The research methodology consisted of several steps (Fig. 1). First, we identified peer-reviewed scientific articles satisfying the search criteria. Second, we accessed grey literature by searching websites of key institutions and NBS databases for NBS projects and initiatives to ensure that NBSs advanced by development agencies but not scientifically studied were not missed. The peer-reviewed scientific articles were accessed through Scopus, ScienceDirect and Web of Science, and Google Scholar. Grey literature was searched for on the websites of 12 key institutions, including United Nations agencies and Local Governments for Sustainability (ICLEI), and 11 NBS databases (Table S1 in the Supplement). Eligibility was checked according to inclusion and exclusion criteria, and a thematic analysis was carried out following this paper selection process.

Search terms were selected after an initial scoping of other review papers on NBSs and related terms (Du Toit et al., 2018; Ruangpan et al., 2020; Thorn et al., 2021) and a review of NBSs and green infrastructure definitions, typologies and practices (Koc et al., 2017; Somarakis et al., 2019). Specific terms used during the search process were related to NBSs, green infrastructure, ecosystem services, urbanisation, hydro-meteorological risks and SSA (Table 2).

According to Donatti et al. (2020), NBS-related concepts like ecosystem-based adaptation can be advanced as on-the-ground actions or enabling activities. On-the-ground actions include ecosystem protection and restoration efforts, agricultural forest and conservation management practices, urban gardens, and green infrastructures. Enabling activities formulate policies, develop strategic plans and advance awareness-raising campaigns. In many cases, both approaches are married in the NBS roll-out. However, the literature search excluded papers only focused on enabling activities, since we aimed to document specific and tangible actions implemented to help address hydro-meteorological risks.

The grey-literature search was conducted on the websites of key institutions and the NBS databases from 23–30 April 2022. Peer-reviewed scientific papers were searched using Publish or Perish software, version 8.2, considering the time window from 1 January 2008 to 31 December 2021. These years were selected because 2008 was when the concept of NBSs emerged (Ruangpan et al., 2020). The literature search also allowed papers published in English and French, the top two official languages used by countries in SSA. In all, 3530 scientific peer-reviewed papers and 759 papers of grey literature were found.

Screening was performed by examining the titles and abstracts and, subsequently, the full text of the papers. The screening and selection process followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines, according to Page et al. (2021). Eligible papers had to meet the criteria defined in Fig. 1. Generally, papers included in the review had to provide data on NBSs that address specific hydro-meteorological risks; and they had to have an SSA city or peri-urban area – as several SSA countries lack a clear delineation of urban and rural areas – as the study area (Du Toit et al., 2018).

Apart from project documents, technical reports, fact sheets and policy briefs, non-peer-reviewed literature such as blog posts, news, magazine articles, commentaries and editorials was excluded to ensure that only papers following scientific standards were used for the review. Two people did the screening: one of the authors and a research assistant. Forty-five papers were deemed eligible for the study. Of them, 18 were peer-reviewed papers, while 27 were publications of grey literature. Only 1 paper, a publication of grey literature, was published in French. The remaining papers were published in English.

The quality and strength of evidence are essential to the systematic review process (Movsisyan et al., 2018). In this study, we used a 14-point framework to assess the quality of included papers (Table S2). We asked a series of questions on three themes – quality of reporting (six questions), risk of bias minimisation (five questions) and appropriateness of conclusions (three questions) – to ensure that quality research was done (Venkataramanan et al., 2018). For each paper, a score of 0, 0.5 or 1 was given for each of the 14 questions, and the scores were then converted to percentages to compare across themes (Fig. S1). The studies were rated from the perspective of social-ecological research methods as being of high quality (score of ≥10 to 14), medium quality (score ≥5 and <10) or low quality (score <5).

The data from the selected papers were extracted into Notion version 2.0.21, a project management software developed by Notion Labs Incorporated, for assessment. The coded information included:

study title,

A narrative summary of the papers is then given with the aid of tables, graphs and figures. ArcGIS Pro (version 2.8) by Esri (2022) was used to create maps to visualise the location of NBSs.

By conducting this study using a systematic review methodology, we could establish general trends in the literature on NBSs and hydro-meteorological risk mitigation in urban areas of SSA. However, factors such as the finite selection of keywords and poorly written abstracts could have led to the exclusion of important papers from the review. The impacts of implemented NBSs were not assessed to determine whether they were successful or if any lessons could be drawn due to the lack of the requisite data. In addition, the search was limited only to floods, extreme heat and drought, the most frequent hydro-meteorological risks in SSA. However, other risks like landslides and wildfires are recorded in the sub-region. Even though excluded languages like Portuguese and Kiswahili are not as widely spoken as English and French in SSA, the exclusion of papers published in these languages may also limit this study. Furthermore, because the focus was only on reported NBSs, some likely implemented or ongoing NBSs, which went unreported, were not captured in the analysis.

For floods, most NBSs were implemented in Dar es Salaam (n=4) and Kampala (n=3), both located in eastern Africa. Two NBSs were implemented in Nairobi and Gazi Bay, both in Kenya in eastern Africa; Accra in Ghana and Lagos in Nigeria in western Africa; and Durban and Johannesburg in South Africa and Nacala and Quelimane in Mozambique in southern Africa.

Regarding extreme-heat mitigation, most NBSs (n=6) were implemented in southern Africa. Three NBSs were implemented in eastern Africa, with most in Dar es Salaam (n=2). There was only one NBS in western Africa, in Lagos, Nigeria, and none were reported in central Africa.

For drought mitigation, the city of Johannesburg in South Africa was reported to have the most NBSs implemented (n=2). Only one NBS was implemented in each of the remaining cities. However, the majority of the NBSs were clustered in western Africa (n=9), followed by eastern Africa (n=8) and then southern Africa (n=3). Figure 5 presents the locations where the NBSs were implemented.

Green NBSs (n=20) were the most widely used for flood mitigation, followed by blue NBSs (n=17). Hybrid NBSs (n=7) were the least used. For extreme-heat mitigation, most NBSs were green (n=9), while a couple were found to be hybrid. There were no recorded blue NBSs. Seven green NBSs, three grey measures and one blue NBS were reported for drought mitigation. Figure 6 presents the link between NBS types and the hydro-meteorological risks addressed.

A total of 36 green, 18 blue and 12 hybrid NBS practices were reported for mitigating floods, extreme heat and drought in SSA. They summed up to 66 different NBS practices, with 44 deployed for addressing floods, 11 for addressing extreme heat and 11 for mitigating drought.

In terms of flood mitigation, the most reported NBS practices were mangrove restoration (n=10), wetland restoration (n=7), urban agriculture (n=5) and marine conservation (n=5). For extreme-heat mitigation, reforestation (n=10), urban forests (n=8), green conservation (n=7), gardens (n=6) and green/open spaces (n=6) were the most reported practices. For drought, the most common practices reported were agroforestry (n=3), conservation agriculture (n=2), integrated soil management (n=2) and sustainable agriculture (n=2). Table 3 presents a detailed list of NBS types and practices used for hydro-meteorological risk mitigation in SSA.

Ecosystem services are provisioning, regulatory or cultural. Intrinsically, NBSs used for mitigating hydro-meteorological risks provide regulatory ecosystem services, whether flood control, reversing the impact of extreme heat or addressing drought. However, we also explored if other ecosystem services were provided beyond the hazard mitigation services studied (Fig. 7).

Twenty-four different ecosystem services made up of 5 different provisioning services (20.8 %), 15 regulatory services (62.5 %) and 4 cultural services (16.7 %) were identified. In all, 88.9 % (40 papers) reported at least one type of ecosystem service, while 11.1 % (5 papers) reported none. Furthermore, 13.3 % (6 papers) reported on only one type of ecosystem service, 46.7 % (21 papers) reported on two types of ecosystem service and 28.9 % (13 papers) reported on all three types of ecosystem service.

After conducting this systematic review, we find that SSA is critically understudied in the area of NBSs for hydro-meteorological risk mitigation. Du Toit et al. (2018) found that only 38 % of cities in SSA had any research carried out on them on green infrastructure and ecosystem services. The review of Choi et al. (2021) on green infrastructure found that only 1 % of the papers included were from Africa. Nevertheless, there may be more NBS initiatives in SSA, although they are unreported or were not captured within the search terms used in this study. Such unreported NBSs most likely draw on local knowledge and are community-based, which makes documenting them challenging as a result of the ineffective data management culture in SSA (Malgwi et al., 2020; Manteaw et al., 2022). It is also likely that those locations in which NBSs are reported in the scientific literature are places where research funds have been made available for their investigation. What is more, there may be other activities that could qualify as NBSs but are not described as such. For example, African farmers have been using NBS-like practices such as agroforestry, stone bunds, grass strips and sustainable land use through techniques like observing fallow periods for generations without calling them NBSs (Keesstra et al., 2018). As such, it is unclear where a fine line should be drawn between age-old traditional practices and NBSs or whether they should be considered NBSs at all. Adopting the jointly created citizen science approach, which brings lay people and experts together for knowledge co-creation (Gill et al., 2021), could help incorporate such practices, which are effective, into NBSs and promote inclusivity and sustainability. The present study, therefore, affirms the assertions that the literature on NBSs and hydro-meteorological risk mitigation in SSA is scant, though this may be due in part to a lack of documentation and the use of different terminologies.

The results show that most papers were from South Africa, Kenya, Nigeria and Tanzania. This could be because these countries are among the biggest economies in SSA – South Africa and Nigeria, in particular, are the two biggest economies in SSA (Kamer, 2022) – and are basically leaders in their respective sub-regions. The four countries have also been forerunners in incorporating concepts like green infrastructure in urban planning, especially South Africa (e.g., Frantzeskaki et al., 2019; Russo et al., 2017; Venter et al., 2020). Furthermore, they boast some of the best educational and research institutions, which places them in an excellent position to advance research on urbanisation, climate change, and concepts like NBSs and ecosystem services.

Most reported NBSs were implemented on a national scale. This is likely because major climate funds like the Global Environment Facility and Green Climate Fund are more easily accessible to national governments than to non-profit and community-based organisations. Nonetheless, local-scale NBSs are the second most common kind. Such initiatives are often grassroots-driven, thus enabling local people to maximise benefits. However, many challenges often constrain local governance in SSA: decentralisation mechanisms may be ineffective, local-level capacity may be weak and financial resources may be limited (Hjerpe et al., 2014). For many SSA countries, development and climate adaptation often occur only when they are grassroots-driven by non-state actors or when local institutions are robust enough to lead or coordinate initiatives (Mubaya and Mafongoya, 2017). The Local Action for Biodiversity project advanced by ICLEI (which focused on improving the capacity of local governments and political actors, including mayors, on biodiversity and ecosystems) presents a good case study of how national, as well as even regional and international, projects can support local communities to develop more sustainably. International and regional NBSs also promote knowledge sharing, which is essential, especially in applying a novel concept like NBSs and in the context of the shared climate crisis that confronts all regions of the world.

Somalia, South Sudan and populations along the coast of Mozambique are identified as the most vulnerable to hydro-meteorological risks due to poor household and community resilience, high population densities, and weak governance systems (Busby et al., 2014), even though they are not located in the areas the IPCC predict will receive the harshest climate impacts in SSA. In this review, only Mozambique, among these most vulnerable countries, reported NBSs.

Based on the total deaths recorded from climate-related disasters, Somalia, Mozambique and Nigeria have been the most affected (CRED, 2019) (Table 4). However, only Nigeria, third on the list, is among the countries most studied in this review.

The factors behind very few papers from the countries most at risk could be attributed to political instability. Somalia, in particular, is third globally and first in SSA on the global Fragile States Index (Nasri et al., 2021). South Sudan, fourth globally and second in SSA on the Global Fragile States Index, is a relatively new country. Other reasons may be a lack of capacity for developing winning proposals for accessing climate funds and dwindling climate finance globally. The exclusion of papers published in Portuguese – because the language is not as widely spoken as English and French – could have also led to the low identification of papers in countries like Mozambique, Sao Tome and Principe, and Angola. Therefore, the reported NBSs for hydro-meteorological risk mitigation in SSA are in areas where risks exist but not where they are most severe.

In SSA, blue NBSs have been the most used when addressing floods, while green NBSs are more popular for extreme-heat and drought mitigation. However, in Europe, hybrid practices are the most popular when addressing floods, while green NBSs are more prevalent when responding to heatwaves and droughts. Blue NBSs are used the least (Sahani et al., 2019). NBS implementation often demands land (e.g., river restoration), which is often unavailable due to urbanisation (Pugliese et al., 2022). In Europe, 90 % of floodplains have been ecologically degraded (Entwistle et al., 2019), and the sections of urban areas vulnerable to floods increased by 1000 % between 1870 and 2016 (Paprotny et al., 2018). These factors have hampered the uptake of blue and green NBSs, which is why practitioners have had to settle for hybrid NBS practices. In SSA, the rapid rate of urbanisation often makes it challenging for city officials to keep up with urban environmental change, which is characterised by green depletion and environmental degradation (Cobbinah et al., 2019). Much of the Global North went through this period, especially between the 18th and 20th centuries, which saw the depletion of green spaces (Colding et al., 2020; Paprotny et al., 2018) and the degradation of several water-related ecosystems (Wantzen et al., 2019), which is why much attention has been on restoration even through NBS uptake (EC, 2016). In 2018, Europe was 4.2 % built-up (EUROSTAT, 2021) compared to 0.16 % in SSA (Karamage et al., 2018). A study on the extent of development in and around protected areas from 1975 to 2014 found that built-up areas were highest in Europe and Asia and lowest in Africa and Oceania (De La Fuente et al., 2020). Thus, the proliferation of blue and green NBSs in SSA implies that decision-makers can structure urbanisation using lessons from the Global North to avoid counterproductive practices and develop in a climate-resilient way. In particular, lessons can be drawn from NBSs like the Isar River Restoration in Germany (Pugliese et al., 2022) and the implementation of constructed wetlands, bio-swales, permeable pavements and other NBSs in the sponge city concept in China (Li and Zhang, 2022), both for flood mitigation, as well as ambitious greening efforts across Europe (Pauleit et al., 2019), Singapore and Hong Kong to improve thermal comfort (Aflaki et al., 2017).

Out of 66 NBS practices identified, most were implemented for flood mitigation. Earlier studies have found that 64 % of hazard events in Africa from 2000 to 2019 were flood-related (CRED, 2019). Many identified NBSs were reported to address multiple risks (Fig. 3). This demonstrates the multifunctionality of NBSs and highlights their relevance for SSA in addressing the variety of challenges in the sub-region within the context of limited climate adaptation funds. Comparatively, Sahani et al. (2019) found 205 NBSs used for addressing floods, heatwaves and drought in Europe. In a review in the German Alps, Zingraff-Hamed et al. (2021) also found 156 NBSs used to address floods and landslides. While NBSs are gradually becoming popular in SSA, it has not seen the level of wide uptake in the Global North, despite being the most vulnerable to hydro-meteorological risks.

Regarding flood risk mitigation, the most reported NBSs were mangrove restoration and wetland restoration. For extreme-heat mitigation, reforestation, urban forests and green conservation measures were the most reported NBSs. In Europe, NBSs like river and floodplain restoration (Zingraff-Hamed et al., 2021) and natural water retention measures (Hartmann et al., 2019) are more widely used for flood mitigation, while different green infrastructure types are used for heatwave mitigation (Pauleit et al., 2019). In this review, the most reported NBSs for drought mitigation were agroforestry, conservation agriculture, integrated soil management and sustainable agriculture. Consequently, there may be many similarities between NBS practices used in SSA and Europe. However, food production appears to be a critical necessity for many SSA locals, even in the uptake of NBSs for hydro-meteorological risk mitigation. Indeed, the agricultural sector is one of the most sorely affected by climate change in SSA (Stringer and Dougill, 2013), and it is predicted that yields could drop to up to 50 % by 2100 (FAO, 2009). This could explain why communities often lend more support to NBS projects that provide provisioning ecosystem services like fruits from tree crops (Etshekape et al., 2018).

NBS practices that are not common in SSA but are more widely used in the Global North were identified in SSA. These include green roofs, vertical greening, constructed wetlands and soil remediation. Green roofs are building rooftops where plants are grown in extensive or intensive ways. The review showed the increasing use of green roofs in many locations in Nigeria (Adegun et al., 2021). Vertical greening systems are plants grown along the vertical axis of buildings, either on the façade or in the interior. Studies in Nigeria found the practice improved thermal conditions and provided edible and medicinal plants (Akinwolemiwa et al., 2018; Oluwafeyikemi and Julie, 2015). Soil remediation is the process through which soils are returned to their original form of ecological stability before being disturbed. In Kenya, this method was used to help address floods through reduced runoff and improved access to co-benefits such as agricultural lands (Mulligan et al., 2020). These buttress the assertion that there may be many similarities between NBS practices used in Europe and those used in SSA.

SSA's most critical challenges include food and water insecurity, poverty, unemployment, and climate change (World Economic Forum, 2019). In SSA, 50 % of people live in urban areas (Kelsall et al., 2021), and over 43 % of this urban population live below the poverty line (Du Toit et al., 2018). Most of these people live in informal spheres and lack access to decent and affordable housing, food and water, and other necessities of life (Güneralp et al., 2017). Provisioning ecosystem services such as food, water and fuel are therefore necessary. This explains the popularity of NBSs, which are closely related to food provision – agriculture already employs most of the labour force – such as agroforestry and climate-smart agriculture. Also, the urban poor are the most vulnerable to climate change impacts, and the fact that NBSs can provide livelihood options is welcomed by locals. For decision-makers, the evidence that NBSs can promote climate action through carbon sequestration, mitigate heat and beautify cities, among other things, constitutes significant benefits and drivers of adoption (Lupp et al., 2021b; Thorn et al., 2021). Aside from delivering hazard mitigation services, NBSs could help address some of SSA's developmental challenges concurrently.

Cultural ecosystem services provide non-material benefits such as recreation, education and intellectual appreciation; physical and mental benefits; aesthetic significance; spiritual and symbolic appreciation; and enjoyment (Roux et al., 2020). Many of the papers did not report on cultural ecosystem services. This paper then adds to a long list of studies highlighting how cultural ecosystem services are little researched (e.g., Jones et al., 2022; Milcu et al., 2013). The lack of data in this sense makes it challenging to demonstrate the full spectrum of the benefits and disadvantages of NBSs. It reiterates calls by earlier authors to scientists to produce ecosystem service assessment frameworks, especially for cultural ecosystem services, to improve reporting (Christie et al., 2019; Schäffler and Swilling, 2013).

Most of the papers included in the review reported that NBSs created livelihood opportunities. Creating livelihood opportunities, mainly green jobs, which are more sustainable, is essential for a youthful region like SSA, where 60 % of the population is 25 years or younger (Mo Ibrahim Foundation, 2019). This is also relevant in addressing crime and insecurity, which is often rife among the 50 % and over people who reside in informal spheres in urban SSA due to a lack of economic opportunities. Improving life standards may also reduce the destruction of natural habitats and enhance natural restoration. Despite this, livelihood generation needs to be studied in detail, especially in river conservation and restoration projects because, in some instances, NBSs have led to the loss of local people's livelihoods. These have often occurred where risk responses have required the resettlement of populations such as with an NBS found to be used in SSA in this study (Douglas, 2018; Kita, 2017; Thorn et al., 2021). While its consideration as an NBS on its own may be contestable, Douglas (2018) indicates that relocation of informal settlements within riparian zones is a significant part of conservation and restoration initiatives in many locations in SSA, such as in Nairobi, Kenya. When such informal settlers were offered compensation and alternative livelihood options and relocated, they preferred to move back to these riparian areas, even if they were at risk of being impacted by floods, because their livelihoods were tied to these areas. When river corridors were also improved, it increased the value of such riparian lands, which became more attractive to developers and displaced the original informal settlers. This mirrors concerns with conventional engineering solutions like wastewater treatment plants, raises critical social justice concerns and could lead to a critique of the NBS concept.