Flood risk management in Africa
Darren Lumbroso
Abstract
The continent of Africa comprises 54 states, and its climatic conditions are incredibly diverse, ranging from equatorial to desert. Rainfall and river flows in Africa show high levels of variability across a range of spatial and temporal scales (Conway et al., 2009; Hamandawana, 2007; Laraque, Mahé, Orange, & Marieu, 2001; Sutcliffe & Knott, 1987). This poses several complex challenges for the management of floods on the continent. These range from estimating extreme flood flows in ungauged catchments in arid zones in north Africa and managing flood regimes in large, transboundary river basins such as the Niger and Zambezi to reducing the vulnerability to floods of the 238 million people in sub-Saharan Africa who live in informal settlements (United Nation [UN], 2019). Based on the number of people affected, over the past 30 years, floods and droughts are the two natural hazards that have the largest humanitarian impacts in Africa (Lumbroso, Brown, & Ranger, 2016). However, in the past decade, across Africa, floods have overtaken droughts in terms of the number of people that they impact. The 2018–2019 South-West Indian Ocean cyclone season resulted in a level of flood damage previously unseen in Africa (Masters, 2019). The main cause was Cyclone Idai, which primarily affected Mozambique and Zimbabwe. It commenced in March 2019 as a tropical depression over Malawi, which caused widespread flooding, affecting almost 1 million people. This moved back out to sea, forming Cyclone Idai, which hit the east coast of Mozambique before dissipating in eastern Zimbabwe (Masters, 2019). The storm resulted in at least 602 deaths in Mozambique, 299 fatalities in Zimbabwe (Centre for Research on the Epidemiology of Disasters, 2019), and caused damage to critical infrastructure estimated to be $1 billion (Hillard, 2019). This special issue of the Journal of Flood Risk Management on Africa brings together the eight papers that have been published in the past 12 years which explicitly focus on aspects of flood risk in Africa. While most people in Africa still live in mainly rural areas, the continent is rapidly urbanising. Africa's urban population is expected to almost triple from 548 million in 2018 to around 1.5 billion in 2050 (UN, 2018). In their paper, Egbinola, Olaniran, and Amanambu (2017) give an insight into the impacts of flooding on the city of Ibadan in Nigeria. A survey of stakeholders in Ibadan, carried out as part of this work, illustrates the devastating impact that floods have on their lives. Of the people surveyed, 96% indicated that, when a flood occurs, they are unable to save the contents of their house as they try to escape from the inundated areas. In Ibadan, as in many African cities, the blockage of watercourses and drainage ditches, as a result of the dumping of refuse, was perceived to be a significant factor in causing floods (Egbinola et al., 2017), an issue that is raised by Adelekan (2016) in his paper on flood risk management for the city of Lagos. Adelekan (2016) also found that, of the 1,000 households that were surveyed in Lagos, 80% do not benefit from any external assistance when flooding occurs and that the dissemination of flood warnings, as well as awareness of flood risks, needs to be improved. Urban flooding is a pressing issue on the continent and one that, in many African countries, requires transformative change, which will be challenging to deliver in the face of a changing climate. In their paper, Tazen et al. (2018) investigate the trends in flood events and their relationship to extreme rainfall in the city of Ouagadougou in Burkina Faso. In keeping with the work of Egbinola et al. and Adelekan, the results indicated that the significant increase in flood risk in Ouagadougou in recent years is not only related to extreme rainfall events but also a result of human and environmental factors. These include unplanned urbanisation accompanied by low levels of investment in flood-resilient infrastructure and flood management (Tazen et al., 2018). Gebremedhin et al.'s paper discusses a novel approach to flood modelling and mapping in Dar Es Salaam, Tanzania. This builds on community mapping projects in vulnerable areas of the city and uses OpenStreetMap as a data source (Gebremedhin, Basco-Carrera, Jonoski, Iliffe, & Winsemius, 2020). Rapid urbanisation of the city over the past 20 years has led to a significant increase in the flood risk. The engagement with at-risk communities at the commencement of and throughout flood modelling and mapping activities has led to a more accurate depiction of the urban flood risk than would otherwise have been achieved (Gebremedhin et al., 2020). Most cities and urban centres in Africa are now regarded as flood disaster risk hotspots (Baker, 2012). The unplanned urbanisation in Africa and the associated increase of people living in floodplains have led to an increase in the number of fatalities related to floods in African cities (Di Baldassarre et al., 2010). Urban floods cannot be managed in isolation at the city scale, and responses to ameliorate their impacts are complicated by interlinked political, socio-economic, and environmental changes (Di Baldassarre et al., 2010). Urban water management in most African countries is based on outmoded practices, with practitioners receiving very little continuous professional development. For Africa's rapidly urbanising cities, flood risk needs to be taken into consideration in formulating future plans to make cities more resilient and sustainable. Successful flood risk management requires that city governments develop clear, robust, and forward-looking strategic plans that are informed by rigorous research, administrative data gathering, dialogue with the public, evaluation, and learning (Egbinola et al., 2017). Currently, over half of Africa's population lives in rural areas, with many reliant on floodplains and rivers for their livelihoods. In wide floodplains, such as those found in the Zambezi River basin, where the population density is low, a structural approach to flood risk management cannot be justified on both environmental and economic grounds (Lumbroso, Ramsbottom, & Spaliveiro, 2008). In their paper, Lumbroso et al. outline non-structural solutions to reduce rural communities' vulnerability to floods in Mozambique by raising their awareness of floods and improving preparedness. These initiatives included the development of games to help educate schoolchildren about flooding. This allows students to gain an understanding of the important issues and share what they have learned with their parents (Lumbroso et al., 2008). Other adaptation actions, such as rainwater harvesting infrastructure to provide a source of clean water during floods, are useful to the communities not only in times of floods but also on a day-to-day basis (Lumbroso et al., 2008). There have been several studies which indicate that the damage to roads from climate change, relative to population and gross domestic product (GDP), will be higher in Africa than in any other region in the world (African Development Bank [AfDB], 2011). Much of the weather-related damage to road infrastructure is caused by floods. There are approximately 330,000 bridges in the existing road network across sub-Saharan Africa (Cervigni, Losos, Neumann, & Chinowsky, 2016). It has been estimated that 31% of these are critical links for primary or secondary roads (Cervigni et al., 2016). Although roads are the main means of transport in Mozambique, accounting for half of the freight traffic and 98% of the passenger traffic, the road density is low, and the network has few redundancies (Mozambique National Statistics Institute, 2015). This means that a loss of an individual road link, as a result of a flood event, often leaves areas without a viable road connection to markets or essential services. A recent study on the road network in Mozambique suggested that some of the worst effects of flooding are on the bridge components of road transport networks. During Cyclone Idai, an estimated 29% of Mozambique's national road network was damaged, and 20 bridges were destroyed (World Bank, 2019a). In his paper, Monsef (2018) investigates strategies to reduce flood risk to highways in Egypt. Increasing the flood resilience of roads on Egypt's Red Sea coast is particularly important given the rapid increase in the importance of tourism in the area over the past 30 years. The paper identifies flood adaptation measures for roads, which include the construction of a series of contour stone bunds in the upstream catchments, with attenuation dams further downstream (Monsef, 2018). These dams also act as artificial infiltration basins to help recharge aquifers, as well as to reduce erosion (Monsef, 2018). The impacts of soil degradation in Africa, not only on agricultural production but also on flood flows, have long been recognised (Feddema, 1998). Degradation and erosion of watersheds in Ethiopia, as a result of intensive cultivation, overgrazing, increases in population, and deforestation, have led to a significant decrease in crop yields and an increase in flood risk (Gebretsadik, 2014). In his paper, Gebretsadik recognises that, although there are significant institutional barriers, a participatory approach to watershed management would improve vegetation cover and water retention, reduce soil erosions, and reduce flooding (Gebretsadik, 2014). Estimating flood flows in ungauged catchments is an ongoing challenge for those working on flood risk management studies. The problem is particularly acute in many African countries where the hydro-meteorological networks have been in decline over the past 40 years. In their paper, Gericke and Smithers (2018) discuss an improved and consistent approach to estimating catchment response time parameters for ungauged catchments in South Africa. Their research found that large errors in estimates of peak flood discharge in medium to large catchments in South Africa can be largely ascribed to significant errors in the estimation of the catchment response time. This is mainly a consequence of the use of inappropriate time variables and the inadequate use of a simplified convolution process between rainfall runoff time variables. Their paper also highlights the sensitivity of design flows to estimated time parameter values in rainfall runoff methods in South Africa and the considerable effort that is required to ensure that time parameter estimations are representative and consistently estimated (Gericke & Smithers, 2018). There is clearly much work to be done in flood risk management in Africa. For example, a recent study by Lumbroso et al. (2016) found that, of the 54 sovereign states in Africa, there were only 7 where stakeholders perceived risk assessments for floods to be effective. Of these seven countries, South Africa has produced flood hazard maps for the areas of highest risk for at least the past two decades, and in Mozambique, since the floods of 2000, there has been a concerted effort to carry out flood mapping for the major river basins (Lumbroso et al., 2016; Lumbroso, Rance, Pearce, & Wade, 2014). The picture is similar for flood forecasting and early warning systems in that most African stakeholders do not perceive them to be particularly effective in reducing the humanitarian impacts of floods (Brown, Lumbroso, & Wade, 2014; Lumbroso, 2018; Lumbroso et al., 2016; Lumbroso, Rance, Pearce, Brown, & Wade, 2014; Lumbroso, Rance, Pearce, & Wade, 2014). Africa accounts for approximately 17% of the world's population, and this is projected to reach 23% by 2050. In 2019, the increase in the continent's GDP was expected to be 4%, higher than in other emerging and developing economies as a whole but lower than in China and India (AfDB, 2019). However, these figures are not reflected in the amount of research published by Africans. From 1996 to 2012, the number of research papers published in scientific journals with at least one African author more than quadrupled, from about 12,500 to over 52,000, and during the same time, the share of the world's articles with African authors almost doubled, from 1.2% to around 2.3% (Schemm, 2014). Despite this increase, the number of African peer-reviewed journal papers published is low. Unfortunately, this trend is also mirrored by the Journal of Flood Risk Management. The first issue of the Journal of Flood Risk Management was published in 2008. Of the approximately 700 papers published in the journal to date, only 8 of these, ~1%, have focused explicitly on Africa. There are numerous reasons for the relative paucity of African peer-reviewed journal papers. In 2011, when worldwide expenditure on research was 1.77% of the total global GDP, Kenya spent 0.1% of its GDP and South Africa spent 0.76% of its GDP on research (Kumwenda et al., 2017). The African Union has set a target of 1% of GDP to be invested on research and development, but data available show that only three sub-Saharan African countries are close to this target: South Africa, Kenya, and Senegal (around 0.8% in all three countries), and most African countries fall below 0.5% (United Nations Educational, Scientific and Cultural Organization, 2018). High-income countries such as Finland, Japan, Sweden, and the Netherlands have over 5,000 full-time equivalent researchers per million of the population, compared to sub-Saharan Africa where the highest number of researchers per million of the population is 494 in South Africa (World Bank, 2019b). Other challenges faced by researchers in Africa include scarcity of mentors, lack of funding, lack of writing skills, lack of motivation, and low demand for research by policymakers (Kumwenda et al., 2017). Many of the future flood risk management challenges that Africa faces would be most effectively met through research in Africa that is led by African institutions (Odume & Slaughter, 2018). However, this requires investment in the continent's universities to run water-related professional courses and degree programmes (Odume & Slaughter, 2018). To meet the continent's needs, these programmes should include localised practical development content and innovations to adapt technologies to fit the local social–cultural context and challenges. There are some promising signs. The African Research Universities Alliance was set up in 2015 to facilitate collaboration, knowledge transfer, equipment sharing, pooling of resources, and development of mutually beneficial partnerships across Africa. It has proposed the formation of Centres of Excellence across Africa related to key thematic areas including water (Odume & Slaughter, 2018). We would welcome more submissions to the Journal of Flood Risk Management on topics specifically related to flood risk management in Africa so that, in 10 years' time, the percentage of Africa-focused papers published significantly increases. The author acknowledges Dr Bob Andoh and Dr Paul Samuels for their comments and suggestions.