preliminary attempt to illustrate relative challenges and
capacities between cities and thus to inform decision makers of the great-
est needs. It is hoped that the index will generate a dialogue to improve
evaluation, and incorporate data inputs from cities and other stakeholders
with the view to future improvements.
Notes
1. According to data from UNDESA, 2012.
2. UNDESA, 2012.
3. See database on http://water.worldbank.org/AfricaIUWM (forthcoming).
4. The Economist Intelligence Unit and Siemens, 2011.
157
APPENDIX 3
Indicators for the 31 Cities
Diagnostic*
* See 31 Cities Diagnostic Database, http://water.worldbank.org/AfricaIUWM.
158
Table A3.1 Selection of Indicators for the 31 Cities Diagnostic
Variable Indicator Type Units Dashboard
Comparative
table Index Notes and sources
Urbanization challenges
1. City growth rate, 1995–2010 Quantitative % — √ √ UNDESA, 2012.
2. National population growth,
2010–2025
Quantitative % √ — — UNDESA, 2012.
3. Share of city in national
population, 2010
Quantitative % √ — — UNDESA, 2010.
4. Share of city in national
population, projection, 2025
Quantitative % √ — — UNDESA, 2010.
5. City density Quantitative Population/km2√ — — Demographia, 2011.
6. Percentage of city population
living in informal areas
Quantitative % √ √ √ Various sources (e.g., UN-Habitat, 2008; UN-Habitat,
2007a; World Bank, 2010b).
Solid waste
management
7. Percentage of solid waste
produced collected (public and
private collection)
Quantitative % √ √ √ Various sources (e.g., Parrot et al., 2009; UN-Habitat,
2010; UN-Habitat, 2011a; World Bank, 2009b).
8. Percentage of solid waste
disposed of in controlled sites
(landfill)
Quantitative % √ √ √ Various sources (e.g., Parrot et al., 2009; UN-Habitat,
2010; UN-Habitat, 2011; World Bank, 2009b).
(continued on next page)
159
Variable Indicator Type Units Dashboard
Comparative
table Index Notes and sources
Water resources availability
9. Intra-annual precipitation and
temperature
Figure mm and °C √ — — Annual variation of mean monthly precipitation (in
mm) and temperature. WorldClim climate layers (2.5
arc-minutes and 0.5 degrees resolution per pixel). For
full methodology see Hijmans, et al., 2005.
10. Inter-annual precipitation Figure mm/year √ — — Deviation from mean annual precipitation, 1901–
2006. CRU 3.0 database for 1901 to 2006
(approximately 50–60 km2 per pixel). For full
methodology see Mitchell, et al., 2003.
11. Climate change impact Figure % relative change
to base period
√ — — Statistics for a selection of parameters for three
emission scenarios and 23 Global Circulation Models
for the time period 2050–2059. Source: World Bank
Data, Climate Change Knowledge Portal. For detailed
methodology see Strzepek et al., 2011.
12. River basin map Map — √ — — World Bank Data, Climate Change Knowledge Portal.
For detailed methodology see Strzepek et al., 2011.
13. Average annual runoff Quantitative Million cubic
meters (MCM)/
year
— √ √ World Bank Data, Climate Change Knowledge Portal.
For detailed methodology see Strzepek, et al., 2011.
Average modeled runoff at basin scale for years
1961–1999.
(continued on next page)
Table A3.1 (continued)
160
Variable Indicator Type Units Dashboard
Comparative
table Index Notes and sources
Water resources availability
14. Basin yield Quantitative MCM/year — √ — World Bank Data, Climate Change Knowledge Portal.
For detailed methodology see Strzepek et al., 2011.
Maximum sustainable reservoir releases within the
basin for years 1961–1999.
15. Annual high flow (q10) Quantitative MCM/year — √ — Annual runoff exceeded 10 percent of the time for
years 1961–1999. Source: World Bank Data, Climate
Change Knowledge Portal. For detailed methodology
see Strzepek et al., 2011.
16. Annual low flow (q90) Quantitative MCM/year — √ — Annual runoff exceeded 90 percent of the time for
years 1961–1999. Source: World Bank Data, Climate
Change Knowledge Portal. For detailed methodology
see Strzepek et al., 2011.
17. Groundwater baseflow Quantitative MCM/year — √ — Sustained flow in a river resulting from groundwater.
Source: World Bank Data, Climate Change Knowledge
Portal. For detailed methodology see Strzepek et al.,
2011.
(continued on next page)
Table A3.1 (continued)
161
Variable Indicator Type Units Dashboard
Comparative
table Index Notes and sources
Water supply service
18. Total design capacity of water
supply infrastructure
Quantitative Cubic meters per
day (m3/day)
√ — — Various sources (e.g., AfDB, 2010; IBNET, various years;
World Bank, 2011c).
19. Total production by water
supply infrastructure
Quantitative m3/day √ — — Various sources (e.g., AfDB, 2010; IBNET, various years;
World Bank, 2011c).
20. Residential water
consumption in city or utility
coverage area
Quantitative Liters per capita
per day
√ √ √ Total residential water consumption, in liters per
capita per day. Relates to population served by utility
or population living in city, depending on the source.
Various sources (AfDB, 2006; African Development
Fund, 2007; IBNET, various years).
21. City population served by
utility
Quantitative Persons √ — — IBNET, various years.
22. Percentage of city population
with improved water coverage
Quantitative % √ √ √ Improved water coverage as per source’s definition.
Various sources (e.g., AfDB, 2009; IBNET, various years;
World Bank, 2008a).
23. Percentage of water sold by
utility
Quantitative % — — √ Various sources (e.g., World Bank, 2008; IBNET,
various years).
24. Percentage of collection rate
from population billed
Quantitative % √ √ — Various sources (e.g., IBNET, various years; Pinsent
Masons, 2012; World Bank, 2008).
(continued on next page)
Table A3.1 (continued)
162
Variable Indicator Type Units Dashboard
Comparative
table Index Notes and sources
Water supply service
25. Percentage of nonrevenue
water
Quantitative % √ √ √ Percentage of water produced and lost before
reaching the customer, either through leaks, theft, or
legal use for which no payment is made. Various
sources: (e.g., IBNET, latest year; Pinsent Masons, 2012;
Hove and Tirimboi, 2011).
26. Utility operating ratio Quantitative % √ — — Utility annual operating revenues/annual operating
costs (IBNET, latest year available).
27. Future performance Quantitative % √ — — Inverse of the Water Utility Vulnerability Index (WUVI).
For methodology details see Moffitt et al., 2012. The
WUVI gives an estimation of the probability of a water
utility to face financial difficulty in the next two years;
hence the inverse estimates the probability of strong
future financial performance. To create this indicator,
the inverse is taken: i.e., a high value indicates strong
financial utility performance. Source: IBNET, latest year
available.
Sanitation service
28. Percentage of population
with access to improved
sanitation
Quantitative % √ √ √ Various sources (e.g., AfDB, 2004; República de
Mozambique, 2011; World Bank, 2009a).
29. Percentage of wastewater
treated
Quantitative % √ √ √ Percentage of wastewater treated by treatment plant
system of percentage of wastewater collected.
Various sources (e.g., Mtethiwa, et al., 2008; UN-
Habitat, 2011a; World Bank, 2011c).
30. Number of cases of cholera
from last outbreak
Quantitative Persons — — √ Various sources (e.g., WHO, 2011; WHO, 2006).
(continued on next page)
Table A3.1 (continued)
163
Variable Indicator Type Units Dashboard
Comparative
table Index Notes and sources
Flood hazard
in river basin
31. Frequency of flood events Quantitative Number of
events/100 years
√ √ √ Estimate of flood frequency as the expected average
number of events per 100 years (hydrologic model of
peak-flow magnitude). Sources: UNEP/GRID-Europe
PREVIEW flood data set; Strzepek et al., 2011;
Dartmouth Flood Observatory, Dartmouth College.
Economic and institutional strength
32. Unbundling of bulk water
production and distribution
Qualitative Text √ — — Two different entities responsible for bulk water
production and distribution at national level. Source:
Banerjee et al., 2008.
33. Separation of business lines
water-wastewater
Qualitative Text √ — — Separation of water and wastewater services from
supply in urban area at national level. Source:
Banerjee et al., 2008.
34. Private de facto participation
in water utility
Qualitative Text √ — — Participation of private utilities for water or
wastewater service provision (participation in at least
one of the two). Source: Banerjee et al., 2008.
35. Existence of utility
performance contract with
government
Qualitative Yes = 1
No = 0
— — √ World Bank, Water Operators Partnership database.
36. Existence of regulator Qualitative Yes = 1
No = 0
— — √ At the national level. Source: Banerjee et al., 2008.
37. Existence of water utility
targets for access to services in
informal settlements
Qualitative Yes = 1
No = 0
— — √ World Bank, Water Operators Partnership database.
38. Existence of river basin
authority
Qualitative Yes-No √ — √ Existence of river basin authority within the basin the
city abstracts its water for supply (and year of
creation). Various sources (e.g., Volta River Authority;
Zambezi River Basin Authority).
(continued on next page)
Table A3.1 (continued)
164
Variable Indicator Type Units Dashboard
Comparative
table Index Notes and sources
Economic and institutional strength
39. Existence of water master plan Qualitative Yes-No √ — √ Various sources (e.g., AfDB, 2004; República de
Mozambique, 2011; World Bank, 2009a).
40. Existence of solid waste
management plan
Qualitative Yes-No √ — √ Various sources (e.g., Lusaka City Council, 2003;
République du Congo, 2010; UN-Habitat, 2010).
41. Existence of urban master
plan
Qualitative Yes-No √ — √ Various sources (e.g., Cities Alliance, 2008; JICA, 2012;
UN-Habitat, 2007a).
42. Existence of wastewater
master plan
Qualitative Yes-No √ — √ Various sources (e.g., AfDB, 2005; Maoulidi, 2010;
World Bank, 2007).
43. GDP/capita Quantitative Current US$ √ — √ Using national data. GDP/cap as a measure of wealth
with lowest value = 0 and highest value = maximum
for 31 cities. Values assigned range from 0 to 100%.
Source: World Bank Data, http://data.worldbank.org/
indicator/NY.GDP.PCAP.CD.
44. Country Policy and
Institutional Assessment (CPIA)
Number Index unit √ — √ Using national data. CPIA is a World Bank cluster
index measuring the quality of governance, national
policy, and institutional frameworks. Values are from
lowest = 1 to highest = 6, assigned values in
dashboard normalized from 0 to 100%. Source: World
Bank data, http://data.worldbank.org/indicator/IQ.
CPA.PUBS.XQ.
45. Human Development Index
(HDI)/capita
Number Index unit √ — — Using national data. HDI is a measure of human
development with lowest value = 0 and highest
value = 1, assigned a value of 0 and 100%
respectively. Source: UNDP data, http://hdr.undp.org/
en/statistics/hdi/.
Table A3.1 (continued)
165
APPENDIX 4
Objectives
City-level data for African cities are scarce and spatial information even
more so. City leaders and city and other decision makers need data to
make informed planning and investment decisions. As part of the eco-
nomic and sector work on integrated urban water management (IUWM),
the World Bank undertook a major data collection exercise. As part of
this we developed (based on satellite imagery) a set of maps of urban
extent for 31 cities in Africa. The maps show historic urban extent from
1990 (when available) to today and illustrate likely future urban extent
by 2025. The purpose hereof is to contribute to better decision making
by providing an illustrative tool of the likely spatial consequences of con-
tinued urban growth.
Methodology
Spatial modeling of urban growth has been developing since the 1960s,
and has increasingly moved toward the use of cellular automata models.
These models represent space as a grid where each cell in that grid is
subject to a certain possibility of transition in use based on a defined set
of spatial rules and probabilities. These simple systems can give rise to
Methodology for Urban
Extent Maps*
* For more details, see Duncan, Blankespoor and Engstrom. “Urban Extent Map for 31 cities in Africa”. World Bank,
Washington, DC. Available at http://water.worldbank.org/AfricaIUWM.
166 The Future of Water in African Cities
complex dynamics across spatial scales and have been increasingly used
since the 1990s following the increasing availability of computing power
(Santé et al., 2010).
Four categories of urban models have emerged over the last several
decades (Landis, 2001). In addition to cellular automata models, there are
spatial interaction models that model choices for sitting firms and houses
using economic and travel distance (for example, Wegener, 1998); agent-
based modeling where a reduced set of representative decision makers
populate the landscape and convert land uses based on a set of rules
largely focused on the household- or firm-level (for example, UrbanSim
designed by Waddell, 2002); and urban future models that model site-
level transitions based on spatial relationships such as neighborhood
statistics, distance to infrastructure, demography, economic conditions, and
regulatory conditions (for example, California Urban Futures; Landis, 2001).
One particular model that has been widely studied and used is the
SLEUTH model (Clarke et al., 1997). This model defines rules for cell
transitions from nonurban to urban based on a core set of determinants
and iterates over time. SLEUTH stands for slope, land use, exclusion, urban,
transportation, and hillshade, describing the core input data sets that drive
the model behavior. This form of cellular automata model simulates growth
in time steps, and uses historical data sets to calibrate model parameters
for forecasting in the future. Monte Carlo simulations are run and the
results averaged to give a best estimate of future urbanization.
This modeling effort used a simplified cellular automota approach that
was necessitated by limits in suitable consistent historical data over all
cities, and a limit in computational resources. The first step in the model-
ing approach was to define the suitability of each cell to urbanization
based on site characteristics and neighborhood relationships. Suitability
was defined as a combination of enablers and constraints on future
growth. The primary enablers of urban growth are proximity to existing
urban areas and transportation infrastructure, while the primary con-
straints on urban growth are slope, water, and exclusionary land use zon-
ing (Clarke et al., 1997). Population growth determined the area of
suitable land that might be urbanized, as this growth is the primary driver
of urbanization. Once the suitability rankings were mapped, additional
populated areas based on measured urban densities and projected urban
growth were allocated to successively decreasing suitability levels. Next,
the stages of suitability calculation are described, followed by the method
of urban area allocation.
Appendix 4 167
Step-By-Step Description of Map Construction
Model Space
• Cells were defined as 28.5 meter by 28.5 meter squares, following work
done by the Lincoln Land Institute (Angel et al., 2010).
• All data sets were projected into the Albers Equal Area Conic projec-
tion and aligned to a digital elevation model raster.
• All vector data were converted to raster data.
• The modeling extent was defined by a 20-kilometer buffer around the
urban area from the most recent time period.
• All operations were executed in Python programming language or the
ArcGIS 10 software suite using tools from ArcGIS 10 (Esri, 2011) and
Python scripts.
Definition of Urban Extent over Time
• Primary input data sets: existing urban areas were defined through
digitization of satellite imagery for 31 cities for three time periods (as
close to 1990, 2000, and 2010 as possible for each city).
• Delineation steps:
– The boundaries for each of the cities were visually identified for
three different dates circa 1990, 2000, and 2010 from satellite imag-
ery. The satellite imagery was Landsat Thematic Mapper or Landsat
Enhanced Thematic Mapper (depending on the year) imagery data.
These imagery data have a 30-meter spatial resolution with seven
multispectral bands. The date and type of imagery varied between
the cities. Table A4.1 at the end of this Appendix has the exact date
and type of imagery that was used in the analysis. Also, each city
boundary was projected in the local Universal Transverse Mercator
(WGS 84, UTM) projection. The UTM projection for each city is
also listed in Table A4.1.
– Within the imagery, each built-up portion of the city (where build-
ings were visible) was identified and then a polygon was drawn
around the visually identified area. The polygon was extended to
areas that were within 1.5 kilometers of the neighboring area. There-
fore, if a visually identified area of buildings had a gap between it and
the next area went out over 1.5 kilometers, then these areas where
not added to the city polygon. However if it was within 1.5 kilome-
ters, then the area was included.
168 The Future of Water in African Cities
• The final outputs of this stage were vector polygons for each target
time period (1990, 2000, 2010) showing the urban extent during that
time period. See Table A4.1 for information about the imagery files
used in classification.
Influence of Existing Urban Areas
• Primary input data sets: urban extent delineations from previous step.
• Modeling steps:
– Concentric zones were defined around the most recent urban extent
at distances of 5, 10, 25, and 60 cells (some cities used 40 cells as the
outer extent).
– Cells in each zone were weighted less the further away from the
existing urban areas they fell. These weights were 100 for the closest
ring, 75 for the next closest ring, 50 for the next closest, and 25 for
the outer ring.
• The final output of this step is the urban influence raster.
Influence of Roads
• Primary input data sets: Open Street Map data from 2011 (Open Street
Map, 2011). All roads were considered regardless of type, except for
Capetown, Durban, and Johannesburg, where road data comes from
the Africa Infrastructure Country Diagnostic (AICD N.D.).
• Modeling steps:
– The influence of roads was determined by the location of existing
roads, the cells immediately adjacent to existing roads, and cells
within a certain buffer distance of existing roads.
– Road cells themselves were defined as unsuitable for urbanization.
– A buffer distance of 15 cells was used to define near-road areas.
• The final outputs of this stage were three rasters—roads, next-to-roads,
and near-roads.
Influence of Slope
• Primary input data sets: 90-meter void-filled Shuttle Radar Topo-
graphic Mission (SRTM) digital elevation model developed by the
Consultative Group on International Agricultural Research (CGIAR)
(Jarvis et al., 2008), except for Accra, where HydroSHEDS 90-meter
void-filled SRTM data (Lehner et al., 2008) were used.
• Modeling steps:
– Slope was calculated from the elevation data and resampled to
model resolution.
Appendix 4 169
– Cells with slopes greater than a 100 percent rise were considered
impossible to urbanize (that is, they were given a weight of 0).
– Cells between 25 percent and 100 percent were considered difficult
to urbanize, and were given a weight of 50.
– Cells between 0 and 25 percent were considered easy to urbanize,
and given a weight of 100.
• The final output of this stage was a raster classifying pixels into one of
three categories of slope.
Influence of Exclusions
• Exclusions were defined as national parks, water bodies, and existing
urban areas.
• Primary input data sets:
– Water: Modis IGBP (International Geosphere-Biosphere Pro-
gramme) landcover data (Schneider et al., 2009) and the Global
Reservoirs and Dams database (Lehner et al., 2011);
– National parks: World Protected Areas Database (IUCN and UNEP,
2010), specifically all reserves with IUCN category status of Ia, Ib, or
II;
• Modeling steps:
– Resample landcover data set to model resolution.
– Convert landcover to binary water/not-water classification.
– Convert reservoir data to binary water/not-water classification.
– Combine these with urban/not urban classification.
• The result of this stage is a raster of cells showing where development
will be excluded, which was binary and thus not weighted.
Calculating Suitability
• The derived data sets from the previous sets (urban, slope, road, next-
to-road, near-road, and exclusions) were combined with relative
weights to create a suitability score for each pixel.
• The equation used was:
– Suitability = Exclude × (SlopeWeight × Slope + UrbanWeight ×
Urban + (RoadNextWeight × Next-To-Road + RoadNearWeight ×
Near-Road)).
– The following user-defined parameters were used for all runs:
o SlopeWeight = 50
o UrbanWeight = 100
o RoadNextWeight = 75
o RoadNearWeight = 40.
170 The Future of Water in African Cities
Allocating a New Urban Area
• The final stage calculated a potential new area needed to accommodate
anticipated population growth at a defined density.
• Input data sets:
– Urban density: Demographia measurements of urban densities in
2011 (Demographia, 2011) for all 31 cities. Lacking consistent data,
these density levels were assumed to remain unchanged into 2025.
– Population growth: UN population estimates (UNDESA, 2011) for
all cities but Blantyre were obtained for 2010 and 2025. Blantyre
was too small, so Demographia data on Blantyre’s 2010 population
was combined with Lilongwe’s annual growth rate to generate
approximate population pressure into 2025.
• Modeling steps:
– Calculate expected population change.
– Use density to calculate new urban area needed in number of cells.
– Starting with highest suitability, assign all cells as urban in each
decreasing suitability level until reaching the suitability level where
there are more cells available than required to meet demand.
– For this final suitability level, allocate remaining needed pixels ran-
domly.
Post-Modeling Adjustments to Displayed Urban Growth Potential
• Cut out rivers.
• Erase obvious places where effect of barrier was not captured.
• In cases where area was allocated to large areas of low suitability, creat-
ing very dispersed random placements of pixels, this last stage of alloca-
tion was removed from the final display (see room for improvement).
Post-Modeling Adjustments to Displayed Urban Growth Potential
In general, the results of this modeling exercise should not be taken as
assertions about where development will happen, but rather those areas
most likely to be urbanized by 2025. The goal was to use some basic spa-
tial characteristics of the context of each city to develop a more nuanced
picture of where growth may happen than could be achieved just by
assuming uniform expansion. Some specific limitations include:
• Roads
– We did not model changes in transportation infrastructure.
Appendix 4 171
– We did not model roads as rigorous spatial networks, so travel times
along roads could not be considered, only straight-line distances to
existing urban areas.
– We did not consider differences in road quality.
• Urban area
– The choice of what constitutes “urban” involves some subjectivity
when doing digitization from imagery, as well as choosing which
satellite settlements to include as part of the named urban area.
• Barriers
– Barriers to the movement of people, such as a large river, were not
incorporated in the model, as the primary consideration was prox-
imity, not how a cell would be accessed.
• Water
– The landcover data used is very coarse, and may miss smaller water
bodies that are nonetheless important for constraining urbanization.
• Zoning
– Land use conversions are highly dependent on regulatory regimes
such as land use zoning, which could not be captured here because
of inadequate, comparable data.
• Other exclusions
– Due to a lack of consistent and readily available data, we did not
include swamps or airports in these simulations.
Room for Improvement
As this modeling effort was a quick pass to illustrate the potential impact
that future urbanization could have on nearby water resources, there
are a number of ways that the approach and results could be improved.
Particular options include:
• Sensitivity analysis and more rigorous parameterization would improve
the model.
• As a last step of urbanization allocation, the lowest suitability zone to
be urbanized has pixels assigned randomly, with no consideration of
spatial neighborhood—it would be better to preference cells closer to
the city for random allocation (that is, density decay grid).
• Road quality, such as the difference between roads connecting second-
ary cities versus minor roads, may provide additional spatial information
on the suitability of urban growth along these different types of roads.
172 The Future of Water in African Cities
• The change of density can be calculated from the UN population data
for the urban extent year of record. A density measure of the historical
data could provide additional insight and better parameterization of
the model for projecting future changes.
Additional Cartographic References
In addition, each map made cartographic use of the following data sets:
Dams and reservoirs:
Lehner, B., C. Reidy Liermann, C. Revenga, C. Vörösmarty, B. Fekete,
P. Crouzet, P. Döll, M. Endejan, K. Frenken, J. Magome C. Nilsson, J.
Robertson, R. Rödel, N. Sindorf, and D. Wisser. 2011. “High Resolution
Mapping of the World’s Reservoirs and Dams for Sustainable River Flow
Management.” Frontiers in Ecology and the Environment 9(9): 494–502.
Available online at http://www.gwsp.org/85.html.
Rivers:
For South African cities: Rivers of South Africa at 1:500000 Scale.
Resource Quality Services and Chief Directorate of National Geo-Spatial
Information, Department of Water Affairs, Republic of South Africa.
Last updated in 2003. Available online at http://www.dwaf.gov.za/iwqs/
gis_data/river/All.html.
For all other cities: USGS HydroSHEDS, described in Lehner, B., K.
Verdin, and A. Jarvis. 2008. “New Global Hydrography Derived from
Spaceborne Elevation Data.” Eos, Transactions, AGU 89(10): 93–94.
Available online at http://www.worldwildlife.org/hydrosheds.
Major towns:
Location and population data set created by MaxMind, available from
http://www.maxmind.com/; using population data compiled by World
Gazetteer, and available online at http://world-gazetteer.com/.
Political boundaries:
National boundaries come from Esri’s World Countries data set. Last
updated November 2011. Available online at http://www.arcgis.com/
home/item.html?id=3864c63872d84aec91933618e3815dd2.
Appendix 4 173
Illustrating the Modeling Process, using Kumasi, Ghana, as an
Example*
Figure A4.1 Initial Landscape
Figure A4.2 Influence of Roads: Next-to-Road
* For the urban extent maps of all 31 cities, see http://water.worldbank.org/AfricaIUWM.
174 The Future of Water in African Cities
Figure A4.3 Influence of Roads: Near Road
Figure A4.4 Influence of Slope: Slope Ranking
Appendix 4 175
Figure A4.5 Excluded Areas: Exclusion Binary Ranking
Figure A4.6 Suitability Ranking for Urbanization
176 The Future of Water in African Cities
Figure A4.7 New Area Allocated to Urban Use
Figure A4.8 Final Map of Urban Extent for Kumasi, Ghana
177
Table A4.1 Imagery Used to Classify Urban Extent for Each City and Each Time Period
Country City Target period Path Row Date Platform Quality
UTM
projection
Angola Luanda 1990 182 66 6/30/91 Landsat 4-5 TM Good 33N
Angola Luanda 2000 182 66 6/14/00 Landsat 7 Good 33N
Angola Luanda 2010 182 66 7/14/08 Landsat 5 TM Good 33N
Benin Cotonou 1990 191 56 12/27/90 n.a. Poor - clouds 31N
Benin Cotonou 1990 192 55 06/04/91 n.a. Good 31N
Benin Cotonou 1990 192 56 06/04/91 n.a. Good 31N
Benin Cotonou 2000 192 55 12/13/00 n.a. Good 31N
Benin Cotonou 2000 192 56 12/13/00 n.a. Poor - opaque 31N
Benin Cotonou 2000 191 56 12/09/01 n.a. Good 31N
Benin Cotonou 2010 191 56 10/02/11 n.a. Poor - clouds +
striping
31N
Benin Cotonou 2010 192 55 12/12/11 n.a. Poor - opaque +
striping
31N
Benin Cotonou 2010 192 56 12/12/11 n.a. Poor - opaque +
striping
31N
Burkina Faso Ouagadougou 1990 195 51 9/10/90 Landsat 4-5 TM Good 30N
Burkina Faso Ouagadougou 2000 195 51 2/14/01 Landsat 7 Good 30N
Burkina Faso Ouagadougou 2010 195 51 1/20/10 Landsat 5 TM Good 30N
Cameroon Douala 1990 186 57 12/21/86 n.a. Good 32N
Cameroon Douala 2000 186 57 12/17/99 n.a. Poor - clouds 32N
Cameroon Douala 2010 n.a. n.a. 01/01/12 Bing Imagery Not rated 32N
Cameroon Yaoundé 1990 185 57 03/30/87 n.a. Good 33N
Cameroon Yaoundé 2000 185 57 05/18/00 n.a. Good 33N
Cameroon Yaoundé 2010 185 57 12/11/11 n.a. Good 33N
(continued on next page)
178
Country City Target period Path Row Date Platform Quality
UTM
projection
Côte d’Ivoire Abidjan 1990 196 56 12/24/88 n.a. Poor - clouds 30N
Côte d’Ivoire Abidjan 1990 195 56 01/02/89 n.a. Good 30N
Côte d’Ivoire Abidjan 2000 196 56 12/31/02 n.a. Good 30N
Côte d’Ivoire Abidjan 2000 195 56 02/26/03 n.a. Poor - clouds 30N
Côte d’Ivoire Abidjan 2010 195 56 01/15/11 n.a. Poor - opaque +
striping
30N
Côte d’Ivoire Abidjan 2010 196 56 02/10/12 n.a. Poor - opaque +
striping
30N
Democratic
Republic of
Congo
Kinshasa 1990 182 63 01/10/87 n.a. Poor - clouds 33N
Democratic
Republic of
Congo
Kinshasa 1990 182 63 08/09/94 n.a. Poor - opaque +
clouds
33N
Democratic
Republic of
Congo
Kinshasa 2000 182 63 09/02/00 n.a. Poor - opaque +
clouds
33N
Democratic
Republic of
Congo
Kinshasa 2000 182 63 04/30/01 n.a. Good 33N
(continued on next page)
Table A4.1 (continued)
179
Country City Target period Path Row Date Platform Quality
UTM
projection
Democratic
Republic of
Congo
Kinshasa 2010 182 63 08/16/11 n.a. Poor - clouds +
striping
33N
Democratic
Republic of
Congo
Lubumbashi 1990 173 68 6/15/91 Landsat 4-5 TM Good 35N
Democratic
Republic of
Congo
Lubumbashi 2000 173 68 5/1/01 Landsat 7 Good 35N
Democratic
Republic of
Congo
Lubumbashi 2010 173 68 7/2/09 Landsat 5 TM Good 35N
Democratic
Republic of
Congo
Mbuji-Mayi 1990 176 64 9/8/91 Landsat 4-5 TM Good 34N
Democratic
Republic of
Congo
Mbuji-Mayi 2000 176 64 6/7/01 Landsat 7 Good 34N
Democratic
Republic of
Congo
Mbuji-Mayi 2010 176 64 7/7/09 Landsat 5 TM Good 34N
(continued on next page)
Table A4.1 (continued)
180
Country City Target period Path Row Date Platform Quality
UTM
projection
Ethiopia Addis Ababa 1990 168 54 11/21/89 Landsat 4-5 TM Good 37N
Ethiopia Addis Ababa 2000 168 54 12/5/00 Landsat 7 Good 37N
Ethiopia Addis Ababa 2010 168 54 1/10/11 Landsat 5 TM Good 37N
Ghana Accra 1990 193 56 01/10/91 n.a. Poor - opaque 30N
Ghana Accra 2000 193 56 12/26/02 n.a. Good 30N
Ghana Accra 2010 193 56 01/17/11 n.a. Good 30N
Ghana Kumasi 1990 194 55 01/11/86 n.a. Good 30N
Ghana Kumasi 2000 194 55 05/07/02 n.a. Good 30N
Ghana Kumasi 2010 194 55 02/06/10 n.a. Good 30N
Guinea Conakry 1990 202 53 12/24/90 Landsat 4-5 TM Good 28N
Guinea Conakry 2000 202 53 12/19/00 Landsat 7 Good 28N
Guinea Conakry 2010 202 53 2/22/10 Landsat 5 TM Good 28N
Kenya Nairobi 1990 168 61 10/17/88 Landsat 4-5 TM Good 37N
Kenya Nairobi 2000 168 61 2/5/00 Landsat 7 Good 37N
Kenya Nairobi 2010 168 61 8/19/10 Landsat 5 TM Good 37N
Malawi Blantyre 1990 167 71 7/1/89 Landsat 4-5 TM Good 36N
Malawi Blantyre 2000 167 71 5/26/02 Landsat 7 Good 36N
Malawi Blantyre 2010 167 71 6/6/09 Landsat 5 TM Good 36N
Malawi Lilongwe 1990 168 70 5/11/91 Landsat 4-5 TM Good 36N
(continued on next page)
Table A4.1 (continued)
181
(continued on next page)
Table A4.1 (continued)
Country City Target period Path Row Date Platform Quality
UTM
projection
Malawi Lilongwe 2000 168 70 7/17/01 Landsat 7 Good 36N
Malawi Lilongwe 2010 168 70 8/22/11 Landsat 5 TM Good 36N
Mozambique Maputo 1990 167 78 6/5/91 Landsat 4-5 TM Good 36N
Mozambique Maputo 2000 167 78 5/7/01 Landsat 7 Good 36N
Mozambique Maputo 2010 167 78 9/7/08 Landsat 5 TM Good 36N
Nigeria Abuja 1990 189 54 12/21/87 n.a. Good 32N
Nigeria Abuja 2000 189 54 01/09/01 n.a. Good 32N
Nigeria Abuja 2010 189 54 01/18/10 n.a. Good 32N
Nigeria Ibadan 1990 191 55 04/12/86 n.a. Good 30N
Nigeria Ibadan 2000 191 55 02/06/00 n.a. Poor - opaque 30N
Nigeria Ibadan 2010 191 55 01/03/11 n.a. Good 30N
Nigeria Kano 1990 188 52 2/16/88 Landsat 4-5 TM Good 32N
Nigeria Kano 2000 188 52 10/28/99 Landsat 8 Good 32N
Nigeria Kano 2010 NA NA 1/1/12 Bing Imagery Good 32N
Nigeria Lagos 1990 191 55 12/18/84 n.a. Good 31N
Nigeria Lagos 1990 191 56 12/27/90 n.a. Poor - clouds 31N
Nigeria Lagos 1990 191 55 12/27/90 n.a. Poor - clouds 31N
Nigeria Lagos 2000 191 56 12/09/01 n.a. Good 31N
Nigeria Lagos 2000 191 55 12/09/01 n.a. Good 31N
Nigeria Lagos 2010 191 56 10/02/11 n.a. Poor - clouds +
striping
31N
182
Country City Target period Path Row Date Platform Quality
UTM
projection
Nigeria Lagos 2010 191 55 01/06/12 n.a. Poor - opaque +
striping
31N
Republic of
Congo
Brazzaville 1990 182 63 01/10/87 n.a. Poor - clouds 33N
Republic of
Congo
Brazzaville 2000 182 63 09/02/00 n.a. Poor - opaque +
clouds
33N
Republic of
Congo
Brazzaville 2000 182 63 04/30/01 n.a. Good 33N
Republic of
Congo
Brazzaville 2010 182 63 08/16/11 n.a. Poor - clouds +
striping
33N
Senegal Dakar 1990 205 50 10/15/89 Landsat 4-5 TM Good 28N
Senegal Dakar 2000 205 50 11/4/99 Landsat 7 Good 28N
Senegal Dakar 2010 205 50 10/12/11 Landsat 5 TM Good 28N
South Africa Cape Town 1990 175 84 6/29/91 Landsat 4-5 TM Good 34N
South Africa Cape Town 2000 175 84 6/13/00 Landsat 7 Good 34N
South Africa Cape Town 2010 175 84 4/17/11 Landsat 5 TM Good 34N
South Africa Durban 1990 168 81 4/9/91 Landsat 4-5 TM Good 36N
South Africa Durban 2000 168 81 6/28/00 Landsat 7 Good 36N
South Africa Durban 2010 168 81 5/12/09 Landsat 5 TM Good 36N
South Africa Johannesburg 1990 170 78 5/6/90 Landsat 4-5 TM Good 35N
(continued on next page)
Table A4.1 (continued)
183
Table A4.1 (continued)
Country City Target period Path Row Date Platform Quality
UTM
projection
South Africa Johannesburg 2000 170 78 1/7/02 Landsat 7 Good 35N
South Africa Johannesburg 2010 170 78 5/26/09 Landsat 5 TM Good 35N
Sudan Khartoum 1990 173 49 12/10/89 Landsat 4-5 TM Good 36N
Sudan Khartoum 2000 173 49 12/24/00 Landsat 7 Good 36N
Sudan Khartoum 2010 173 49 1/26/10 Landsat 5 TM Good 36N
Tanzania Dar es Salaam 1990 166 65 12/09/89 n.a. Poor - clouds 37N
Tanzania Dar es Salaam 2000 166 65 02/09/01 n.a. Poor - clouds 37N
Tanzania Dar es Salaam 2010 166 65 07/01/09 n.a. Good 37N
Uganda Kampala 1990 171 60 2/27/89 Landsat 4-5 TM Good 36N
Uganda Kampala 2000 171 60 11/27/01 Landsat 7 Good 36N
Uganda Kampala 2010 171 60 1/28/10 Landsat 5 TM Good 36N
Zambia Lusaka 1990 172 71 3/1/90 Landsat 4-5 TM Good 35N
Zambia Lusaka 2000 172 71 2/22/02 Landsat 7 Good 35N
Zambia Lusaka 2010 172 71 5/30/11 Landsat 5 TM Good 35N
Zimbabwe Harare 1990 170 72 6/23/90 Landsat 4-5 TM Good 36N
Zimbabwe Harare 2000 170 72 9/30/00 Landsat 7 Good 36N
Zimbabwe Harare 2010 170 72 5/26/09 Landsat 5 TM Good 36N
Source: World Bank.
Note: In some cases, multiple images from slightly different dates were used to classify a single urban extent for a target period. n.a. = not available.
