Pathways for resilience to climate change in African cities

African cities are largely less-built with agile informal settlements and multiple ecologies that harbor different pathways for resilience to climate change. We undertook a qualitative systematic review of academic and policy evidence, to address the question of what interventions are emerging at neigbourhood to city scale to enhance resilience to climate change in Africa. Resilience at neigbourhood scale often stems from harnessing the local resource base and technologies for urban agriculture and forestry; alternative energy from wastes; grassed drainages for protection against erosion; recreation along dry riverbeds; fog-water harvesting; and adjustments in irrigation schedules. At city scale, planning is targeted at buildings, mobility and energy service delivery as the objects to be made resilient. The review established that evidence on comparisons across regions is mainly on East, West and South African cities, and much less on cities in Northern and Central Africa. Ecological comparisons are majorly on coastal and inland cities, with minimal representation of semi-arid and mountainous cities. Resilience efforts in capital cities are the most dominant in the literature, with less emphasis on secondary cities and towns, which is necessary for a deeper understanding of the role played by inter-municipal and inter-metropolitan collaborations. African cities can bring context-sensitivity to global debates on climate resilience, if theoretical perspectives are generated from emerging interventions across case studies. We conclude with suggestions on what future research needs to take on, if evidence on resilience to climate change in African cities is to be strengthened.

In terms of regional differences in exposure, projections show increased precipitation in areas of Western Africa and Eastern Africa (Endris et al 2019, Odoulami et al 2019, and this has been linked to changes in the characteristics of the El Niño Southern Oscillation (ENSO) and Indian Ocean Dipole (IOD) for the case of northern and equatorial part of East Africa. Coastal areas are especially subject to flooding as a result of both precipitation events and sea level rise. An estimated 54 million Africans live in vulnerable Low Elevation Coastal Zones (LECZ)-defined as areas 10 meters or less above sea level-and this figure is projected to rise to over 100 million by 2030 (Becker et al 2019). As shown in figure 1, cities around rivers and creeks are susceptible to riverine flooding, though localized flooding occurs outside these areas as well, especially in more heavily developed settings , Ponte et al 2019, Soares et al 2019, Ziervogel 2019. In Bamenda (Cameroon), approximately 20% of the 250 000 residents live on floodplains and roughly 7% live in informal settlements on steep slopes. Land clearance for settlement and for quarrying and sand mining, along with other land-use changes caused by urban expansion, have further created serious problems of soil erosion. Soil that is washed down the hills blocks drainage channels and changes peak water flows. Degradation of the land has exacerbated problems with floods (Tume et al 2019). An estimated 17% of Mombasa's area (4600 hectares) could be submerged by a sea-level rise of 0 3 meters, with a larger area rendered uninhabitable or unusable for agriculture because of waterlogging and salt stress Sandy beaches, historic and cultural monuments, and several hotels, industries and port facilities would also be negatively affected. Mombasa already has a history of disasters related to climate extremes, including floods that have caused serious damage and loss of life nearly every year (Okaka and Odhiambo 2019).
An assessment of the vulnerability in the cities of Alexandria, Rosetta and Port Said (Egypt), suggests that with a sea-level rise of 50 cm, more than 2 million people would have to abandon their homes, 214 000 jobs would be lost, and the cost in lost property value and tourism income would be over US $35 billion, which does not include the immeasurable loss of world famous historic, cultural and archaeological sites (Elshazly et al 2019). In Banjul (Gambia), the city is less than 1 meter above sea level and flooding is common after heavy rains in settlements established on reclaimed land in dried-up valleys, and in settlements close to mangrove swamps and wetlands. Problems with flooding are likely to intensify under a warmer climate with an increase in the strength and frequency of tropical storms, coupled with a negative influence on fish assemblage in the Central River Region (CRR) (Gomez et al 2020). Although some important areas of Abidjan (Côte d'Ivoire) lie on a plateau and may escape the direct effects of sea-level rise, major economic centers, including the nation's largest port and much of the international airport, are on land less than 1 meter above sea level. A sea-level rise in Abidjan is likely to inundate 562 square kilometers along the coastline of the Abidjan region, as lowland Marshes and lagoons dominate the coastal zone. Average retreat will vary from 36 to 62 m (Anouman et al 2019).
Given the location in hot regions, African cities are likely to face increased exposure to excessive heat over the coming decades. Projections show that aggregate exposure in African cities will increase by a multiple of 20-52, reaching 86-217 billion persons per day by 2090, depending on the scenario. The most exposed cities are located in Western and Central Africa, although several East African cities indicate an increase of more than 2000 times the current level by 2090 (Rohat et al 2019). Average temperatures in North African cities are expected to rise between 1.5°a nd 3°Celsius by 2060, with a risk of increased desertification. The ElNiño-Southern Oscillation (ENSO) events that ran from October 2014 to June 2016, followed by a La Niña from July 2016 to January 2017, are responsible for heightening the drought conditions in the cities of Djibouti, Mogadishu and Addis Ababa, since the year 2017 (Qu et al (2019)). Coral reefs that have the potential of protecting human and marine life in Djibouti, are also vulnerable to rising sea surface temperatures (SSTs), wind and heat waves along the Gulf of Aden, which leads to coral bleaching and anthropogenic impacts (Cowburn et al 2019). High rates of urbanization coupled to drought and high demand for water in Tripoli-Libya, has made groundwater levels decline and exceed 12 m in the Southern area while in the Northern area near the coastal line, depletion is continuous and more than 70 wells will be confronted by saline water intrusion by the year 2100. The salinity levels in these wells will make the groundwater unfit for human consumption (Aswad et al 2019). An almost similar situation has occurred in Niamey-Niger, where ponds have exhibited a change in groundwater discharge behavior due to consecutive dry days and changes in the spatial pattern of rainfall (Boko et al 2020). Drought impacts on water availability and security have also been witnessed in Harare-Zimbabwe, where severe drought episodes have been observed in 1991-1992, 1994-1995, 2002-2003(Frischen et al 2020, Tanyanyiwa 2020. The studies presented indicate that although rapid onset disasters like cyclones often have devastating effects on informal settlement and African urban ecologies, slow onset climate events, such as drought and flush floods are also negatively impacting lives and infrastructures, which warrants responses from different disciplines and communities of policy and practice.
The framing and operationalization of resilience to address climatic change in cities Building resilience to climatic change in cities is closely linked to how the world will resolve the twin challenges of rapid urbanization and accelerated change in global temperature (Parnell 2016, Simon et al 2016, Holmstedt et al 2017, Patel et al 2017, Rybski et al 2017. The unpacking of pathways to climate resilience in cities, however, is usually dualistic in nature, viewing the resourcefulness of local community actors as a downstream approach that is parallel to, and sometimes conflicting with upstream interventions, which are influenced by municipal regulations, economic trends, technological advancements, geopolitics, and other global forces. The upstream lays emphasis on the role of multi-lateral and government agencies in brokering and facilitating leadership coalitions, through reforming urban policies across sectors, to enhance collaboration amongst municipalities, and deliver on sector-wide plans for recovery from  Conversely, the downstream approach usually centers on actions and networks that pervade the realm of government and municipal interventions. Such actors operate through neighborhood groups, citizen coalitions, associations of informal urban service delivery operators, and federations of city traders, who seek to coalesce their efforts towards re-building local communities in face of flooding and other consequences of extreme weather events. Examples include green roofing and rain water harvesting movements in urban informal settlements of cities in the global south, which are targeted at improved stormwater management, better regulation of building temperatures, reduced urban heat-island effects, while promoting the conservation of green spaces for reduced pollution and intensity of surface run-off (Oberndorfer et al 2007). The downstream approach has also been researched widely in the global north with concepts such as a 'first-responder city', which stem from studies that illuminate the opportunities associated with increasing the capacity of local communities to avoid, prepare for, and effectively respond to extreme weather events (Ebi 2011, Schmeltz et al 2013, Rosenzweig and Solecki 2014. The other component of studies on the downstream approach are onground volunteers and local nonprofits that influence urban development dynamics for enhanced community resilience, by for example disseminating early warning information to low-income urban residents, which is framed as the 'civic infrastructure' for resilient neigbourhoods by Graham et al (2016) and Elkin and Keenan (2018). Quantitative studies on downstream interventions have provided metrics for measuring the effectiveness of community-led efforts, by factoring in the extent of community engagement, emergency supplies, communication with neighbors, civic engagement, and collective efficacy (Eisenman et al 2016, Kwok et al 2018. What is common across the framing and operationalization of downstream and upstream approaches, is the notion that climate resilience in cities is an incremental, multi-actor and cross-scalar process. The incrementalists like Abdoumaliq (2016) and Dovey (2016), argue that resilience starts in what is gradual and integral for people with regard to the natural and built landscapes in which they live, the talents and skills of stakeholders chosen to represent them, rather than emphasis on what is lacking in terms of opportunities and resources. The interdependence amongst actors and their multiple decisions for climate resilience is key from a multi-actor perspective, where for example communities and inter-municipal collaborations take lead to build networks that catalyze efforts towards confronting extreme weather events (Giest and Howlett 2013, Fünfgeld 2015, Bansard et al (2017). Both incremental and multi-actor processes are characterized by interlocking and dynamic political, social, economic and spatial factors, which means that the hindrances to and enablers for climate resilience are cross-scalar in nature (Bahadur et al 2013, Bahadur and Tanner 2014, Tanner et al 2015. But advances in framing climate resilience in cities have not taken on a broader perspective when it comes to the contextual diversity and local priorities of urban regions in Africa. Cities of Africa are not only shaped by informality, which is the opposite of centralized planning systems in China and much of the global north (Bai et al 2014, Shatkin 2016, but are also characterized by multiple ecologies and infrastructures that perhaps require unique pathways to resilience in the face of climatic hazards (Schäffler andSwilling 2012, Fraser et al 2017). However, our understanding of African urban pathways to climate resilience has remained patchy and minimally represented, even in global assessments such as those by the Intergovernmental Panel on Climate Change (IPCC). This paper addressed this gap through a systematic review of evidence, from both science and policy, on the interventions that are emerging at neigbourhood to city scale, to enhance resilience to climate change in Africa.

Methods and Materials Approach
The search for academic literature was done qualitatively, as one of the methodological approaches deemed appropriate in understanding institutional, ecological and social dimensions of climate resilience in cities (Ford et al 2011, Berrang-Ford et al 2015. Academic material was generated from Scopus, using the keyword search strategy and in line with the review question. The review question is what interventions are emerging at neigbourhood to city scale to enhance resilience to climate change in Africa?The key words used are: Climate Change (CC), Climate Resilience (CR) and African Cities (AC). Literature published in languages other than English was not considered and therefore the material selected may not be representative of Anglophone and Francophone cities of Africa.

Inclusion and exclusion criteria
The key terms used in the search for material were: Climate Change (CC), Climate Resilience (CR) and African Cities (AC) (see supplementary file 1 available online at stacks.iop.org/ERL/15/073002/mmedia for String of Search Terms). The initial search result was 1051, covering the period of 2006-2019. But when subjected to the inclusion and exclusion criteria (figure 2), 95 academic articles appeared from Scopus for selection, 72 were excluded and 25 articles were included, based on their tittles, abstracts, texts, methodology and results. Papers were also included if they contained climate-related terms such as: flooding, storm surges, drought, landslides, heat waves, wind storms and wild fires. We also considered papers with abstracts, texts and results on West, East, South, Central, and North Africa, as a way of representing different geographical locations in the continent. The search further considered papers with ecological terms namely; in-land, coastal, semi-arid or arid and mountainous cities. For pathways to CR, we included papers with abstracts, texts and results on: (i) local community interventions; (ii) interfaces amongst different actors, (iii) climate change policies and climate change plans; and (iv) local and technological innovations. Papers with terms such as adaptation, mitigation, environmental change, sustainability, risk reduction, urban transitions and community resilience were also included the review (see supplementary data S1 of the search string table).
Titles, abstracts and texts were extracted and screened independently against the inclusion criteria by three reviewers, namely: Buyana Kareem (BK), Shuaib Lwasa (SL) and Jaqueline Walubwa (JW). The abstracts, texts, methodology and results of the included papers were reviewed using three complimentary strands: (1) identification of titles and abstracts on interventions for CR in AC; (2) internal validity of the material with regard to how well the methodology and results answered the research question(s) set in the paper being reviewed; and (3) external validity of material which refers to how well the methodology and results answered the review question. Each paper was reviewed independently twice, with disagreements resolved through discussion and consensus with the rest of the co-authors as third party reviewers. Full text articles were reviewed by Paul Mukwaya (PM) and Samuel Owuor (SO) for final decisions regarding inclusion, with disagreement resolved by consulting the other reviewers, that is; Denis Tugume (DT), Peter Kasaija (PK), Hakimu Sseviiri (HS), Gloria Nsangi (GN), and Disan Byarugaba (DB).

Quality appraisal
Quality appraisal of the included academic literature was undertaken using the ten-item Critical Appraisal Skills Programme (CASP) checklist for qualitative research (Critical Appraisal Skills Programme 2018). Despite its demerits, the CASP tool is commonly used in qualitative research (Hannes et al 2010). Three evaluators (BK, SL and JW) independently assessed the quality of each study with discrepancies resolved through consensus and discussion with third party evaluators (PM and SO). No studies were excluded on the basis of quality, as per the summary of the results in supplementary data S2.

Systematic mapping of policy and planning documents
Academic literature was triangulated with a purposive online search for climate change strategies and action plans mounted by African city authorities of government. The plans were purposively selected to achieve representation across regions of East, West, North and Southern Africa. The plans that were mapped are: (i) the City of Alexandria Energy and Climate Change Action Plan (North Africa); (ii) the Durban Climate Change Strategy (South Africa); (iii) the Kampala Climate Change Action Strategy (East Africa); and (iv) Lagos State Climate Change Adaptation Strategy (West Africa). Review of evidence from policies and plans focused on the objectives, key stakeholders, targets and object(s) to be made resilient as well as implementation programmes under each strategy and action plan. The documents were useful in discerning the notions and pathways to mainstreaming climate resilience in city planning and governance processes. City plans published in languages other than English were not considered in the mapping and therefore the evidence presented may not necessarily speak for Anglophone compared to Francophone cities.

Results
Resilience at neigbourhood scale means harnessing the local resource base and technologies Evidence indicates that resilience at neigbourhood scale often stems from harnessing the local resource base and available technologies. As shown in table 1 below, some of the key pathways to resilience are urban agriculture and forestry for alternative food and income sources (Lwasa et al 2015); household energy alternatives from wastes; and plot-level technologies of retention for runoff, using for example vetiver grass for protection against erosion (Eckart et al 2011). Recreation along fractured dry riverbeds has taken shape (Sareh et al 2016), coupled with reliance on ecosystems, such as mangroves and community reforestation along coastal areas, for defence against storm water surges (Roberts et al 2012). Fog-water harvesting is being promoted (Marzol and Sánchez 2008), alongside kitchen or toilet facilities that capture methane gas at the top of the dome for re-use as cooking gas  (Thorn et al 2015). Evidence revealed that resilience to climatic change in African cities means harnessing the positive interactions amongst interventions at neigbourhood and city scale. The positive interactions are the conditions for scaling out what is existing and emerging to wider scales. One of the manifestations of positive interactions, is the emphasis on urban forestry, which supports neigbourhoods on food security and jobs, while promoting tree planting closest to pedestrian walkways, buildings and parking lots at city scale, for a cooler climate, protection against surface run-off and dust storms.
City planning for resilience targets buildings, mobility, and energy service delivery but without explicit mechanisms on cross-sector collaboration Across all the city strategies and action plans in the review (table 2 below), the objects to be made resilient are mainly energy, mobility, and buildings, with a few means to civic engagement with local communities. The plans reviewed are: (i) the City of Alexandria Energy and Climate Change Action Plan (North Africa); (ii) the Durban Climate Change Strategy (South Africa); (iii) the Kampala Climate Change Action Strategy (East Africa); and (iv) Lagos State Climate Change Adaptation Strategy (West Africa). As indicated in table 2, all the overarching targets set by the city of Alexandria are centered on energy and buildings, namely: 'By 2030, all new buildings will be carbon neutral; by 2050, reduce GHG emissions by 80% below 2005 levels; by 2050, 80% of City's energy will be from clean, renewable sources.' In the case of Kampala city, there are targets that interconnect energy with mobility services. These are: 50% of motorists using mass public transport (Buses and Train); 50% of city roads tarmacked; 25 km of nonmotorized transport (cycle and Pedestrian) lane length constructed. Evidence on the implementation of plans revealed that emphasis is on the mandates of individual administrative structures at city and national scales, as the basis for division of responsibilities. Although the energy and mobility targets set by the cities are aligned to the goals of emission reduction and reconfiguration of urban service delivery, there is no explicit indication of building flexible management systems and inter-agency networks, which are usually suited to respond to shocks to energy and mobility urban systems (Frantzeskaki et al 2014).
The targets and strategies on energy are devoid of linkages with health interventions, such the effective use of natural ventilation and green roofing for buildings. Health interventions are essential to reducing indoor temperatures in warm months and hot climates to avert risks from heat stroke, indoor dampness and mold, thus providing a natural urban health asset to low-income households, who rarely afford air conditioners, fabricated with hydro fluorocarbons (HFCs) with global warming potential (Rao et al 2013). Planning should take into account the trajectories of mobility and energy use at city level and the multiple linkages with individual health at micro-scales. Public transportation as well as the positioning of buildings and choice of building materials have the potential to increase emissions of carbon dioxide (CO 2 ), periodic dust storms, excessive heat and dampness-all of which are risk factors for respiratory illnesses and other cardiopulmonary diseases, which lessen the capacity of people to handle or recover from other forms of ill health within the urban environment (Mukwaya 2012, Lwasa 2017. This means that policy and planning targets that are based on health-energymobility interactions offer opportunities for either trade-offs or co-benefits for scaling up efforts on resilience within and across different sectors.
All the climate strategies in the review are silent on how to harness the opportunities associated with the agile nature and varying degrees of informality in African cities. Evidence from academic literature shows that institutional planning for urban transformations to resilient cities in Africa, calls for the active participation and knowledge from informal city dwellers and businesses, such as street vendors and neigbourhood associations, who are for example involved in the promotion of waste economies, energy-efficient cooking stoves, green roofing for cooler indoor climates, and federations informal transport operators and city traders (Lindell et al 2019). It is possible to draw on and leverage the biographies and urban life course perspectives of informal land and business owners in African cities, to understand and design ways in which social networks and relations facilitate junctures and intermediation between sustainability-oriented informal urban structures and city planning for resilience to climate change (Buyana et al 2019, Van Breda and Swilling 2019, Monteith and Camfield 2019. For instance, efforts around the mobilization of spatial media technologies to digitally map informal settlements, has potential for use in monitoring and

Discussion
As noted by Nagendra et al (2018), the ways in which resilience plays out in cities that are less-built with agile informal settlements and multiple ecologies, is different. However, only two studies in the review contained an explicit definition of what resilience means. Chirisa et al (2016) defines resilience as, 'the ability of a social or ecological system to absorb disturbances while retaining the same basic structure and ways of functioning, the capacity for self-organization, and the capacity to adapt to stress and change. ' Ziervogel et al (2017) on the other hand, focuses on rights and justice as the point of departure for resilience to support risk management and inclusive development, as opposed to infrastructures, technical engineering and ecosystem services. However, both studies note that much of the resilience agenda has been shaped by policies and discourses from the global north, and therefore its practicability for cities of the global south, particularly African cities, has not been sufficiently addressed. There is need for theoretical perspectives from Africa that shine light on urban transformations to resilient cities, guided by emerging processes across case studies. In order to bring some structure to the debate on resilience in African cities, we postulate three interlinked streams: (i) social resilience as the lens for understanding interconnected networks of local community actors poised for sustainability-oriented urban experiments and learning; (ii) ecological resilience that means harnessing the positive interactions amongst interventions at neigbourhood to city and inter-regional scales; and (iii) institutional resilience, whose major constitute element is inter-municipal and intra-metropolitan collaborations with flexibility and capacity for coordinated response to climate risks and impacts on systems that make city regions functional (figure 3).
Recommendations for future research Inter-regional and ecological comparisons require expansion in scope and coverage Understanding pathways to resilience at different regional and ecological urban scales in Africa, is key to finding novel and context-sensitive ways for implementing climate action. However, evidence on interregional differences was mainly amongst East, West and South African cities, and much less amongst cities in Northern and Central Africa. Though there were two studies drawing on cases across Africa  (2011) and Di Ruocco et al (2015). This means that North and Central African cities were the least represented in the review, while cities in East, West and South Africa were the most represented.
From an ecological point of view, most of the comparisons majorly drew on coastal and in-land cities, with minimal representation of scalable interventions across semi-arid and mountainous cities, like Maseru in Lesotho. The comparative studies on coastal and in-land cities that met the inclusion criteria were 6 out of the 25, whereas studies on semi-arid/arid cities were the least represented (3 out of 25). Many of the excluded studies (72 articles) focused mainly on analyzing the magnitude and severity of risks and not existing or emerging interventions. This means that there is need to expand the scope of inter-regional and ecological comparisons on pathways for resilience to climate change in African cities.
More research is needed on the role played by intermunicipal and inter-metropolitan collaborations in enhancing resilience to climate change in Africa The growing magnitude of risks and impacts in African cities, means that interdependence amongst actors and decisions in capital and secondary cities is critical for effective planning on the resilience of urban systems. To reckon with, is the recent cyclone that affected urban mobility and housing systems in Mozambique, named Tropical Cyclone Idai. The disaster was reported to be gathering strength over the Indian Ocean that hit the city of Beira in Sofala Province with intensive flooding, but with governance implications for not only Maputo, which is the capital city, but also road connections between Zimbabwe and coastal ports as well as hydro-power projects in neighboring cities and countries (Muller 2019).
Nonetheless, the cities that were most represented in the review are the leading administrative, financial and tourism capitals of Africa. These include: Cairo, Kampala, Nairobi, Mombasa, Dakar, Abidjan, Lagos, Banjul, Bissau, Freetown, Monrovia, Accra, Kumasi, Lome, Conakry, Durban, Cape Town, Port Elizabeth and Alexandra in South Africa. Only 1 out of the 25 studies had a particular focus on the role played by inter-municipal collaborations, that is: Leck and Simon (2018). Therefore there is need for more research that sheds light on the linkages between climate resilience and network planning at various levels of connectivity amongst urban systems in Africa, including: inter-city connectivity, intra-metropolitan connectivity and local-level connectivity.

Data availability statement
Any data that support the findings of this study are included within the article.