The production of flood vulnerability in Accra,Ghana: Re-thinking flooding and informal urbanisation

Abstract

Chinese

To date, understandings of flood vulnerability in African cities have been conceptually and practically limited. A dominant focus on the geophysical and biophysical causes of flood events restricts the analyses to the source of flood events and does not pay critical attention to the internal actors, dynamics and processes of informal urbanisation where the burden of flood impacts usually fall. This paper challenges these analyses by approaching the problem of flood vulnerability through an understanding of informal urbanisation. The study is based on experiences from selected informal communities in Accra, Ghana. Drawing on mixed qualitative methods including community focus group discussions, hazard victims’ interviews and institutional consultations/surveys, the study reveals that flood vulnerability in informal settlements has co-evolved with the dynamics of informal urbanisation and dwelling processes. The paper therefore makes a case for a re-look at the epistemology and ontology of urban flooding in rapidly and informally urbanising areas in the Global South.

Keywords

Accra flood management flood vulnerability informality urbanisation

Introduction

In cities of the developing world, informality, poverty and human vulnerability to flood events are inextricably intertwined (Few, 2003). For instance, in many countries in Asia and Africa, flood vulnerability in urban areas has been somewhat associated with informal urbanisation, urban poverty and marginalisation and occupation in flood zones (Gencer, 2013). However, informal urbanisation and flood vulnerability are usually researched separately, with minimal or no linkages. The few studies that focus on the relationship between informality and flood vulnerability are particularly in South-East Asia and to some extent Latin America (Braun and Aßheuer, 2011; Chatterjee, 2010; Zoleta-Nantes, 2000, 2002) Many studies on urban informality and flood vulnerability in African cities have treated these two rather related themes differently (Amoani et al., 2012; Appeaning-Addo and Adeyemi, 2013). At best, these research works only focus on the impacts of flood hazards on poor slum dwellers, perceived as passive victims with no control over what happens to them (Parker, 2000; Wisner et al., 2004).

To date, studies that connect flood vulnerability to informal urbanisation as a fundamental process in cities of the developing world are limited in the literature (Zoleta-Nantes, 2000). Issues such as unclear land tenure systems, uncontrolled urban land use development, urban poverty and marginalisation are usually not associated with flood vulnerability in cities of the developing world. Informal urbanisation processes are rather seen as internal dynamics in the face of an uncontrolled global threat over which poor urban households have no control (Few, 2003). A number of empirical studies can be referred to for emphasis (Braun and Aßheuer, 2011; Gencer, 2013; Sakijege et al., 2012).

Most studies on urban flood vulnerability in the developing world appear to overlook the contextual dynamics of informal urbanism and its role as both determinant and consequence of flood hazards in that part of the world. How therefore can we understand flood vulnerability as a complex socio-material and political process among informal urbanites in African cities? This paper is of the view that this question can be answered by approaching flood vulnerability in African cities through an understanding of their informal urbanisation processes. The paper argues that flood vulnerability in informal communities must first be contextualised and understood by examining how urban informality is produced and creates vulnerabilities among slum dwellers. As a result, instead of focusing on the geophysical and biophysical causes of flooding, we conceptualise flood vulnerability in informal settlements in Accra as part of the multiple outcomes of the city’s rapid informal growth.

Study context and entry points

Accra, the capital of Ghana, has been selected as the primary case because of its long history of flood hazards with several devastating events recorded since the 1950s (Karley, 2009; Rain et al., 2011); the latest being in June 2015 which killed over 150, displaced and injured several others (UN Country Team (UNCT) Ghana, 2015). Many such flood events have occurred in informal settlements (Douglas et al., 2008). As a result, the three flood-prone informal communities of Agbogbloshie, Old Fadama and Glefe were selected for the study (see Figure 1) because of their informality and vulnerability to yearly flood events.

Figure 1.

Flood areas in Accra with the locations of the study communities.

The selection of these communities was purposively done in consultation with officials of Accra Metropolitan Assembly (AMA), metropolitan office of the National Disaster Management Organisation (NADMO); Hydrological Services Department (HSD) and the Water Research Institute – the key institutions in charge of physical development and flood management in Accra. By their varied tenure securities, Agbogbloshie, Old Fadama and Glefe can be classified into the three main types of informal settlements in Accra (AMA and UN Habitat, 2011: 8–10; WRI (Water Research Institute), 2011: 13–18) as follows: (1) Agbogbloshie is a state recognised indigenous Ga¹ settlement which has developed into a slum; (2) Old Fadama, a spillover settlement from Agbogbloshie, is unrecognised by the state and seen as ‘illegal’ by both the traditional land owners and state institutions (Farouk and Owusu, 2012; Grant, 2006) and (3) Glefe is a peri-urban informal coastal settlement with unclear state recognition growing as a result of Accra’s rapid physical expansion.

Agbogbloshie and Old Fadama are exposed to similar flood hazards (Figure 1). Bounded by the Odaw River and Korle Lagoon draining Accra central (Figure 1), the two communities are vulnerable to perennial flood hazards from the overflow of the two water bodies. The two communities are on about 146 ha of land comprising 31 ha of state lands on which Old Fadama is located and 115 ha of customary lands on which Agbogbloshie is located (COHRE, 2004). In terms of demography, the 2010 Ghana’s population and housing census pegs the populations of the two communities as 10,000 and 24,164, respectively for Agbogbloshie and Old Fadama. However, Farouk and Owusu (2012), in a community-driven enumeration study at Old Fadama in 2004, 2006, 2007 and 2009, gave the community’s population as 79,684 as at 2009. Again, officials of the Accra Metropolitan Assembly (AMA) interviewed estimated the current population of the two communities as ‘over 50,000 inhabitants’.² According to the Metropolitan Planning Director, this population figure keeps changing ‘because of the community’s role as one of the main entry points for unskilled rural–urban migrants into Accra’.

This population is said to have doubled over the last decade because of an unprecedented rate of migration of young people, mostly unskilled girls between the ages of 10 and 30 years from northern Ghana into Accra in search of economic opportunities. The two communities were studied together because of their contiguous locations and exposure to similar flood risks.

Glefe is sandwiched between the confluence of two lagoons; Gbugbe and Gyatakpo to the north and the Atlantic Ocean to the south (Figure 1); and exposed to perennial flood hazards from storm surges and overflow of two lagoons. It is located on a 40-ha customary land³ under the ownership of the Sempe stool.⁴ The most current population of the community was captured in 2010 as 3969 at an annual growth rate of 5% (GSS, 2012). The community is a heterogeneous settler community with a myriad of ethnic groups from all over the country, notably: Ewes, Adas, Akans, Ga Damgmes and some ethnic groups from northern Ghana.

Approach and methods

Mixed qualitative methods were adopted and used for the study. These include interviews, focus group discussions, field observations and document reviews. Forty officials of 18 institutions were purposively sampled and interviewed. Eleven institutions related to urban management and flooding had been selected before the fieldwork; while seven others were selected and interviewed based on snowball sampling resulting from interviews with the pre-selected institutions and informal interactions within the study communities (Table 1). Some flood victims, members of traditional authorities, land owners and other community leaders were also purposively selected and interviewed.

Table 1.

List of institutions interviewed.

Pre-selected institutions	No. of officers
National Disaster Management Organization	2
Town and Country Planning Department	5
Environmental Protection Agency	3
Ghana Meteorological Agency( Accra Metropolitan Office)	2
Centre for Remote Sensing and Geographic Information Systems	3
Department of Urban Roads (Accra Metropolitan Office)	2
Water Research Institute, (Surface Water Division)	2
Hydrological Services Department	2
Accra Metropolitan Assembly	2
Drainage Maintenance Unit of AMA	2
Water Resources Commission	2
Institutions selected through snowball sampling
Urban Development Consultants working in study communities	2
Researchers in the Three Ghanaian Universities	3
UNDP (UN Habitat, Accra Office)	1
People’s Dialogue on Human Settlements	1
Ghana Federation of the Urban Poor (GHAFUP)	2
Old Fadama Development Association (OLDFADA)	2
Glefe Community Development Association	2
Total	40

Source: Authors’ fieldwork in Accra, Ghana, April–August, 2014.

Officials were interviewed based on their functional and operational knowledge on the topic under investigation. Their interviews revealed a number of relevant documents which were also reviewed as part of the field data collected for the study (Table 2).

Table 2.

Documents collected and reviewed.

	Documents obtained and reviewed
Government/state institutions
Accra Metropolitan Assembly	• Medium-Term Development plan of AMA – 2008–2013 • Participatory Slum Upgrading and Prevention – 2011
Metropolitan Town and Country Planning Department	• Greater Accra Metropolitan Area Strategic Plan – 1991/1992 • National Urban Policy Framework – 2012 • Ghana National Urban Policy Action Plan – 2012
Environmental Protection Agency	• Ghana Environmental Action Plan – 1991 • Climate Change Vulnerability and Adaptation Assessment on Water Resources of Ghana – 2000 • Climate Change Vulnerability and Adaptation Assessment of Coastal Zone of Ghana – 2000 • Ghana State of Environment Report – 2004 • Report on Climate Change Impact Assessment in Ghana – 2008 • National Environmental Sanitation Strategy and Action Plan – 2010
National Disaster Management Organisation	• National Disaster Management Plan – 2010 Disaster Annual Action Plan for Year 2014
Water Resources Commission	• Ghana National Water Policy – 2007 National Buffer Zone Policy for Managing River Basins in Ghana – 2008
Water Research Institute	• Update of Flood Phenomenon in Ghana – Draft Report 2011
Non-state institutions
UH Habitat/ United Nation Development Programme	• Rapid Disaster Waste Management Assessment – 26 • October Flash Flooding, Central Accra, Ghana – 2011
USAID	• Ghana Climate Change Vulnerability and Adaptation Assessment – 2011
Glefe Community Development Association	• Letters of correspondence and minutes of meetings

Source: Authors’ fieldwork in Accra, Ghana, April–August, 2014.

At the community level, six major focus group discussions were carried out, two in each community. There was an average of 25 participants at each session – between the ages of 20 and 70 years and dominated by males. Discussions at these meetings focused on the following: perceptions, experiences and local knowledge of causes and impacts of flooding in the communities. Other methods used include mini workshops of between four and five participants, field observations and informal interactions. A total of 18 of these mini workshops were held as follows: five in Agbogbloshie; ten in Old Fadama; and three in Glefe covering a total of 80 flood victims between the ages of 30 and 70 years – also male dominated. Field observations took the form of transect walks, picture taking of flooded areas, and informal interactions with some flood-affected households.

Flooding, informality and human vulnerability in Accra

Three main types of flood hazards were identified by the National Disaster Management Organization (NADMO) and the Water Research Institute (WRI) as affecting the city of Accra. These are: (1) localised flooding resulting from inadequate drainage and poor waste management; (2) flooding from overflow of streams, rivers and lagoons into settlements built in waterways and reclaimed wetlands; and (3) coastal flooding from the sea through storm surges and tidal waves. All the three types of flood hazards fit in with informal communities owing to their processes of growth, locations and internal structures.

Accra’s exposure and vulnerability to flood hazards has been a major concern to city authorities and residents alike. As a result, many government-sponsored studies and consultancy services have been carried out on the subject, some of these reports were obtained as shown in Table 2. These documents have sought to outline the various perceived and actual causes and impacts of flooding and responses by victims and state institutions. However, these works have not shown an understanding of how informal urbanisation has shaped flood vulnerability and flood risk in the city. Accra’s flooding has been associated with the intensity and frequency of rainfall and occupation on flood plains (Amoako and Frimpong Boamah, 2015; Karley, 2009).

Over the last 15 years, Accra has experienced some of the worst flood events. The city experienced serious damage to life and properties caused by floods between 2001 and 2005. The widespread flooding caused by the torrential rain in June 2001 left 12 people dead and over 100,000 homeless, and seriously disrupted traffic in Accra (Karley, 2009: 27). Accra’s floods in June 2004 and March 2005 also washed away sections of the city’s main roads (WRI, 2011: 16). The city also recorded another devastating flood event in June 2007.The local media reported five deaths in Gbawe, Santa Maria and Kwashiebu – all informal settlements (WRI, 2011). Other affected informal settlements include: Sowutuom, Sakaman and Ofankor. According to NADMO officials interviewed, the June 2007 also displaced over 13,000. Studies by the Accra Metropolitan Assembly and UN Habitat also showed that there have been increases in communities and residents located in flood-prone zones since 2000, because of rapid informal urbanisation (AMA and UN Habitat, 2011). It was estimated that flooding in July and August 2008 caused more than US$1 million worth of damage in Accra alone.

A monthly rainfall volume of 313.8 mm recorded in June 2009 also resulted in six deaths in some suburban areas in Accra (WRI, 2011: 16). Many residential areas and offices in the city were also flooded, destroying over US$2 million worth of assets. Another devastating flood event occurred in October 2011, killing 14 people and displacing 17,000 others (UNEP (United Nations Environment Program)/OCHA (Office for the Coordination of Humanitarian Affairs), 2011: 9). The most recent flood devastation was recorded in June 2015, killing over 150 people and injuring approximately 60 others at Kwame Nkrumah Circle, a central location in the city (UNCT Ghana, 2015). Between 2001 and 2015, flood hazards killed over 250 and displaced some 178,750 people. Over the same period, more than 70,000 informal livelihoods have been destroyed, running into about US$50 million (Amoako and Frimpong Boamah, 2015). Karley (2009) even values assets destroyed annually in Accra due to flood events as exceeding US$6 million. As a result the director of research at the NADMO headquarters described flooding as:

… the singular major cause of displacement, related deaths [direct and indirect], physical and material losses as well as economic destruction in Accra …

Most flood-affected communities in Accra are informal settlements (Douglas et al., 2008). Over 90% of flood-prone communities obtained from the NADMO are informal settlements. Residents of these informal settlements live in very poor and vulnerable physical and socio-economic conditions – along banks of major river catchments and lagoons in the city, fringes of unused industrial and dump sites among others (Amoako and Frimpong Boamah, 2015: 112). The AMA, with the support of the local office of UN Habitat, has identified 82 such informal communities within the city (AMA and UN Habitat 2011).These communities are home to some 1,652,374 people representing 38.4% of the city’s population. According to NADMO, an estimated 366,823 of residents of informal settlements forming about 22.2% live within unapproved flood zones and are therefore vulnerable to flood hazards (NADMO, 2010). This figure keeps increasing rapidly because of uncontrolled informal urbanisation.

Nonetheless, the impacts of flood hazards are more easily measured in high income and formal areas. The unavailability of data and uncoordinated data gathering processes make it difficult to estimate the actual economic damage costs of flood hazards in informal communities (Okyere et al., 2013: 63). Thus while damages because of floods in high-income areas are more tangible and easily measurable in economic terms, because they are usually insured (Karley, 2009: 37; Okyere et al., 2013: 62–63), flood vulnerability, impacts and losses within informal communities are less tangible and often not accurately measured.

Flood characterisation and measurement in informal communities

Like all other informal settlements in Accra, there are no weather stations within the study communities to provide accurate rainfall figures that cause floods, their frequencies and intensities. However, through shared memories of experiences with flood events, residents have various ways of measuring and comparing flood trends and intensities. Respondents in all the three study communities reported increasing flood events over the last two decades. Agbogbloshie and Old Fadama have experienced annual flood events since mid-1986 while Glefe’s first reported case of flooding was in 1995 (Table 3). The differences in periods of recorded flood events could be attributed to the differences in their locations, security of tenure and historical development. The three study communities had developed densely at different times under varied socio-political circumstances. Agbogbloshie dates back to the 1930s, had spilled over to Old Fadama in the early 1980s and the two communities developed rapidly through the 1980s to the early 1990s. Thus flood events in the area could be traceable to the rapid physical developments in the 1980s which led to encroachments on the catchments around the Odaw River and Korle Lagoon. Glefe developed in the 1970s as a small peri-urban community, was not known to be congested and flood-prone until the mid-1990s and early 2000s (Appeaning-Addo and Adeyemi, 2013).

Table 3.

Community reported cases of flooding and impacts of flood hazards.

Years of reported major flood events	Agbogbloshie/Old Fadama			Glefe
Nature	Reported causes	Major impacts/hazards	Nature	Reported causes	Major impacts/hazards
1986	Devastating	Heavy rainfall/ overflow of lagoon	Not reported	Not reported	Not reported	Not reported
1995	Devastating	Heavy rainfall	Six lives lost Houses destroyed Over 2000 people displaced	Devastating	Heavy rainfall Overflow of lagoon Storm surge	Houses destroyed Over 500 people displaced Household items destroyed
1999	Devastating	Heavy rainfall	Houses destroyed Over 3300 people displaced	Devastating	Heavy rainfall Overflow of lagoon Storm surge	Houses destroyed Over 300 households displaced Cholera outbreak
September 2001	Devastating	Heavy rainfall	Four lives lost Houses destroyed Many people displaced	Not reported	Not reported	Not reported
January 2002	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported
June 2002	Devastating	Heavy rainfall Overflow of Odaw River and Korle Lagoon	Houses destroyed Many people displaced Threat of eviction	Devastating	Heavy rainfall Overflow of lagoon Storm surge	Houses destroyed households displaced
January 2003	Normal	Heavy rainfall Overflow of Odaw River	Houses and household items destroyed	Normal	Heavy rainfall	Household items destroyed
June 2003	Normal	Heavy rainfall	Household items destroyed	Normal	Heavy rainfall	Household items destroyed
June 2004	Not reported	Not reported	Not reported	Normal	Heavy rainfall	Household items destroyed
March 2005	Not reported	Not reported	Not reported	Normal	Heavy rainfall	Household items destroyed
19 June 2009	Devastating	Heavy rainfall Overflow of Odaw River	Households displaced Cholera outbreak	Devastating	Heavy rainfall Overflow of lagoon Storm surge	Houses and households displaced
20^th June 2010	Devastating	Heavy rainfall	Houses destroyed Households displaced	Normal	Heavy rainfall	Household item destroyed
25/26 October 2011	Devastating	Heavy rainfall Overflow of Odaw River	Over 10,000 people affected	Devastating	Heavy rainfall Overflow of lagoons	Over 2000 people affected and death of a pregnant woman
31 May 2013 5/6 June 2013	Not reported	Not reported	Not reported	Normal	Heavy rainfall Storm surge	Houses destroyed and households displaced

Sources: Community focus group discussions at Agbogbloshie, Old Fadama and Glefe in Accra, Ghana July/August 2014.

The nature of flood events were described by respondents as resulting from moderate to very fast storm run-off carrying huge amounts of domestic and industrial waste from various parts of their communities. While the flood water presents major risk to household assets and human life, the transportation of waste raises concerns about waste management practices and the lack of storm drains in Accra. Based on the above qualitative criteria, community members categorised the reported flood events into ‘normal’ and ‘devastating’. Normal was explained as when no human life was lost, no major community infrastructure destroyed and residents returned to their daily livelihoods after the flood day. Flood events were described as ‘devastating’ if human lives were lost, households displaced, livelihoods destroyed and economic activities halted for a number of days and received attention from politicians and the media.

Perceived causes and impacts of flood events

Respondents in all the three communities indicated that rainfall or storm surges trigger all flood events. However, they attributed their exposure and vulnerability to flood hazards to the internal structure of their settlements. These are characterised by the manifestation of informality such as haphazard housing development, reclamation of wetlands for residential development and poor waste management practices. The three communities have grown on their own without adequate state interventions in infrastructure provision. Rooted within this argument is the neglect or absence of state and city institutions in the provision of flood management infrastructure and decent housing for slum dwellers (Amoako, 2016). Respondents also indicated that flood devastations are exacerbated by the lack of access to water and sanitation facilities. In a study by Douglas et al. (2008: 194) in Alajo, another neighbourhood in Accra, respondents associated flood events with ‘improper city planning, poor drainage, over population and lack of consultation by officials with the poor and insensitivity to their problems’. Thus the central role of informal urbanisation as influencing flood vulnerability in Accra is emphasised.

Aside the overflow of water bodies within their localities and storm surges along the shores of Glefe, there were no significant differences among the three communities, in what respondents’ perceived to be the immediate physical causes of flood events (Table 4). From Table 4, the perceived causes of floods can be summarised under three key frames. The first groups of perceived causes of floods are heavy rainfall and/or high tides in coastal communities. Respondents could not relate the amount and duration of rainfall to the occurrence of floods since there are no official rainfall data for their communities. Nonetheless officials of NADMO indicated that an average daily rainfall volume of 55 mm and above can result in major flood events in flood-prone informal settlements. This is an indication that flood vulnerability might not be directly related to the amount of rainfall recorded. Second, perceived causes of flood relate to lack of or inadequate municipal infrastructure and inadequate official responses to flood hazards – these are directly related to their informal growth and neglect of city authorities. And third, haphazard residential development in the form of building on reclaimed wetlands with poor housing materials and poor waste management practices. All the three broad perceived causes of floods mentioned are well known and discussed in the literature on flooding in Accra (Afeku, 2005; Karley, 2009; Rain et al., 2011). However, it is worthy to note that these causes have evolved and are continuously evolving with the processes of informal urbanisation and the struggle for social and economic spaces in the study communities.

Table 4.

Perceived causes of flooding and flood hazards.

Agbogbloshie and Old Fadama		Glefe
Ranked causes of flood hazards	No. of times mentioned	Ranked causes of flood hazards	No. of times mentioned
Heavy and unpredictable rainfall	102	Heavy and unpredictable rainfall	84
Overflow of river and lagoon	100	Dam and diversion drains of Pambros Salt Production Ltd	77
Neglect of city authorities and official insensitivity to communities’ problems	90	Overflow of lagoons and high tides	77
Poor or lack of storm drainage system	89	Neglect of city authorities and official insensitivity to communities’ problems	76
Poor urban planning	77	Sand mining along the community’s shoreline	76
Overpopulation and rapidly increasing population	76	Poor or lack of storm drainage system	73
Poor waste management practices	76	Reclamation of wetlands and building in waterways	52
Reclamation of wetlands and building in waterways	70	Poor waste management practices	52
Inability of government to complete the Korle Lagoon Ecological Restoration Project (KLERP)	55	Act of God	30
Act of God	38	−

Source: Victims’ interviews/mini workshops and focus group discussions at Agbogbloshie, Old Fadama and Glefe in Accra, Ghana July/August 2014.

Perceptions of flood causes among residents of informal settlements present a significant point of departure from the causes attributed to flooding by city officials. Officials of the NADMO, AMA, EPA and the Metropolitan Planning office associated flood occurrences to three main causes: illegal/unapproved occupation in flood plains; poor management of Accra’s surface water resources and impacts of climate change. Encroachment in waterways is ranked top on the officially documented causes of flooding in Accra (AMA and UN Habitat, 2011: 8). For example a report by the WRI (2011: 21) states that Accra’s ‘flooding problems have been aggravated by building and road structures that act as barriers to floodwaters hence contributing to persistent flooding’. A commissioned report by the United Nations Environment Programme (UNEP) after the devastating floods of 25/26 October 2011 also ranked unauthorised occupation on wetlands and waste disposal into rivers and lagoons as the major causes of flood hazards in the city (UNEP/OCHA, 2011: 12). An official at NADMO’s headquarters also indicated that:

Almost all informal settlements are located at hazardous and flood prone areas [of the metropolis]. These settlements are usually located on unapproved, unplanned or illegally acquired parcels of land [not recognised by the city authorities].

One officer at the metropolitan town and country planning department indicated that building in waterways and buffer zones of water bodies are the main breaches of development control and major causes of flood events in informal settlements. For this officer, ‘urban poverty and lack of affordable housing cannot be tangible excuses for flouting planning regulations and living in flood zones’. He likens it to ‘justifying robbery with poverty’.

Closely related to occupation of flood zones is the issue of poor management of surface water bodies in Accra. Many officials blamed the inability of state and city authorities to manage the Odaw River, Korle and Sakumo lagoons on informal and/or illegal occupation along their banks. For these officials, demolition, forced eviction and relocation of these informal settlements should be the main interventions for Accra’s flooding problem. This shows the contrasting views between city officials and slum residents on the role of formal planning and the processes of informal urbanisation in shaping flood vulnerability. City authorities prescribe formal planning as the only solution to flooding in Accra while ignoring the broader processes and contexts of plural urban land management systems and informal urbanisation. On the other hand, residents in the study communities associate flooding with strict regulations of urban planning and refusal of city authorities to provide appropriate flood management infrastructure in informal communities.

The third officially endorsed cause of floods, which departs from the perceptions of residents in the study communities, is climate change. Over the last decade several academic and government studies/reports have made references to the role of climate change in flood occurrences in Accra (Cudjoe et al., 2014; Douglas et al., 2008; EPA, 2008; Rain et al., 2011). This has come under enormous debate in the literature on flooding in Accra (Afeku, 2005; Karley, 2009). However responses received in the study communities appear to challenge the prevailing view that climate change is a cause of increased flooding because their local experiences tell them otherwise. For instance at a mini workshop a female participant at Old Fadama asked: ‘Does it matter if the climate changes or not if you live at a place like this?’. After which she made a statement that appears to have received endorsement from the other participants:

When there are rains in Accra, there are only selected communities in the city that experience flood hazards, this means that the problem is caused by what is happening in those communities and not changes in weather conditions. (Hajia, 50, flood victim and unit committee member)

Community perceptions on climate change/variability and its role in flood hazards are still those heard from officials of NGOs, city authorities and the local media. However, they do not necessarily link it directly to increasing episodes of flood hazards. Only a few respondents attributed flood events to climate change/variability (Table 4).

Ironically, flood impacts in the three communities also appear to have shaped residents’ exposure and vulnerability to more floods. Both residents and officials of city government institutions recounted flood impacts that further marginalise residents of informal settlements. Experiences of flood impacts appear rather similar in the three communities (see Table 5). From Table 5 all reported flood impacts could be grouped into three main frames: destruction of lives, informal livelihood sources and housing structures. The key link between these flood impacts and informality is that almost all residential structures in the three communities also double as places of informal economic activities and daily livelihoods. Heavy flood events destroy wooden and concrete structures used by poor households for residential and business purposes rendering them homeless and jobless. Unpaved alleys and paths in between individual houses are flooded and rendered inaccessible; cutting off sections of the settlements for days or even weeks. Reported indirect impacts of flooding include threats of forced evictions from AMA and outbreak of water-borne diseases. Floods expose residents of the informal settlements, particularly children, to malaria, typhoid, cholera and other diseases. Transportation into the communities is curtailed for days while school children are unable to go to school.

Table 5.

Ranked community-level impacts of flood hazards.

Agbogbloshie and Old Fadama		Glefe
Reported flood impacts	Sections of community affected	Reported flood impacts	Sections of community affected
Loss of human and animal lives	Flood victims’ household	Community cut off – no movement in or out for any purpose	Entire community
Justification for eviction of Old Fadama by City authorities	Entire community	Loss of land along shoreline and reduction of community land size	Entire community
Destruction of storm drains and community roads	Entire community	Destruction of private properties along the shoreline and within community	Affected victims or households
Flooding of community toilets and pollution of domestic water	Entire community	Loss of wetland (lagoons) because of erosion along shoreline	Entire community
Transportation of liquid and solid waste through community	Entire community	Flooding of educational and health facilities	Mostly women and children of school age
Outbreak of water-borne diseases	Entire community	Loss of home-based livelihoods	Affected victims or households
Flooding of markets and transport terminals	Affected victims and owners of business in the market, drivers, vehicle owners, etc.	Socio-political tension/conflicts between community leadership and salt production company and other stakeholders	Affected victims and Pambros Salt Production company
Destruction of livelihoods	Affected victims or households	Choked gutters and blocked access roads	Entire community
Theft and looting by rescuers in the market and private homes	Affected victims or households	Presence of pools of stagnant water leading to outbreak of diseases	Entire community
Close down of community health and educational facilities	Mostly women and children of school age	Filthy and smelly environment through the transportation of garbage	Entire community
Flooding of private houses	Affected victims or households	Social stigmatisation of residents	Entire community

Source: Community focus group discussions at Agbogbloshie, Old Fadama and Glefe in Accra, Ghana July/August 2014.

Two major flood events were identified as having the most devastating impacts over the last decade in the study communities. Participants at Agbogbloshie and Old Fadama identified the October 2011 floods as the worst in modern times – these claimed 14 lives, destroyed several informal livelihoods and rendered thousands of households homeless in Agbogbloshie and Old Fadama (UNEP/OCHA, 2011). The most affected areas were the surrounding markets and e-waste hub, where several livelihoods were lost. A young man in his late 20s at Old Fadama who was affected narrated his personal experience as:

I used to sell second hand clothes. When it rained at the time, I lost everything in my room … everything. The worst part was that I had just got some second hand clothes which I was sorting and making ready to sell. The dirty flood water soaked and destroyed all of them. I lost everything. Since then I have been out of business.

In the case of Glefe, the worst flood event was reported to have occurred in June 2013 and had rendered several households along the community’s shoreline homeless. Many displaced households sought refuge with friends and relatives at unaffected parts of the community. Glefe was cut off for a day without any movements to or from the community. There was neither rescue nor evacuation of affected victims over the period. Though very devastating, to the residents of the community, the event largely went unnoticed by city authorities except for some media reports. Again, attention is drawn to the informal growth of the three communities and how that has shaped their vulnerability to flood hazards and vice versa.

The production of informality and flood vulnerability at Agbogbloshie, Old Fadama and Glefe

Exposure and vulnerability of the three communities to flood hazards are similar in characteristics and can be attributed to their informal growth at flood-prone zones without planning interventions. This is well explained by their historical development, political geographies and informality. The growth, socio-cultural, political and land management processes have played important roles in producing and shaping the exposure and vulnerability of Agbogbloshie, Old Fadama and Glefe to perennial flood events in Accra.

Agbogbloshie dates back to the 1930s under the British colonial administration (Cudjoe et al., 2014) and developed adjacent to the then European town which is now the CBD of Accra. The historical development of the community reveals spatial conflict between customary land ownership and formal urban planning under colonial administration. The dialectic development between formally planned European town and unplanned native town of Agbogbloshie marked the beginning of a dichotomy between formal and informal communities. Although Agbogbloshie was accepted by the colonial government as ‘legitimate/formal’ indigenous settlement it did not receive any formal planning under the British colonial government. Grant and Yankson (2003: 67) reported that the area became ‘crowded, cluttered, and congested with poor structures and unhealthy conditions … attracted local and rural migrants and was physically separated’ from the European town. Two main reasons given for the uncontrolled growth of Agbogbloshie are: the colonial Government’s ‘neglect of urban planning’ (Grant and Yankson, 2003: 67) and indigenous ‘rejection of urban planning’ as a result of their traditional and spiritual attachments to the Korle Lagoon (Grant, 2006: 8). As a result, the indigenes resisted the colonial government’s proposal to compulsorily acquire and convert the area around the lagoon into harbour facilities and develop the adjoining community into an industrial and commercial hub (Grant, 2006: 8).

The period after Ghana’s independence in 1957 saw the compulsory acquisition of land at Agbogbloshie for various public uses such as markets, railway station and other industrial and commercial uses. Later in 1961 the independent government acquired the Korle Lagoon and the areas around it for what it claimed to be the ‘Korle Lagoon Development’ project (Grant, 2006: 9). However, the project was abandoned after the overthrow of the government in 1966. Still without formal planning, Agbogbloshie which was recognised as indigenous Ga settlement under customary land administration started developing into a slum community surrounded by several informal residential and commercial developments. The study revealed that the overall effects of these historical growth factors on Agbogbloshie have been the reduction in the community’s land size, inadequate community infrastructure and city authorities’ claim of illegal spillover of the community into the catchment of the Odaw River and Korle Lagoon and hence prone to perennial flood hazards from their overflow during rainy seasons.

The failed attempt of Ghana’s first independent government to implement the ‘Korle Lagoon Development’ Project and the inability of successive governments to complete it or find appropriate use for the compulsorily acquired land around the lagoon gave birth to the Old Fadama community. Thus Old Fadama had developed as spillover from Agbogbloshie in the early 1980s (Afenah, 2009: 12) and has been condemned by city authorities and a section of the local media as an ‘illegal’ squatter settlement (Grant, 2006: 2). The settlement evolved on the unused government land acquired in 1961 for the development of the Korle Lagoon into an ecological tourist destination. The study revealed that the growth of Old Fadama during the period between the early 1980s and late 1990s was largely influenced by the decision of the state to ‘temporarily’ locate Ghanaian returnees deported from Nigeria, relocate street hawkers in Accra for the Non-Aligned Movement (NAM) conference, Accra’s main local yam market and displaced ethnic conflict victims from the northern region of Ghana in the Area. The community is now a densely populated area made up of self-built wooden shacks, kiosks and lacks adequate community infrastructure. Owing to its location along the banks of Odaw River and Korle Lagoon without planning and provision of storm management infrastructure, Old Fadama is exposed and vulnerable to annual floods during the rainy season.

Apart from the overflow of the Korle Lagoon and Odaw River as a trigger of annual floods, exposure and vulnerability of residents of Agbogbloshie and Old Fadama is also caused by the generally low elevation of the area and residential developments on lands reclaimed from the Korle Lagoon using ‘saw dust’ and other materials such as plastic bags, cardboard and plywood. In addition, haphazard housing constructions mainly made from wood, plywood and aluminium/iron roofing sheets make the communities increasingly vulnerable to flood risks and hazards.

Glefe started as a small peri-urban fishing community in the mid-1960s and has grown into one of the most vulnerable informal coastal communities in Accra (Appeaning-Addo and Adeyemi, 2013). It was a fishing community with houses made from coconut fronds. Owing to its informal growth and unclear recognition by city authorities, there are contested population figures and historical development narratives. For instance, the first census to recognise the community was in 1984 during which the population of Glefe was recorded as 978. However, older members of the community interviewed refuted this claim, indicating that their community dates back to the period immediately after Ghana’s independence in 1957 and that their population would have been higher than what was recorded in 1984.

Over the years Glefe has grown without planning into one of the most hazardous informal communities in Accra, manifesting such informal characteristics as haphazard housing development, reclamation of wetlands for residential development and poor waste management practices (Amoani et al., 2012; Appeaning-Addo and Adeyemi, 2013). The settlement also has problems of low elevation, haphazard physical development, sea erosion and tidal waves. Aside these geophysical factors, annual flood events have been attributed to the community’s closeness to an officially declared wetland known locally as Densu Wetlands⁵ (Amoani et al., 2012).

From the foregoing, Glefe’s vulnerability to floods can also be attributed to its informal growth in a hazardous zone. Over the last four decades the community has grown without any land use plan and clear security of tenure among residents. Aside its unplanned informal growth, Glefe’s flood vulnerability is shaped by two main factors. First is a booming local sand mining industry along the shores of the community. This industry has been the alternative informal economic activity to the local fishing industry since the early 1970s and is said to have been a major attraction for the community’s population growth during the period (Amoani et al., 2012). The second factor is salt production by a local company, Pambros Salt Production company Ltd, which has dammed one of the lagoons for its operations. Residents attribute the annual overflows of the lagoons to this dam. In contrast, city authorities argue that the settlement must either be replanned or relocated because of its hazardous location. Traditional leadership of the area, who have customary title over the land, do not agree to any of these options. None of these proposed interventions by the Accra Metropolitan Assembly have been implemented.

From the foregoing, human vulnerability to flood hazards at Agbogbloshie, Old Fadama and Glefe cannot be seen just as a product of their physical location and geophysical occurrences but also as a product of their historical development, informal growth, politics of space and urban citizenship. Although flood events in the three communities have been associated with rainfall and tidal waves (Amoako and Frimpong Boamah, 2015; Rain et al., 2011) flood vulnerability in the study communities are produced and shaped by the processes underlying their informal and uncontrolled growth in flood-prone zones. For instance, Grant (2006: 8) recounted the prevention of formal planning in the then native town, now Agbogbloshie, owing to traditional attachments of the early native Gas because of their ‘designation of the Korle Lagoon and its tributaries as sacred’. The lack of urban planning coupled with rapid influx of local and rural migrants to stay and do business in the area have substantially influenced their informal development in flood-prone zones. A key result of this is the informal spillover into Old Fadama within the catchments of the Odaw River and Korle Lagoon producing flood vulnerability in the two communities. The case of Glefe represents informal development without clear tenure arrangements. Hence the three communities’ exposure and vulnerability to floods are connected to their informal growth processes and security of tenure.

Though different in locations, historical developments and political geographies, the three communities have largely developed outside the formal planning system that lays so much emphasis on tenure security to the neglect of affordable housing for slum dwellers. Not only have their locations close to wetlands influenced flood vulnerability, but the dynamics of their informal growth without city authorities’ intervention is also fundamental to the production of flood vulnerability among the residents of the three communities. The interactions of these factors produce what Cutter (1996: 532) refers to as ‘vulnerability as tempered response’. This view highlights the ‘social construction of vulnerability as a condition rooted in the historical, cultural, social and economic processes’ that have shaped not only the growth of the communities but ‘their capacity to anticipate, cope with, resist and recover from the impact of a natural disaster’ (Wisner et al., 2004: 11).

Conclusions

In many African countries, city authorities approach urban flood vulnerability among residents of informal settlements in two main ways: First, as caused by residents who have defied planning regulations and built sub-standard houses in flood-prone zones; and second, as received by residents who are passive victims of flood hazards because of their disadvantaged locations in the city. Based on these two views, city authorities in African countries have used forced evictions and distribution of relief items as their two main flood management approaches (UN Habitat, 2003, 2007). This paper challenged both these analyses and approaches to flood vulnerability by taking a critical look at the nexus between informal urbanism and flood vulnerability. In doing so, we argue that vulnerability to flood hazards in urban informal settlements in developing countries is a complex network of political and socio-cultural factors rooted in their historical development and indigenous land management systems that together produce informal urban spaces.

The link between informal urbanisation and flood vulnerability in the three communities can be understood through the complex overlaps between Ghana’s customary land administration and the state’s partially functional urban planning systems. Interactions between indigenous land ownership and formal urban planning in Accra, produce and shape informal settlements and their vulnerability to flood hazards. We are of the view that the inability of the city’s administrative system to mainstream the customary land market, deal with housing supply deficits, provide affordable housing and community infrastructure has contributed directly to the development of informal settlements in flood-prone areas.

City authorities in Accra have been overwhelmed by the rapid informal developments in flood zones. This has heightened the perception that failures or inefficiencies in urban governance are both causes and consequences of flood vulnerability in informal communities. This view had been explained earlier by Satterthwaite et al. (2007: 49) who attribute the lack of sustainable flood management in cities of the developing world to the deliberate attempt by city governments to ignore slum dwellers because of their illegal/unclear status and inadequate data to demonstrate the extent of the problem. This approach also appears to be informed by the judgement of informality as ‘evil’ and the key setback to urban development in the Global South.

Within the broader literature and conceptual framing of informal urbanisation in the Global South, this signifies the production of urban informality, marginalisation and vulnerability by the failure of urban planning system to produce environmentally safe spaces for poor urban citizens (Roy, 2005, 2011; Watson, 2009: 176–177). Watson (2009: 177) points out how planning regulations have been used ‘opportunistically’ to justify forced evictions and marginalisation of poor urban households and slum dwellers. These actions of city governments through formal planning regulations do not only push/relegate poor people to hazardous marginal lands but also worsen their vulnerability to current and future environmental hazards. The locations and growth of informal settlements without infrastructure and the continuous struggle of poor urban households for affordable housing and livelihoods are central in shaping the causes of flooding in all the three communities – regardless of their different securities of tenure.

Accra’s flood vulnerability among residents of informal settlements also depicts a direct relationship between growing urban poverty and vulnerability as discussed by Few (2003) and Gencer (2013). Few (2003: 48) asserts that in many urban areas of the developing world, the ‘most vulnerable are low-income people, [rural–urban] migrants, those living in flimsy houses … or living near drainage channels …’ Thus in most cities of developing countries, spatial marginalisation and vulnerability to environmental hazards are closely related. However, Chan and Parker (1996: 314) caution that though poverty and vulnerability may not always be related, poor urban dwellers are more likely to be adversely affected by natural hazards – especially floods and fire as was the case in some part of Accra in June 2015 mentioned earlier.

The crust of our argument is not to ignore the importance of biophysical factors such as rainfall, overflow of rivers and lagoons, and high tides as important in triggering flood events, but to unearth the complexities of ways in which socio-material and political processes of informal urbanisation in urban areas may shape flood vulnerability of residents. These concerns bring to the fore the argument for a re-look at flood vulnerability research and intervention approaches to consider urban informality as a key ontological and epistemological lens. Cannon (2000: 45–46) for example pushes for analysis of flood vulnerability not only as an exposure to a specific hazardous event but also as a condition of people that derives from their political-economic position. He argues that if residents of flooded communities are perceived only or mainly as victims then the problem of what causes vulnerability may be evaded (Cannon, 2000: 45). Situated in this political-economic frame of flood vulnerability, this paper has revealed that while flood events are associated with geophysical occurrences, flood vulnerability in informal communities are produced and shaped by complex socio-material, cultural and political processes that underlie their informal urbanisation.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

References

Accra Metropolitan Assembly (AMA) and UN Habitat (2011) Participatory Slum Upgrading and Prevention: Millennium City of Accra, Ghana. Accra: UN Habitat.

Afeku

(2005) Urbanization and flooding in Accra, Ghana. Master’s Thesis, Department of Geography, Miami University, USA.

Afenah

(2009) Conceptualizing the Effects of Neoliberal Urban Policies on Housing Rights: An Analysis of the Attempted Unlawful Forced Eviction of an Informal Settlement in Accra, Ghana. DPU Working Paper (Vol. 139). London: UCL Development Planning Unit.

Amoako

(2016) Brutal presence or convenient absence: The role of the state in the politics of flooding in informal Accra. Geoforum 77: 5–16.

Amoako

Frimpong Boamah

(2015) The three-dimensional causes of flooding in Accra, Ghana. International Journal of Urban Sustainable Development 7(1): 109–129.

Amoani

Appeaning-Addo

Laryea

(2012) Short-term shoreline evolution trend assessment: A case study in Glefe, Ghana. Journal of Disaster Risk Studies 4(1): 1–7.

Appeaning-Addo

Adeyemi

(2013) Assessing the impact of sea-level rise on a vulnerable coastal community in Accra, Ghana. Journal of Disaster Risk Studies 5(8): 1–8. doi: 10.4102/jamba.v5i1.60.

Braun

Aßheuer

(2011) Floods in megacity environments: Vulnerability and coping strategies of slum dwellers in Dhaka/Bangladesh. Natural Hazards 58(2): 771–787.

Cannon

(2000) Vulnerability analysis and disasters. In: Parker

(ed.) Floods. London: Routledge, pp. 45–55.

10.

Center on Housing Rights and Evictions (COHRE) (2004) A Precarious Future: The Informal Settlement of Agbogbloshie. Accra: COHRE.

11.

Chan

Parker

(1996) Response to dynamic flood hazard factors in peninsular Malaysia. The Geographical Journal 162(3): 313–325.

12.

Chatterjee

(2010) Slum dwellers response to flooding events in the megacities of India. Mitigation and Adaptation Strategies for Global Change 15(1): 337–353.

13.

Cudjoe

NAS

Owusu

Burkett

(2014) Perceptions, experiences and indigenous knowledge of climate change and variability: The case of Accra, a sub-Saharan African city. Regional Environmental Change 14(1): 369–383.

14.

Cutter

(1996) Vulnerability to environmental hazards. Progress in Human Geography 20(4): 529–539.

15.

Douglas

Alam

Maghenda

et al . (2008) Unjust waters: Climate change, flooding and the urban poor in Africa. Environment and Urbanization 20(1): 187–205.

16.

Environmental Protection Agency (EPA) (2008) Report on Climate Change Impacts Assessment in Ghana. Accra: Netherlands Climate Change Studies Assistance Programme (NCAP 2).

17.

Farouk

Owusu

(2012) If in doubt, count: The role of community-driven enumerations in blocking eviction in Old Fadama, Accra. Environment and Urbanization 24(1): 47–57.

18.

Few

(2003) Flooding vulnerability and coping mechanisms: local responses to a global threat. Progress in Development Studies 3: 43–58.

19.

Gencer

(2013) The Interplay between Urban Development, Vulnerability, and Risk Management. New York: Springer.

20.

Ghana Statistical Service (GSS) (2012) 2010 Population and Housing Census Report. Accra: Ghana Statistical Service.

21.

Grant

(2006) Out of place? Global citizens in local places: A study of informal settlements in the Korle Lagoon Environs in Accra, Ghana. Urban Forum 17(1): 1–24.

22.

Grant

Yankson

(2003) City profile: Accra. Cities 3(1): 17–29.

23.

Karley

(2009) Flooding and physical planning in urban areas in West Africa: Situational analysis of Accra, Ghana. Theoretical and Empirical Researches in Urban Management 13(4): 25–41.

24.

National Disaster Management Organization (NADMO) (2010) National Disaster Management Plan. Accra: NADMO and UNDP.

25.

Okyere

Yacouba

Gilgenbach

(2013) The problem of annual occurrences of floods in Accra: An integration of hydrological, economic and political perspectives. Theoretical and Empirical Researches in Urban Management 8(2): 45–79.

26.

Parker

(2000) Floods. Abingdon: Routledge.

27.

Rain

Engstrom

Ludlow

et al . (2011) Accra Ghana: A City Vulnerable to Flooding and Drought-Induced Migration. Case study prepared for Cities and Climate Change Global Report on Human Settlements. Available at: http://www.unhabitat.org/grhs/2011.

28.

Roy

(2005) Urban informality: Toward an epistemology of planning. Journal of the American Planning Association 71(2): 147–158.

29.

Roy

(2011) Slumdog cities rethinking subaltern urbanism. International Journal of Urban and Regional Research 35(2): 223–238.

30.

Sakijege

Lupala

Sheuya

(2012) Flooding, flood risks and coping strategies in urban informal residential areas: The case of Keko Machungwa, Dar es Salaam, Tanzania. Journal of Disaster Risk Studies 4(1): 46–56.

31.

Satterthwaite

Huq

Reid

et al . (2007) Adapting to Climate Change in Urban Areas: The Possibilities and Constraints in Low- and Middle-Income Nations. Human Settlements Discussion Paper, London: IIED.

32.

UN Country Team Ghana (UNCT) (2015) Ghana – Floods Situation Report. Accra: United Nations – Ghana.

33.

UN Habitat (2003) The Challenge of Slums: Global Report on Human Settlements. London: Earthscan Publication Ltd.

34.

UN Habitat (2007) Enhancing Urban Safety and Security; Global Report on Human Settlements 2007. London: Earthscan Publication Ltd.

35.

UN Habitat (2011) Global Report on Human Settlements 2011: Cities and Climate Change. London: United Nations Human Settlements Programme (UN Habitat).

36.

United Nations Environment Program (UNEP) and Office for the Coordination of Humanitarian Affairs (OCHA) (2011) Rapid Disaster Waste Management Assessment of 26th October Flash Flooding, Central Accra, Ghana. Geneva: UNEP/OCHA Environment Unit.

37.

Water Research Institute (WRI) (2011) Prevention of Hydro-Hazard Disasters in Ghana: Update of Flood Phenomenon in Ghana – Draft Report. Accra: UNESCO Cluster Office.

38.

Watson

(2009) The planned city sweeps the poor away …: Urban planning and 21st century urbanisation. Progress in Planning 72: 151–193.

39.

Wisner

Blackie

Cannon

et al . (2004) At Risk: Natural Hazards, People’s Vulnerability and Disaster. New York: Routledge.

40.

Zoleta-Nantes

(2000) Flood landscapes of Metro Manila. In: Liongson

Tabios

Castro

(eds) Pressure of Urbanization: Flood Control and Drainage in Metro Manila. Diliman: UPCIDS, pp. 35–49.

41.

Zoleta-Nantes

(2002) Differential impacts of flood hazards among the street children, the urban poor and residents of wealthy neighbourhoods in Metro Manila, Philippines. Mitigation and Adaptation Strategies for Global Change 7(1): 239–266.