In this section, the WFE nexus history and transition to mainstream researchers, methodologies, simulation framework, and policymaking organs are first presented which is closely followed by a discussion of the cluster topics as they relate to different city sizes following the selected 32,736 case studies and 2,233 cities.
3.1 Water-Food-Energy nexus literature genesis
WFE nexus research history can be classified into two major phases. The first phase commenced in the 1980s based on a two-pronged approach involving only “Food” and “Energy” while the second phase commenced in 2011 with the addition of the “Water” prong as illustrated in Fig. 2. However, the first Food-Energy Nexus Program (FEN-P) was initiated by the United Nations University (UNU) in 1983 to demonstrate mutually-linked issues between the energy and food sectors (Sachs and Silk, 1990). Furthermore, a year later, in 1984, Brazil hosted the first conference on “Food, Energy, and Ecosystems”, thus, introducing an environmental paradigm to the nexus program. In a similar vein, 1986 saw New Delhi (India) host the second international symposium on “Food-Energy Nexus and Ecosystems”. While in the late 1980s and early 1990s, the term “nexus” was spread and used by the World Bank to connect water, food, and trade (McCalla, 1997). Finally, the “Water-Food-Energy Nexus” thinking was officially adopted at the Bonn nexus conference in 2011 highlighting a new concept of “green economy”. All these events were accumulated and mutually reinforced to create a holistic framework that necessitates radical socio-ecological change.
In the last five years (2016 to 2021), the WFE research field rose by about 252% surging from 7,543 publications in 2016 to 26,567 publications in 2021 with 4,112 research case studies put forward in just 2021 alone. A close assessment of these publications indicates that original research article publications in the last five years grew by an estimated 201%. Alternatively, when review article publications are considered, the results illustrate an estimated growth rate of 476% (more than double the growth rate of research articles). In these reviews, scholars often use qualitative methods to examine WFE nexus’ state of the art such as in (Fan et al., 2019) where the authors have discussed the nexus using urban metabolism framework, or in (Gai et al., 2022) where the authors deciphered the methodological aspect of WFE nexus related to political agendas, boundary level, and nexus dimension. However, learning from case studies per se to shape a transferable urban knowledge regarding WFE nexus is largely lacking. Learning from WFE nexus case studies can be extremely helpful in displaying the nature of urban-policy making, and how better sharing of urban information regarding the WEF-nexus might advance the UN Sustainable Development Goals, and most importantly enabling cities with limited urban intelligence to inspire similar WFE nexus policies from peer-cities.
On the other hand, Fig. 3 illustrates the chronology of methods used from 1992 to 2021 within the global WFE nexus research space. The scoping shows that the top four methods employed are Life cycle analysis (LCA) with 3,665 applications, Input-output analysis (IOA) with 2,306 applications, Remote sensing with 901 applications, and Decision Support (DS) with 488 applications. These methods emphasize the measurement of environmental challenges (Albrecht et al., 2018) associated with all life stages of a given resource (water, energy, food) or product (service or commodity) counting in-boundary and out-boundary impacts. Carrying out these computations are necessary, during the period (i.e., from the 1990s to 2021), since the scientific community acknowledges the environmental component cutting across overlapping scales. Hence, overlapping scales now gave birth to transboundary pollution studies (i.e., pollution of a region in a given country could cause damage to other regions across the world). This transboundary effect has given rise to many green models such as the final consumer philosophy (Kitzes, 2013). Conversely, in the IOA framework, environmental impacts are allocated to the final consumers in the form of household consumption, gross fixed capital formation, and government consumption (Lenzen et al., 2007) offering a bottom-up approach to safeguard the earth’s life-support system as it captures several footprints including biodiversity, water, greenhouse gases (GHG), energy (Afionis et al., 2017).
Contrariwise, when the WFE nexus research space is assessed from 2011 since its emergence, three peaks are deciphered as portrayed in Fig. 3. The first peak occurred as a result of the “The Water-Energy and Food Security Nexus – Solutions for the Green Economy” conference held in Bonn (Germany) in November 2011 as previously stated and synthesized in a document titled “Understanding the Nexus” (Hoff, 2011). Following the Bonn conference, the Sustainable Development Goals (SDGs) in 2015 publicized the second peak of the WFE nexus approach space. The SDGs have a mission to provide “a blueprint to achieve a better and more sustainable future for people and the world by 2030”. Within the SGDs framework, 17 actionable goals were put forward for a post-2015 Development Agenda (Department of Economic and Social Affairs, 2022). Now, after the introduction of the SDGs, the third peak occurred at the Conference of the Parties (COP 21) held in Paris (France) in November 2016. COP 21 emphasized finding new innovative pathways to maximize synergies and reduce trade-offs on a city scale as a core element for reducing urban GHG emissions. Therefore, WFE nexus studies are also stimulated by global conference trends.
3.2 Global contrast: Bias towards the Northern Hemisphere
The analysis of the 32,736 documents as they relate to countries in the northern and southern hemispheres illustrated in Fig. 4, shows the current dominance of WFE nexus studies conducted in the Northern Hemisphere (92.3%) compared to 7.6% carried out in the Southern Hemisphere (Nagendra et al., 2018). Regionally, we observe a bias towards North America (representing about 37.8%) and Europe (approximately 35.9%). Hence, North America and Europe contributed about three-quarters of global WFE nexus research output while Africa and South America combined accounted for just 9% of total case studies to date. Again, although we observed an increase in WFE nexus case studies research from Asian cities since the Bonn 2011 conference, the growth of case studies and research production in African and Latin American cities remains stifled.
Hence, following considering demographic growth trends across the globe, WFE nexus research as it relates to global population distribution should be pivoted towards the Southern Hemisphere to cope with its growing population needs (Arrifano et al., 2021) and prepare them for climate impacts (Conway and Vincent, 2021) since the present global population demographics distribution shows Asia hosting about 59.51% of the world’s population, followed by Africa with 17.21%, then Europe (9.61%), Latin America and the Caribbean (8.38%), Northern America (7.74%), and Oceania (0.56%) (GeoNames, 2022). Furthermore, by 2030, Africa will be a hub of the fastest-growing cities while small Asian cities will host the largest share of the global urban population (Lamb et al., 2019). As such, the current lopsided focus on North America and Europe leaves African, Latin American, and Asian cities under-represented which may be due to data unavailability or the cost of research implementation (Kaviti Musango et al., 2020). Besides the challenge of growing enough food to meet the rising population of Africa and Asia, these regions also face the challenge of coping with their population becoming wealthier, and the host cities experiencing increasing infrastructure development and urbanization (Creutzig, 2015).
Hence, from further assessment of the documents scoped, the global WFE nexus research shows a tendency to small cities case studies that contribute about 68.6% share across all regions (see Fig. 5). This focus is justifiable, since small cities contribute 92.41% of the city size worldwide, while medium cities, megacities, and large cities each account for 4.79%, 2.63%, and 0.15% respectively. However, the total population in each of these categorized cities tells a different story. For instance, large cities host 46.12% of the total global population, followed closely by small cities with 22.25%, then megacities and medium cities with 15.94% and 15.66% respectively, as such, WFE nexus research should be shifted also towards large and medium cities to include population dynamics in the global WFE nexus research.
When geoparsing is employed, we identified the most intensive WFE nexus research-focused cities (Fig. 6). The results indicate that for large cities’ case studies, the top four cities are Berlin (Germany) with 818 WFE nexus studies, Singapore city (Singapore) with 510 direct WFE nexus research outputs, Washington (USA) with 175 case studies, and Budapest (Hungary) with 153 research items. On the other hand, the results for Megacities presented the City of New York (USA) as the dominant with a total of 305 case studies, followed by London (United Kingdom) with 193 research items, Paris (France) with 155 case studies, and Beijing (China) with 148 case studies. Still, the results negated the fact that WFE research output should increase with population increments.
3.3 Helm for global sustainability: Mapping WFE nexus knowledge production
WFE nexus is also an umbrella term comprising diverse study domains regarding current global-local concerns (socioeconomic and environmental sustainability), sectors decomposition (transport, waste, household, agriculture), policy and governance level (citizens implication, the role of mayors, environmental taxes), optimization process (energy grid, soil additives, water pipelines) and domains (engineering, urbanism, ecological habitat, diverse ecosystems). Hence, scoping and filtering global WFE nexus research by type of cities and geographical location is relevant to understanding the challenges researchers and policymakers face within a given context with determinant parameters, and how the poorest cities with less urban/rural intelligence (managing capacity) follow a similar context, may utilize the same policy recommendations of another city (also referred to as generalization, upscaling or blueprinting).
Moreover, carrying out WFE nexus research presents two challenges that are firstly wide-ranging and additionally involve identifying the characteristic context of a region(s). Wide-ranging challenges imply all cities face the same generic WFE nexus issues such as water scarcity, increased energy consumption, affluence, urban population, and the trend for infrastructure and urbanization, which place(s) the global WFE networks under severe environmental strain. On the other hand, context characteristics challenges involve such issues as mirroring climate change matters and local climate mitigation strategies that lean on city typology, local culture and beliefs, energy or water price, infrastructure development, environmental education, and level of awareness. Here lies a window of opportunity, whereby fully scoping WFE nexus studies and clustering them according to topic contents, type of cities, and geographical location could present a stepping stone towards reproducing a similar policy in underrepresented location(s) (such as Latin American and African cities).
Therefore, Fig. 7 illustrates the compiled topic clusters of WFE nexus case studies concerning the city size. The topic “Population growth & biomass variation under seasonal variation” (Cluster 2) is the most treated, contributing about 21.35% of the total case studies and buttressing the importance of sustainably meeting population growth demand needs. However, Cluster 4 topic content “Soil properties treatment & cropland management” (i.e., circa. 4.9%) is the least treated within the timeframe. A reason for the low uptake of the cluster topic may be a direct result of the length of time taken to observe the effects of soil treatment. This implies that designing a fully integrated WFE nexus model is geared toward tackling the grand global challenges of climate change (Dang et al., 2022) and land-use sustainable patterns (Laspidou et al., 2018) while promoting healthy consumption choices (Nakamura, 2022).
Moreover, the uneven distribution of topics according to the city size further supports the notion that all cities, regardless of size, face almost the same WFE nexus challenges (food waste, water recycling, energy security, and water scarcity). Nonetheless, a more in-depth assessment indicates that the dominant cluster WFE nexus research in small cities, medium cities, and large cities is “Cluster 2: Population growth & biomass variation under seasonal variation” with 4,547, 927, and 592 case studies respectively while megacities place more emphasis on “Cluster 1: Climate change modeling & sustainable cities policies development” with 319 case studies. However, the second-rated cluster consideration in small cities, medium cities, large cities, and megacities are respectively “Cluster 1: Wastewater treatment & temperatures effects on water availability” with 4,348 case studies, “Cluster 0: Climate change modeling & sustainable cities policies development” with 668 case studies, “Cluster 8: Crops productivity & the impacts of droughts” with 451 case studies, and “Cluster 7: Electricity and biofuel production & GHG emissions from energy production” with 242 case studies. The size of cities, hence, affects the WFE nexus research focus which tends to be directed to climate change and sustainability-related studies as the city size increases. These findings support the claim that due to the high concentration of carbon footprints in affluent cities (Hachaichi and Baouni, 2021), where a small number of local governments acting together can have a disproportionately large impact on global emissions (Moran et al., 2018).
Again, a more in-depth assessment of the nexus reasoning approach for the selected 2,233 cities as illustrated in Fig. 8 shows that not all cities use the nexus paradigm built around the triad “water-food-energy”. Some studies focused on two sectors in the form of water-energy interactions (e.g., Amman, Longueuil, Meknes, Medan, Pensacola) or water-food interlinkages (e.g., Nha Trang, Galveston, Appleton, Athi River, Kharkiv), and food-energy tradeoffs (e.g., Prague, San Rafael, Bratislava). On average, the global WFE nexus case studies leaned more toward “water” (51.3%) assessment than “energy” (25.8%) and “food” (22.7%). On the other hand, there is a bias in the Northern Hemisphere cities to “water” (51.8%) and “energy” (26.2%) while the Southern Hemisphere cities are converging more on “water” (48.6%) and “food” (27.6%). Not surprisingly, because the Southern Hemisphere cities do not meet the food self-sufficiency requirement (Zizipho, 2022), they tend to emphasize food production and water supply (Rakotovao et al., 2022), especially with the increasing negative impacts of climate change on water availability and food security in the region.
As portrayed in Fig. 9, we noticed that the largest cities in Africa (i.e., Kinshasa, Cairo, Ibadan, Alexandria, Algiers, Sousse, Durban, Johannesburg, Cape Town, Tripoli, and Lagos) and Latin America (i.e., Brasilia, Manaus, Lima, Bogota, Sao Paulo, Mexico City, Maracaibo, Leon, Campinas, Havana, and Salvador) exhibit not only scarcity in WFE nexus case studies and knowledge production but also lack topic diversity as evident. Besides, future urbanization challenges related to population growth are being anticipated and examined a priori especially in Lagos (in Africa) and Sao Paulo and Lima (in Latin America), instead of Algiers (Africa) which tends to focus on “Health issues related to pollution and contamination studies”, “Climate change modeling & sustainable cities policies development”, “Wastewater treatment & temperature effects on water availability” and “Electricity and biofuel production & GHG emissions from energy production”. Such topical bias is justifiable since Algeria is a primary energy exporter and ratified the Paris agreement (COP21) aiming to decrease its GHG emissions by 7% (Hachaichi et al., 2020). Consequently, these insights are difficult to spot using a conventional review process. Hence, the potential of machine learning coupled with big literature to reveal the most important large-scale scientific findings to date.
When WFE case studies are assessed on a continental scale (see Fig. 10), it is discerned that South America pays more attention to topics related to “Cluster 2: Population growth & Biomass variation under seasonal variation” with 560 case studies that support the conclusion of the small, medium, and large cities. Conversely, Europe-centric research emphasizes “Cluster 1: Wastewater treatment & temperatures effects on water availability” with 2,402 case studies. North America follows the same premise as South America concentrating on “Cluster 2: Population growth & biomass variation under seasonal variation” with an estimated 3,124 case studies. Africa, on the other hand simultaneously highlights “Cluster 2: Population growth & biomass variation under seasonal variation” with 371 case studies and “Cluster 8: Crops productivity & the impacts of droughts” with 279 case studies mainly because a significant part of the continent is susceptible to drought. Alternatively, Asia places more emphasis on “Cluster 1: Wastewater treatment & Temperatures effects on water availability” with 1,018 case studies and “Cluster 0: Climate change modeling & sustainable cities policies development” with 559 case studies. Finally, Oceania also pivots research towards “Cluster 1: Wastewater treatment & temperatures effects on water availability” with 47 case studies and “Cluster 6: Food production chains technology & Healthy consumption choices” with 28 case studies because of rising sea levels and coastal erosion that directly impact water and food production.
3.4 Mapping WFE nexus peer cities: Enabling a transfer-learning approach between the Northern hemisphere and Southern hemisphere
It can be difficult to produce food while efficiently using water and energy because of the intricate interdependencies of the water-energy-food nexus system. Decisions made in one area frequently have an impact on the other two areas. But what if we can swap successful WFE nexus projects? It is noteworthy most strategies will not be replicable in every city due to context characteristic challenges discussed in Section 3.3. However, to better comprehend the exchangeability of WFE nexus solutions and the comparability of situations/cities, we identified peer-cities (Fig. 11). Peer cities are cities that can shape similar policies to better manage local resources. For instance, we triggered that for cluster 1 (see Fig. 7) encompassing cities in Africa and Asia. Here, the city of Sidi Slimane (Morocco), Taipei (Taiwan), or Ouargla (Algeria) can learn good practices from such cities as Vancouver (Canada), Melbourne (Australia), and Amsterdam (Netherlands) in policies related to wastewater treatment.
To build expertise and learn from the knowledge production in the Northern Hemisphere two procedures are critical for transfer to the Southern Hemisphere. The first procedure concerns the selection of a case study (feasibility) where scholars aim to show that if a phenomenon is not possible (or possible) in one situation, it is more likely to be impossible (or possible) in all circumstances. As such, an alternative would be to cluster similar case studies to represent a wider system and then use random cases to serve as potential options to support generalizability claims (Steinberg, 2015). On the other hand, the second procedure involves a case study location (geography). Again, while we cannot record the motivations behind case study selection (for a given city), their geographic and topic cluster bias present some practical limits. However, learning from case studies is currently being implemented in the IPCC AR6 (both in Working Group III on Mitigation and Working Group II on Impacts & Adaptation), and the Second Assessment report by the Urban Climate Change Research Network (UCCRN) whereby 170 case studies were generated for the report and submitted to the project’s online docking-station. In these situations, the Intergovernmental Panel on Climate Change (IPCC) and UCCRN do not synthesize case studies for mutual and generalizable learning, they show successful implementation cases in boxed sections (e.g., Box 4.5 Singapore, Stockholm, and London). Yet, more importantly, breakthroughs in knowledge synthesis are required to make the best use of the growing literature.