Higher temperatures associated with climate change, through warmer global average temperatures and regional heatwave episodes, will interact with urban systems in a variety of ways (Doblas-Reyes et al., 2021 Box 10.3). Future urbanisation will amplify projected local air temperature increase, particularly by strong influence on minimum temperatures, which is approximately comparable in magnitude to global warming (high confidence) (Arias et al. In Press Box TS14). Within cities, exposure to heat island effects is uneven, with some populations disproportionately exposed to risk including low income communities, children, the elderly, disabled, and ethnic minorities (Quintana-Talvac et al., 2021; Sabrin et al., 2020; Chambers, 2020; and see later in this section).

The risks to cities, settlements and infrastructure from heatwaves will worsen (high confidence) (Leal Filho et al., 2021; see also Sections 6.2.5; 6.3.3.1, Arias et al. In Press Box TS14). Depending on the RCP, between half (RCP2.6) to three-quarters (RCP8.5) of the human population could be exposed to periods of life-threatening climatic conditions arising from coupled impacts of extreme heat and humidity by 2100 (Figure 6.3; Mora et al., 2017; Zhao et al., 2021). Cities in mid-latitudes are potentially subject to twice the levels of heat stress compared with their rural surroundings under all RCP scenarios by 2050, for example Belgian cities (Wouters et al., 2017). A disproportionate level of exposure exists in subtropical cities subject to year-round warm temperatures and higher humidity, requiring less warming to exceed ‘dangerous’ thresholds, for example Nairobi (Scott et al., 2017) and São Paulo (Diniz, Gonçalves and Sheridan, 2020). It is expected that more than 90% of the 300 million people who will be exposed to super- and ultra-extreme heatwaves in the Middle East and North Africa will live in urban centres (Zittis et al., 2021), while the major driver for increased heat exposure is the combination of global warming and population growth in already-warm cities in regions including Africa, India and the Middle East (Klein and Anderegg, 2021).

Locally, the urban heat island also elevates temperatures within cities relative to their surroundings. It is caused by physical changes to the surface energy balance of the pre-urban site from urbanisation, resulting from the thermal characteristics and spatial arrangement of the built environment, and anthropogenic heat release (Oke et al., 2017; Chow et al., 2014; Susca and Pomponi, 2020; Doblas-Reyes et al., 2021 FAQ10.1). A considerable body of evidence exists on how the multi-scale impacts and consequent risks arise when local elevated temperatures within settlements are enhanced by climate change, with specific elements of this affecting megacities (Darmanto et al., 2019). The urban heat island itself is amplified during heatwaves (Founda and Santamouris, 2017), but the extent to which varies regionally and by time of day (Ward et al., 2016a; Zhao et al., 2018b; Eunice Lo et al., 2020). When combined with warming induced by urban growth, extreme heat risks are expected to affect half of the future urban population, with a particular impact in the tropical Global South and in coastal cities and settlements (Huang et al., 2019; Section CCP2.2.2; Table CCP2.A.1).

Heat risk is associated with a range of health issues for urban residents, with the consequences of higher urban temperatures being unevenly distributed across urban populations (high confidence). Clear evidence exists of increased health risks to elderly populations in settlements, especially higher levels of mortality in elderly populations from urban heat islands during heatwave events (Fernandez Milan and Creutzig, 2015; Taylor et al., 2015; Ward et al., 2016a; Heaviside, Macintyre and Vardoulakis, 2017; Gough et al., 2019; Xu et al., 2020a), while health and fitness variables are also major determinants of the effects of heat stress (Schuster et al., 2017) (see also Table 7.2). Heat stress and dehydration are also related to behavioural and learning concerns, with dehydration impairing concentration and cognition for both adults and children (Merhej, 2019). Literature on paediatric heat exposure is associated with increases in emergency department visits for heat-related illnesses, electrolyte imbalances, fever, renal disease and respiratory disease in young children (Winquist et al., 2016), with less severe outcomes such as lethargy, headaches, rashes, cramps and exhaustion negatively affecting children in school and play environments (Vanos, 2015; Hyndman, 2017). Young children in cities are particularly sensitive to heatwaves, and may have little experience or capacity to cope with heat extremes (Norwegian Red Cross, 2019). Such vulnerability of young children to heat is compounded with projected urbanisation rates and poor infrastructure, particularly in South Asian and in African cities (Smith, 2019). There is evidence that socioeconomically disadvantaged populations are more likely to live in hotter parts of cities associated with higher-density residential land use in dwellings with less effective insulation built with poorer or older construction materials (Inostroza, Palme and de la Barrera, 2016; Tomlinson et al., 2011). Specific emerging risks for occupational and related heat illnesses are found in urban tropical or subtropical low- and middle-income countries (Andrews et al., 2018; Green et al., 2019).

There is an emerging risk of diminished indoor thermal comfort due to climate change, evidenced by research into negatively affected thermal comfort indices and/or increased number of overheating hours under future emissions scenarios (medium confidence) (e.g., Liu and Coley, 2015; van Hooff et al., 2014; Vardoulakis et al., 2015; Dodoo and Gustavsson, 2016; Invidiata and Ghisi, 2016; Makantasi and Mavrogianni, 2016; Mulville and Stravoravdis, 2016; Taylor et al., 2016; Hamdy et al., 2017; Pérez-Andreu et al., 2018; Salthammer et al., 2018; Dino and Meral Akgül, 2019; Osman and Sevinc, 2019; Roshan, Oji and Attia, 2019). Decreases in thermal comfort and increases in overheating risks depend on building characteristics, such as thermal resistance, presence of solar shading, thermal mass, ventilation, orientation and geographical location (e.g., Liu and Coley, 2015; van Hooff et al., 2014; Vardoulakis et al., 2015; Dodoo and Gustavsson, 2016; Invidiata and Ghisi, 2016; Makantasi and Mavrogianni, 2016; Mulville and Stravoravdis, 2016; Taylor et al., 2016; Hamdy et al., 2017; Pérez-Andreu et al., 2018; Salthammer et al., 2018; Dino and Meral Akgül, 2019; Osman and Sevinc, 2019; Roshan, Oji and Attia, 2019; Alves, Gonçalves and Duarte, 2021). Most of these studies employed numerical simulations in which different climate scenarios were used to construct future climate data. In hot climates, energy-efficient buildings with high insulation values and high airtightness, which have insufficient protection from solar heat gains and/or limited ventilation capabilities, are generally more vulnerable to overheating than older buildings with lower insulation levels (e.g., van Hooff et al., 2014; Vardoulakis et al., 2015; Makantasi and Mavrogianni, 2016; Mulville and Stravoravdis, 2016; Salthammer et al., 2018; Fisk, 2015; Hamdy et al., 2017; Fosas et al., 2018; Ozarisoy and Elsharkawy, 2019; see also Fox-Kemper et al., 2021 9.7 for building heat mitigation/adaptation links).

Higher urban temperatures result in lower labour productivity levels and economic outputs (medium confidence) (Graff Zivin and Neidell, 2014; Yi and Chan, 2017; Houser et al., 2015; Stevens, 2017; see Section 8.2.1). Globally, urban heat stress is projected to reduce labour capacity by 20% in hot months by 2050 compared with a current 10% reduction (Dunne, Stouffer and John, 2013). Burke et al. (2015) demonstrate a nonlinear relationship between temperature and global economic productivity, with potential global losses of 23% by 2100 due to climate change alone. In specific cases, Zander et al. (2015) estimate heat-related reductions in urban labour productivity in Australia to cost USD 3.6–5.1 billion yr −1, based on self-reported performance reduction and absenteeism among 1726 workers in 2013–142 ; while the high-temperature subsidies given in China at outdoor air temperatures above 35°C are projected to increase to USD 35.7 billion yr −1 after 2030 (compared with USD 5.5 billion yr −1 for 1979–2005) (Zhao et al., 2016)3 .

Higher urban temperatures place unequal economic stresses on residents and households through higher utilities demand during warm periods, for example, electricity in regions where air conditioning is predicted to become more prevalent, and due to medical costs associated with care for heat illnesses and related health effects, missed work and other related impacts (medium confidence) (Jovanović et al., 2015; Liu et al., 2019; Schmeltz, Petkova and Gamble, 2016; Soebarto and Bennetts, 2014; Zander and Mathew, 2019; Zander et al., 2015). Such stresses are projected to increase in many regions associated with continuing global-scale climate change and urbanisation (e.g., Véliz et al., 2017; Ang, Wang and Ma, 2017; Bezerra et al., 2021), although some of these effects in cold-climate cities are offset by reduced stresses in winter associated with urban heat island or rising temperatures more generally (see Section 6.2.2.4).

Thermal inequity can also be seen as a distributive justice risk (Mitchell and Chakraborty, 2018). There are often disproportionate increases of risk for individuals of lower socioeconomic status, especially migrants, from exposure to urban heat. These arise from inadequate housing, less access to air-conditioning, and occupations, such as manual labour and waste picking, that exacerbate heat exposure (Chu and Michael, 2018; Santha et al., 2016). Research from South Africa has shown that housing occupied by poor communities regularly experience indoor temperature fluctuations that are between 4°C and 5°C warmer compared with outdoor temperatures (Naicker et al., 2017); while evidence from the USA indicates that historical housing policies, particularly the ‘redlining’ of neighbourhoods based on racially motivated perceptions, are associated with areas that are exposed to elevated land surface temperatures (Hoffman, Shandas and Pendleton, 2020).

Social surveys from temperate and tropical cities highlight the risk of reduced quality of life during heat events, including increased incidence of personal discomfort in indoor and outdoor settings, elevated anxiety, depression and other indicators of adverse psychological health, and reductions in physical activity, social interactions, work attendance, tourism and recreation (high confidence) (Chow et al., 2016; Elnabawi, Hamza and Dudek, 2016; Obradovich and Fowler, 2017; Wang et al., 2017; Wong et al., 2017; Lam, Loughnan and Tapper, 2018; Alves, Duarte and Gonçalves, 2016). Extreme heat may also have a cultural impact, for example affecting major sporting events, with negative impacts on the athletic performance (Brocherie, Girard and Millet, 2015; Casa et al., 2015) and the experience and health of spectators (Hosokawa, Grundstein and Casa, 2018; Kosaka et al., 2018; Matzarakis et al., 2018; Vanos et al., 2019).

Flood risks in settlements arise from hydrometeorological events interacting with the urban system which exposes settlements to river (fluvial) floods, flash floods, pluvial (precipitation-driven) floods, sewer floods, coastal floods and glacial lake outburst floods (IPCC, 2012). Sea level increase and increases in tropical cyclone storm surge and rainfall intensity will increase the probability of coastal city flooding (high confidence) (WGI Box TS14). Globally, the increase in frequencies and intensities of extreme precipitation from global warming will likely 4 expand the global land area affected by flood hazards (medium confidence) ((Alfieri et al., 2018; Alfieri et al., 2017; Hoegh-Guldberg et al., 2018); Section 4.2.4.2). Mishra et al. (2015) noted that out of 241 urban areas, only 17% of cities experienced statistically significant increases in frequencies of extreme precipitation events from 1973 to 2012. In the future, there is some evidence that changes in high-intensity short duration (sub-daily) rainfall in urban areas will increase (limited evidence, medium agreement ) (Kendon et al., 2014; Ban, Schmidli and Schär, 2015; Abiodun et al., 2017).

Flooding associated with sea level rise is addressed in more detail in Cross-Chapter Paper 2, with detailed regional examples from Africa discussed in Section 9.3. Coastal flooding associated with sea level rise is exacerbated due to the significant number of people living in subsiding areas. As a result of this, the average coastal resident is experiencing (over the last two decades) rates of relative sea level rise three to four times higher than typical estimates due to climate-induced changes (Nicholls et al., 2021). This process can also result in release of coastal waste into urban areas (Beaven et al., 2020).

Urban flooding risks are also increased by urban expansion and land use and land cover change which enlarges impermeable surface areas through soil sealing, impacting drainage of floodwaters with consequent sewer overflows (high confidence) (Arnbjerg-Nielsen et al., 2013; Ziervogel et al., 2016; Aroua, 2016; Kundzewicz et al., 2014). These risks are also driven by increasing societal complexity, urban developmental policy on flood control and long-term economic growth (Berndtsson et al., 2019), including in mega-cities (Januriyadi et al., 2018). The increase in flood risk from urban development can be considerable; based on modelling of two RCP (4.5 and 8.5) scenarios, Kaspersen et al. (2017) noted flooding in four European cities could increase by up to 10% for every 1% increase in impervious surface area. Risks are also compounded by the location of settlements, with greater risks within cities located in low elevation coastal zones subject to sea level rise, potential land subsidence and exposure to tropical cyclones ((Koop and van Leeuwen, 2017; Hoegh-Guldberg et al., 2018; see also Section CCP2.2) and within informal settlements, where generally little investment in drainage solutions exists and flooding regularly disrupts livelihoods and disproportionately undermines local food safety and security for the urban poor (Dodman, Colenbrander and Archer, 2017; Dodman et al., 2017; Kundzewicz et al., 2014; Sections 5.4 and 5.8).

Future risks of urban flooding is increasing in conjunction with continued increases in global surface temperature (high confidence) (IPCC, 2019b; Winsemius et al., 2015; Kulp and Strauss, 2019; Hoegh-Guldberg et al., 2018). In particular, Asian cities are highly exposed to future flood risks arising from urbanisation processes. Between 2000 and 2030, rapid urbanisation in Indonesia will elevate flood risks by 76–120% for river and coastal floods, while sea level rise will further increase the exposure by 19–37% (Muis et al., 2015). In Can Tho, Vietnam, current urban development patterns put new assets and infrastructure at risk due to sea level rise and river flooding in the Mekong Delta (Chinh et al., 2017; Chinh et al., 2016). Flooding in urban areas is exacerbated both by the encroachment of urban areas into areas that retain water and by the lack of infrastructure such as embankments and flood walls, as is the case for large areas of Dhaka East (Haque, Bithell and Richards, 2020). Zhou et al. (2019) have also shown that for the city of Hohhot, China, the increase in impervious surfaces contributes between 2–4 times more to modelled annual flood risk compared with risk induced by climate change.

Global trends in surface water flooding are increasing, which poses risks to vulnerable urban systems depending on current adaptation measures to manage flooding impacts, for example, stormwater management, green infrastructure and sustainable urban drainage systems (Molenaar et al., 2015). The economic risks associated with future surface water flooding in towns and cities are considerable. For example in the UK, expected annual damages from surface water flooding may increase by £60–200 million for projected 2–4°C warming scenarios; enhanced adaptation actions could manage flooding up to a 2°C scenario but will be insufficient beyond that (Sayers et al., 2015). Analyses conducted in South Korea suggests that future flood levels could exceed current flood protection design standards by as much as 70% by 2100, considerably increasing urban flood risk (Kang et al., 2016). Modelling of urban flood damage in the Kelani River Basin in Sri Lanka showed increased frequency of flooding by 2030 could increase potential urban property damage by up to 10.2% (Komolafe Akinola, Herath and Avtar, 2018). Urban flood impacts may also exacerbate health burdens (including disease outbreaks of malaria, typhoid and cholera), which are compounded by damage to medical facilities (e.g., damage to hospitals and disruption of medicinal supply chains), as observed in urban areas of Ghana (Gough et al., 2019). In addition, emerging research shows the cascading consequences of hazard events, in this case urban flooding, on other risks to well-being in ways that are particularly severe for the urban poor, including mental ill-health, incidents of domestic violence impacting children and women, chronic diseases and salinity of drinking water ((Matsuyama, Khan and Khalequzzaman, 2020); Section 4.2.4.5; Section 6.2.4.2; Box 7.2; Section 8.4.5.2).

Urban water scarcity occurs when gaps exist between supply and demand of available freshwater resources (Zhang et al., 2019). Urban water security requires a sustainable quantity and quality of water to meet community and ecosystem needs in a changing climate (Romero-Lankao and Gnatz, 2019; Allan, Kenway and Head, 2018; Huang, Xu and Yin, 2015; Chen and Shi, 2016). Risks arising from urban water scarcity worldwide are very likely increasing due to climate drivers (e.g., warmer temperatures and droughts) and urbanisation processes (e.g., land use changes, migration to cities and changing patterns of water use including over extraction of surface and groundwater resources) affecting supply and demand (high confidence) (Allan, Kenway and Head, 2018; Crausbay et al., 2020; Haddeland et al., 2014; Pickard et al., 2017; De Stefano et al., 2015; Sun et al., 2019; Van Loon et al., 2016; Zhang et al., 2019; Section 4.2.4.4; See Box 8.6 for case study on 2018 Cape Town drought). Flörke et al. (2018) estimates that nearly a third of all major cities worldwide may exhaust their current water resources by 2050. Globally, projections suggest that 350 million (± 158.8 million) more people living in urban areas will be exposed to water scarcity from severe droughts at 1.5°C warming and 410.7 million (± 213.5 million) at 2°C warming (Liu et al., 2018).

Decreased regional precipitation and associated changes in runoff and storage from droughts is exacerbating urban scarcity by impairing the quality of water available for its resource management in cities (high confidence). For example, less runoff to freshwater rivers can increase salinity and concentrate pathogens and pollutants that increases risks of urban water scarcity (Hellwig, Stahl and Lange, 2017; Jones and van Vliet, 2018; Leddin and Macrae, 2020; Lorenzo and Kinzig, 2020; Ma et al., 2020; Mosley, 2015; Zhang et al., 2019; van Vliet, Flörke and Wada, 2017; see also Box 6.2). Drought also changes the dynamics of groundwater pollution, leading to increased environmental health risks when those sources are used for urban water supplies (Kubicz et al., 2021; Moreira et al., 2020; Pincetl et al., 2019). Changes in the nature of droughts, for example, hotter droughts (Herrera and Ault, 2017), snow droughts (Cooper, Nolin and Safeeq, 2016; Mote et al., 2016) or ‘flash’ droughts (Otkin et al., 2016; Otkin et al., 2018; Pendergrass et al., 2020) can exacerbate urban water scarcity, exposing the limitations of engineered water infrastructure designed to accommodate historical patterns of supply and demand (Gober et al., 2016; Ulibarri and Scott, 2019; Zhao et al., 2018a).

Risks of urban water scarcity and security are compounded by vulnerabilities such as service availability and quality of infrastructure to supply water for increased urban demand from in-migration to cities (medium confidence) (Ahmadalipour et al., 2019; Dong et al., 2020; Reynolds et al., 2019; Thomas et al., 2017; Mullin, 2020). Risks to local water security in cities are also exacerbated by drivers such as dependence on imported water resources from distant locales that may be exposed to additional drought risks (high confidence) (Ahams et al., 2017; Li et al., 2019b; Marston et al., 2015; Zhao et al., 2020; Zhang et al., 2020); from considerable projected urban expansion in drought-stressed areas, for example, across drylands of Western Asia and North Africa (Güneralp et al., (2015); and by export of virtual water (i.e., export of water embedded in food and energy) from local sources to distant trading partners (Djehdian et al., 2019; D’Odorico et al., 2018; Fulton and Cooley, 2015; Rushforth and Ruddell, 2016; Verdon-Kidd et al., 2017; Vora et al., 2017).

Droughts interact and manifest in complex ways in interconnected urban areas that likely increase risks of urban water scarcity (Tapia et al., 2017; Rushforth and Ruddell, 2015). Urban interdependencies mean droughts in one region can limit water resources availability in another (e.g., Macao and Zhuhai, Hong Kong, Shenzhen in China, Singapore and Johor, in cities in Pakistan and India, and in the west and southwest USA) (Chuah, Ho and Chow, 2018; Gober et al., 2016; Srinivasan, Konar and Sivapalan, 2017; Zhang et al., 2019; Zhao et al., 2020). Likewise, physical and social teleconnections mean decisions made about water resources in one region or location may impact another in unexpected ways (Moser and Hart, 2015; Liu et al., 2015).

Urban water security risks are confounded by inequities in economic opportunity, risk exposure and human well-being (medium evidence) (Sena et al., 2017; Stanke et al., 2013; Section 4.2.4.5). Water scarcity is felt more acutely among low-income compared with high-income populations (Nerkar et al., 2016), and scarcity on top of inequities and political instability can lead to security issues, for example, conflict between different water users (Cosic et al., 2019; von Uexkull et al., 2016; Ahmadalipour et al., 2019; Döring, 2020 ; Ide et al., 2021), particularly when road infrastructures and access to water are limited (Detges, 2016; Sena et al., 2017). Scarcity risks may also be exacerbated by human and ecosystem needs in water-short years (Srinivasan, Konar and Sivapalan, 2017). Finally, growing populations along with migration into water scarce regions can exacerbate water security issues (Akhtar and Shah, 2020; Singh and Sharma, 2019).

A range of other dynamic climate interactions are relevant for cities, settlements and infrastructure: cold spells, landslides, wind, fire and air pollution.

Cold spells. Although frequencies and intensities of cold spells/cold waves are virtually certain to have decreased globally, and are projected to consistently decrease for most warming levels (high confidence; WGI Table 11.2), cold weather events can periodically occur and impact urban areas and their connected infrastructures. For cities in eastern Canada, the intra-annual distribution of freezing rain events may become more frequent from December to February, and less frequent in other months by 2100 (Cheng, Li and Auld, 2011). Freezing rain is also a risk to urban populations and infrastructure. In general, higher population mortality rates likely occur during the winter season, while more temperature-attributable deaths are caused by cold than by heat in cities located in temperate climates (Gasparrini et al., 2015; Chen et al., 2017; Ryti, Guo and Jaakkola, 2016). Winter mortality is unlikely to significantly decrease due to warming trends, partly because a range of other medical factors (e.g., influenza seasons and elevations in cardiac risk factors) also drive this winter-excess mortality (Kinney et al., 2015). However, the evidence is unclear whether mortality related to cold waves will decrease in coming decades in European (Smid et al., 2019) or US cities (Wang et al., 2016). While projected global cold extremes are expected to decrease in frequency and intensity, the higher regional variability of future climates means that cold waves may remain locally important threats, including in milder regions where there are larger temperature differences between ‘normal’ winter days and extreme cold events, and where there is less capacity to adapt (Ma, Chen and Kan, 2014; Ho et al., 2019). This will be accentuated in many cities, particularly in Europe, by anticipated demographic changes that result in a more elderly population susceptible to cold wave health risks (Smid et al., 2019).

The effects of cold waves on the energy sector include breakdowns in power plants and reduced oil and gas production (Jendritzky, 1999), as well as failures in overhead power lines and towers leading to outages in Moscow and Bucharest (Panteli and Mancarella, 2015; Andrei et al., 2019). Six major power outages associated with cold shocks and ice storms have been recorded since 2010, the majority recorded from large cities in the USA (Añel et al., 2017). Cold waves can also significantly increase energy demand. A cold wave that affected the Iberian Peninsula in January 2017 caused electricity prices to peak at a mean price of 112.8 €/MWh, the highest ever recorded in Spain (AEMET, 2017).

Landslides. While geomorphological events (e.g., land subsidence from permafrost thaw at high latitudes or from groundwater extraction) and factors associated with the built environment (e.g., settlement location adjacent to steep slopes and zonation laws for building construction) are major factors determining urban landslide risk, these can also be influenced by a range of climatic variables, namely precipitation (frequency, intensity and duration), snow melt and temperature change. Some 48 million people are exposed to landslide risk in Europe alone, with the majority in smaller urban centres (Mateos et al., 2020). Travassos et al. (2020) also documented all landslide deaths in the São Paulo Macro Metropolis Region from 2016 to 2019 that occurred from extreme rainfall events in vulnerable areas prone to landslides. An increase in the number of people exposed to urban landslide risks is projected for landslide-prone settlements lying within regions projected to experience a corresponding increase in extreme rainfall (Gariano and Guzzetti, 2016). In addition, human factors such as expansion of towns onto unstable land and land use changes within settlements (e.g., road building, deforestation) are increasing human exposure to landslides and the likelihood of landslides occurring (Kirschbaum, Stanley and Zhou, 2015). Rainfall triggered landslides kill at least 5000 people per year, and at least 11.7% of these landslides occurred on road networks (Froude and Petley, 2018). Although the spatial footprint of an individual landslide might be small (i.e., < 1 km 2), the ‘vulnerability shadow’ cast over an area in terms of regional transport network disruptions can be a significant proportion of a region, and cascade to other infrastructures (Winter et al., 2016).

Landslides tend to occur on moderate to steep slopes, and are thus particularly prevalent in mountainous regions which are also characterised by low infrastructure redundancy (i.e., few alternative routes) and increased impacts from climate change (Schlögl et al., 2019). More robust forecasts of landslides driven by climate risk requires (a) more complete long-term records of previous landslides and (b) baseline studies of the Global South which are currently missing from the literature (Gariano et al., 2017).

Wind. Urban morphology alters wind conditions at multiple spatial scales; generally, increased surface roughness in settlements have resulted in declining trends in both measured wind speed and frequency of extremely windy days (Mishra et al., 2015; Peng et al., 2018; Ahmed and Bharat, 2014; WGI Box 10.3).

Urban wind risks can also be affected by location adjacent to mountains, lakes or coasts with localised wind systems (WGI 10.3.3.4.2; WGI 10.3.3.4.3). In large cities with significant urban heat island, an urban-driven thermal circulation can enhance pollution dispersion under calm conditions (Fan, Li and Yin, 2018) or advect heat to areas downwind of the city (Bassett et al., 2016). Microscale wind conditions within urban canyons also strongly affect ventilation of air pollution dispersion and thermal comfort at pedestrian level, especially in cities located in warm climates (Rajagopalan, Lim and Jamei, 2014; Middel et al., 2014; Lin and Ho, 2016).

In cities, wind risks from climate change hazards can arise from increased exposure from the expanding built environment. Very high wind speeds associated with severe weather systems, for example, tropical cyclones or derechos can cause significant structural damage to buildings and key infrastructure with insufficient wind load, as well as causing human injury through flying debris (Burgess et al., 2014). In particular, there is evidence from North American cities that tornado damage are likely fundamentally driven by growing built-environment exposure (medium confidence) (Ashley et al., 2014; Rosencrants and Ashley, 2015; Ashley and Strader, 2016).

Extreme winds in urban areas can have particularly damaging effects on poorly constructed buildings, including low-income houses in African cities (Okunola, 2019), and on urban trees that may be uprooted by strong wind gusts from downbursts (Ordóñez and Duinker, 2015; Pita and de Schwarzkopf, 2016; Brandt et al., 2016), as well as on disrupting transportation along urban road and railway networks (Koks et al., 2019; Pregnolato et al., 2016).

Fire. Hotter and drier climates in several regions, for example Australia, the Western USA, the Mediterranean and Russia (IPCC, 2018), likely enable weather conditions driving fire events impacting cities within these regions (Section 2.4.4.2, 2.5.5.2). These include wildfires along the margins where cities are adjacent to wildlands, that is, the wildland-urban interface (WUI) (Bento-Gonçalves and Vieira, 2020; Radeloff et al., 2018), or fires in cities with a high degree of informal settlements having greater vulnerability to fire hazards (Kahanji, Walls and Cicione, 2019; Walls and Zweig, 2017; Sections 8.3.3.2). This vulnerability is considerable; over 95% of urban fire related deaths and injuries occur within informal settlements in low- and middle-income countries (Rush et al., 2020).

For wildfires at the WUI, anthropogenic climate change, natural weather variability, expansion of human settlement and a legacy of fire suppression are key factors in determining fire risk (Abatzoglou and Williams, 2016; Knorr, Arneth and Jiang, 2016; van Oldenborgh et al., 2020). Recent wildfires in Australia and California both occurred under hot and dry weather conditions exacerbated by climate change, and resulted in substantial property damage along the WUI, ecosystem destruction and lives lost (Brown et al., 2020; Lewis et al., 2020; Yu et al., 2020). Future climate risk of fires at the WUI are likely (medium confidence), and are compounded by projected urban development along the WUI within several regions, such as in the Western USA (Syphard et al., 2019), Australia (Dowdy et al., 2019) and the Bolivian Chiquitania (Devisscher et al., 2016).

Air Pollution. Despite recent observed improvements in air quality arising from COVID-19 restrictions (Krecl et al., 2020; Naik et al., 2021 Cross-Chapter Box 6.1), significant risks to human health in cities leading to premature mortality very likely arise from exposure to decreased outdoor air quality from a combination of biogenic (e.g., wildfires at the WUI that advect into the urban atmosphere [Reddington et al., 2014; Naik et al., 2021 Chapter 12 Box 12.1]) and anthropogenic sources that are influenced by climate change (e.g., fine particulate matter such as PM2.5, tropospheric ozone, oxides of nitrogen and volatile organic compounds [Burnett et al., 2018; Knight et al., 2016; Turner et al., 2016; West et al., 2016; Chang et al., 2019b; Li et al., 2019a; Alexader, Luisa and Molina, 2016; Naik et al., 2021 Sections 6.5.1, 6.7.1.1, 6.7.1.2]). Risks of premature mortality from indoor air pollution in cities, arising from biomass burning for heating in winter or cooking, indoor pesticide use or exposure to volatile organic compounds from poor thermal insulation in buildings, are also likely to occur (Leung, 2015; Peduzzi et al., 2020; Cross-Chapter Box HEALTH in Chapter 7).

The mortality risk for several pollutants, for example PM2.5, is considerable (high confidence). Current estimates indicate that 95% of global population live in areas where ambient PM2.5 exceeds the WHO guideline of annual average exposure of 10 µg m −3 (Shaddick et al., 2018a; Shaddick et al., 2018b; Chang et al., 2019b). Among the 250 most populous urban areas, estimated PM2.5 concentrations are generally highest in cities in Africa, South Asia, the Middle East and East Asia; PM2.5 in many cities in North Africa and the Middle East is likely due mainly to wind-blown dust, whereas that in South Asia and East Asia are mainly anthropogenic in origin (Anenberg et al., 2019). However, data on PM2.5 concentrations are unavailable in many cities in low- and middle-income countries owing to a lack of measurements (Martin et al., 2019).

For some air pollutants, for example concentrations of PM2.5 in several US, Western European and Chinese cities have recently decreased as a result of clean air regulations that have controlled emissions from sources such as motor vehicles, fossil fuel power plants and major industries (Zheng et al., 2018a; Fleming et al., 2018). These decreases have brought substantial improvements in public health in settlements within these regions (Ciarelli et al., 2019; Zhang et al., 2018). In South Asia, Southeast Asia and Africa, however, concentrations of other air pollutants, for example tropospheric ozone, oxides of nitrogen and volatile organic compounds are likely to continue to grow and peak by mid-century before they subside due to global urbanisation assumptions embedded in the SSPs (Naik et al., 2021 Sections 6.2.1; 6.7.1.2). Broadly, future air pollutant emissions are projected to decline globally by 2050 as societies become wealthier and more willing to invest in air pollution controls, but the trajectories vary among pollutants, world regions and scenarios (Silva et al., 2016b; Rao et al., 2017; Silva et al., 2016c). Whereas cities in East Asia and South Asia currently have large exposure to anthropogenic air pollution, African cities may emerge by 2050 as the most polluted because of growing populations and demand for energy, increased urbanisation and relatively weak regulations to control emissions (Liousse et al., 2014).

Studies modelling climate change impacts on air quality find that the spatiotemporal patterns of concentration changes vary strongly at urban scales, and that often those patterns differ among the different years modelled due to internal variability (Saari et al., 2019) and different models used (Weaver et al., 2009). Changes in PM2.5 due to climate change are less clear than for ozone and may be relatively smaller (Westervelt et al., 2019) as climate change can affect PM2.5 species differently (Fiore, Naik and Leibensperger, 2015). For Beijing, climate change is expected to cause a 50% increase in the frequency of meteorological conditions conducive to high PM2.5 concentrations (Cai et al., 2017). The impacts of future climate change on air quality and consequent risks on human health have been studied at urban (Knowlton et al., 2004; Physick, Cope and Lee, 2014) and national scales (Fann et al., 2015; Orru et al., 2013; Doherty, Heal and O’Connor, 2017); globally, these studies have found a likely net increased risk of climate change on air pollution-related health (low confidence). They have focused mainly on the USA and Europe, with few studies elsewhere (Orru, Ebi and Forsberg, 2017), although the relationship between climate and air quality in megacities is particularly complex (Baklanov, Luisa and Molina, 2016). Silva et al. (2017) found that global premature mortality attributable to climate change (and not from urbanisation) from ozone and PM2.5 will increase by about 260,000 deaths per year in 2100 under RCP8.5, but substantial variance in results exists between individual models.

In both developed and less-developed regions, poverty in urban areas is frequently associated with higher levels of vulnerability (Huq et al., 2020b). This is evident in both rural and urban settlements in a wide range of contexts, including the Philippines (Porio et al., 2019; Valenzuela, Esteban and Onuki, 2020); Bangladesh (Matsuyama, Khan and Khalequzzaman, 2020); Brazil (Lemos et al., 2016), Santiago, Chile (Inostroza, Palme and de la Barrera, 2016); and New York City (Madrigano et al., 2015).

For individuals in urban communities, new literature highlights how differences in vulnerability established by social and economic processes are further differentiated by household and individual variability and intersectionality (Kaijser and Kronsell, 2014; Kuran et al., 2020).This includes differences in wealth and capacity (Romero-Lankao, Gnatz and Sperling, 2016); gender and non-binary gender (Michael and Vakulabharanam, 2016; Sauer and Stieß, 2021; Mersha and van Laerhoven, 2018); education, health, political power and social capital (Lemos et al., 2016); age, including young and elderly, low physical fitness, pre-existing disability, length of residence and social and ethnic marginalisation (Inostroza, Palme and de la Barrera, 2016; Schuster et al., 2017; Malakar and Mishra, 2017). An increasing proportion of refugees and displaced people now live in urban centres, and their characteristics also make them vulnerable to a range of shocks and stresses (Earle, 2016). While some individuals, including children, may be able to exercise agency to reduce their risk (Treichel, 2020), and some indicators are culturally specific, overall, poor, marginalised, socially isolated and informal urban households are particularly at risk (high confidence) (Brown and McGranahan, 2016; Kim et al., 2020b; Huq et al., 2020a; Huq et al., 2020b).

Particularly in low- and middle-income countries, much urban building occurs outside formal parameters and entails a high degree of urban informality. According to the United Nations statistics, the proportion of urban populations living in slums and informal settlements increased from 23% in 2014 to 23.5% in 2018 (United Nations, 2018). Informality is one pathway through which urbanisation generates differentiated vulnerability, tending to increase exposure and susceptibility of physical structures and their occupants to climate-related risks (Dodman et al., 2017; Dobson, 2017) in contexts including Guadalajara, Mexico (Gran Castro and Ramos De Robles, 2019), Kampala, Uganda (Richmond, Myers and Namuli, 2018), Bengaluru, India (Kumar, Geneletti and Nagendra, 2016), and Dar es Salaam, Tanzania (Yahia et al., 2018). In addition to facing emerging water- and heat-related risks, such areas are also more vulnerable to the health impacts of climate change (Scovronick, Lloyd and Kovats, 2015).

Even where formal planning is the norm, this has often remained oriented toward enabling value by adding construction or the protection of existing high-value physical assets, for example infrastructure and built cultural heritage, private residential) rather than enabling disaster risk reduction for all (Long and Rice, 2019). This tendency has been widely documented, including from cases in Australia, Thailand, Indonesia (King et al., 2016), Canada (Stevens and Senbel, 2017), Amman, Moscow and Delhi (Jabareen, 2015), and South Africa (Arfvidsson et al., 2017). Such inconsistencies between the delivery of land use planning and the aims of the SDGs combine with other social structures, economic pathways and governance systems to shape city risk profiles (Dodman et al., 2017).

Migration, displacement and resettlement each play a foundational role in differentiated vulnerability (see Cross-Chapter Box MIGRATE in Chapter 7). The relationship between migration and vulnerability is complex (robust evidence, high agreement ), and is the first of the three components discussed within this section. Climate change, as a push factor, is only one among multiple drivers (political, economic and social) related to environmental migration (Heslin et al., 2019; Plänitz, 2019; Luetz and Merson, 2019). There is consensus that it is difficult to pin climate change as the sole driver of internal (within national boundaries) rural to urban migration decisions owing to, among other factors, the disconnect between national and international policies (Wilkinson et al., 2016), the lack of unifying theoretical frameworks and the complex interactions between climatic and other drivers (social, demographic, economic and political) at multiple scales (Cattaneo et al., 2019; Borderon et al., 2019). Environmental migration, including rural to urban migration, triggered by climate change may ensue from either slow- or rapid-onset climatic events and could be either temporary, cyclical or permanent movement that occurs within or beyond national boundaries (Heslin et al., 2019; Silja, 2017).

A range of specific studies highlight certain elements of vulnerability and migration, including the ways in which slow-onset events affect precarious, resource-dependent livelihoods (such as farming and fisheries) (Cai et al., 2016). In small town Pakistan and Colombia, heat stress increases long-term migration of men, driven by a negative effect on farm income (Mueller, Gray and Kosec, 2014; Tovar-Restrepo and Irazábal, 2013). A study from Mexico reveals that an increase in drought months led to increased rural to urban migration, while increased heat (temperature) led to a ‘nonlinear’ pattern of rural to urban migration that occurred only after extended periods of heat (nearly 34 months) (Nawrotzki et al., 2017). This aligns with other findings that a consistent increase in temperature between 2°C and 4°C in some parts of the world renders involuntary, forced migration inevitable (Otto et al., 2017).

The complexity of migration drivers (as push or pull factors) explains why there is little agreement around quantitative estimates on migration (especially international) triggered by climate change (Silja, 2017; Otto et al., 2017), and why estimates of future displacement attributed to climate change and other environmental causes vary between 25 million and 1 billion in 2050 (Heslin et al., 2019). Many authors are critical of existing perspectives on climate-related migration, and argue for more nuanced research on the topic (Boas et al., 2019; Kaczan and Orgill-Meyer, 2020; Silja, 2017; Sakdapolrak et al., 2016; Singh and Basu, 2020; Luetz and Havea, 2018).

Climate-induced migration is not necessarily higher among poorer households whose mobility is more likely to be limited due to the poverty trap (i.e., lack of financial resources) (high confidence) (Cattaneo et al., 2019; Kaczan and Orgill-Meyer, 2020; Silja, 2017). For example, in Bangladesh, vulnerability of rural populations is increasing, so many of the poorest employ migration as a strategy of last resort (Paprocki, 2018; Penning-Rowsell, Sultana and Thompson, 2013; Adri and Simon, 2018) that occurs as soil salinity (as opposed to inundation alone) increases and is paralleled by economic diversification (i.e., aquaculture) (Chen and Mueller, 2018). There is robust evidence and high agreement that rapid-onset climatic events trigger involuntary migration and short-term, short-distance mobilities (Cattaneo et al., 2019). There is also robust evidence and high agreement that slow-onset climatic events (such as droughts and sea level rise) lead to long-distance internal displacement, more so than local or international migration (Kaczan and Orgill-Meyer, 2020; Silja, 2017), while sea level rise is expected to lead to the displacement of communities along coastal zones, such as in Florida in the USA (Hauer, 2017; Butler, Deyle and Mutnansky, 2016).

Migration, including rural–urban migration, is also recognised as an adaptation strategy in some circumstances, whether this is voluntary or planned (Jamero et al., 2019; Esteban et al., 2020a; Bettini, 2014). Voluntary migration can be an element of household strategies to diversify risk, depending on the nature of the climatic stress, and interacts with household composition, individual characteristics, social networks, and historical, political and economic contexts (Hunter, Luna and Norton, 2015; Carmin et al., 2015; Hayward et al., 2020). For example, in Colombia, rural to urban migration is differentiated across gender depending on the climatic stress whereby men migrate due to droughts, while women migrate due to excessive rain triggers (Tovar-Restrepo and Irazábal, 2013). Especially in Pacific small island developing states, migration can be a strategy for urban settlements or tribal communities to relocate in customary areas, as in the case of Vunidogoloa in Fiji (McMichael, Katonivualiku and Powell, 2019; Hayward et al., 2020); it can be a livelihood strategy as shown in the Cataret Islands in Papua New Guinea (Connell, 2016); or it can be used to enhance education and international networks (i.e., voluntary ‘migration with dignity’) as is the case in Kiribati (Heslin et al., 2019; Voigt-Graf and Kagan, 2017).

The second component, displacement, also plays a crucial role in differentiated vulnerability. The lack of resources and capacities to support mobility limits the effectiveness of migration as an adaptation strategy, therefore leading to both displacement and trapped populations in the future (Adger et al., 2015; Faist, 2018). For example, studies from Colombia (Tovar-Restrepo and Irazábal, 2013), India (Singh and Basu, 2020), Mekong Delta in Vietnam (Miller, 2019) and Pakistan (Islam and Khan, 2018) showed that migration as an adaptation strategy can be constrained due to resource barriers and low mobility potential, and also, to high levels of place attachment such as in the Peruvian Highlands (Adams, 2016), Vanuatu (Perumal, 2018) and the Tulun and Nissan Atolls of Bougainville, Papua New Guinea (Luetz and Havea, 2018). Migration can also be maladaptive for the receiving contexts, whether due to the pressure on and/or conflict over land and/or the urban resources (high confidence) (Faist, 2018; Singh and Basu, 2020; Luetz and Havea, 2018). Other views maintain that migration as adaptation overlooks the agency of people and their resilience, that is the nuances of ‘translocal social resilience’ (Kelman, 2018; Silja, 2017; Sakdapolrak et al., 2016). For example, the ni-Vanuatu prioritise in situ adaptation measures and leave migration as a last resort (Perumal, 2018).

Regardless of the reasons and the initiators for migration, community control over resettlement both at the origin and destination leads to more positive outcomes for both the communities being resettled and the receiving communities (high confidence) (Perumal, 2018; Ferris, 2015; Price, 2019; Mortreux and Adams, 2015; Tadgell, Doberstein and Mortsch, 2018; Luetz and Havea, 2018). The protection of livelihoods contributes to ensuring the well-being (physical and mental) and the protection of the rights of communities (high confidence) (Ferris, 2015; Price, 2019). There is limited evidence but high agreement that the outcomes of resettlement initiatives are complex and multi-faceted (Ferris, 2015). For example, in Shangnan County, northwest China, the Massive Southern Shaanxi Migration Program, based on voluntary participation, reduced risk exposure and improved the quality of life in general, but also disproportionately increased the vulnerability of disadvantaged groups (the poor, migrants, and those left behind) (Lei et al., 2017). Similarly, vulnerability increased due to the loss of connection to place and community bonds in Mekong Delta, Vietnam (Miller, 2019), and due to unsafe construction, poor infrastructure, institutional incapacity and general neglect in resettlement initiatives in Malawi, sub-Saharan Africa (Kita, 2017).

Energy infrastructure underpins modern economies and quality of life. Disruption to power or fuel supplies impacts upon all other infrastructure sectors, and affects businesses, industry, healthcare and other critical services both within and across jurisdictional boundaries (Groundstroem and Juhola, 2019). The economic impacts of climate change risks are significant, for example in the EU, the expected annual damages to energy infrastructure, currently €0.5 billion yr −1, are projected to increase 1612% by the 2080s (Forzieri et al., 2018). In China, 33.9% of the population are vulnerable to electricity supply disruptions from a flood or drought (Hu et al., 2016), whilst in the USA, higher temperatures are projected to increase power system costs by about USD 50 billion by the year 2050 (Jaglom et al., 2014). In a study of 11 Central and Eastern European countries, researchers found that energy poverty is exacerbated by existing infrastructure deficits and energy efficient building stock, as well as income inequality, which can lead to reduced economic productivity (Karpinska and Śmiech, 2020). Climate change is expected to alter energy demand (Viguié et al., 2021), for example heatwaves increase spot market prices (Pechan and Eisenack, 2014), with a disproportionate impact on the poorest and most vulnerable populations. Energy infrastructure are susceptible to a range of climate risks (Cronin, Anandarajah and Dessens, 2018), whilst issues pertaining to energy demand are considered by Working Group III.

Climate change can, for example, influence energy consumption patterns by changing how household and industrial consumers respond to short-term weather shocks, as well as how they adapt to long-term changes (Auffhammer and Mansur, 2014). Recent studies from Stockholm, Sweden, show that future heating demand will decrease while cooling demand will increase (Nik and Sasic Kalagasidis, 2013). A study from the USA showed that climate change will impact buildings by affecting peak and annual building energy consumption (Fri and Savitz, 2014). From an infrastructure standpoint, the vulnerability of current hydropower and thermoelectric power generation systems may change due to changes in climate and water systems and projected reduction of usable capacities (van Vliet et al., 2016; Byers et al., 2016). These examples show how energy infrastructure planning under climate change must take into account a greater number of scenarios and investigate impacts on particular energy segments (Sharifi and Yamagata, 2016).

Electricity generation. Electricity generation infrastructure can be directly damaged by floods, storm and other severe weather events. Furthermore, the performance of renewables (solar, hydro-electric, wind) is affected by changes in climate.

Most thermoelectric plants require water for cooling, many are therefore situated near rivers and coasts and thus vulnerable to flooding. Increases in water temperature or restrictions on cooling water availability affect hydroelectric and thermoelectric plants. A 1°C increase in the temperature of water used as coolant yields a decrease of 0.12–0.7% in power output (Mima and Criqui, 2015; Ibrahim, Ibrahim and Attia, 2014). Excess biological growth, accelerated by warmer water, increases risk of clogging water intakes (Cruz and Krausmann, 2013). While some regions are expected to experience increased capacity under climate change (namely India and Russia), global annual thermal power plant capacity is likely to be reduced by between 7% in a mid-century RCP2.6 scenario and 12% in a mid-century RCP8.5 scenario (van Vliet et al., 2016). Worldwide, hydroelectric capacity reductions are projected at 0.4–6.1% (van Vliet et al., 2016). Analysis of the UK’s water for energy generation abstractions showed that an energy mix of high nuclear or carbon capture technologies could require as much as six times the current cooling water demands (Byers, Hall and Amezaga, 2014; Byers et al., 2016).

Increasing temperatures improve the efficiency of solar heating but decrease the efficiency of photovoltaic panels, and deposition and abrasive effects of wind-blown sand and dust on solar energy plants can further reduce power output, and the need for cleaning (Patt, Pfenninger and Lilliestam, 2013). Projected changes in wind and solar potential are uncertain; the trends vary by region and season (Burnett, Barbour and Harrison, 2014; Cradden et al., 2015; Fant, Schlosser and Strzepek, 2016). In an RCP8.5 scenario, Wild et al. (2015) conservatively calculate a global reduction of 1% per decade between 2005 and 2049 for future solar power production changes due to changing solar resources as a result of global warming and decreasing all-sky radiation over the coming decades. However, positive trends are projected in large parts of Europe, the south-east of North America and the south-east of China.

Electricity Transmission and Distribution. Electricity transmission and distribution networks span large distances, with overhead power lines often traversing exposed areas. Power lines and other assets, such as substations, are often located near population centres, including those in floodplains. Structural damage to overhead distribution lines will increase in areas projected to see more ice or freezing rain (e.g., most of Canada), snowfall (e.g., Japan) or wildfires (e.g., California, USA) (Bompard et al., 2013; Mitchell, 2013; Sathaye et al., 2013; Jeong et al., 2018; Ohba and Sugimoto, 2020). Electricity outages may last for prolonged periods of time and across vast areas, in addition to potentially disproportionately affecting poorer or more vulnerable communities. Increases in windstorm frequency and intensity increase the risk of direct damage to overhead lines and pylons, in many locations this is limited but Tyusov et al. (2017) calculate an increase as high as 30% in parts of Russia. Where the mode of failure is recorded, transmission pylons are seen to be more susceptible to wind damage, whilst distribution pylons are more likely to be affected by treefall and debris (Karagiannis et al., 2019). Increased temperatures can lead to the de-rating (lower performance) of power lines, whose resistance increases with temperature with efficiency reductions of 2–14% being projected by 2100 (Cradden and Harrison, 2013; Bartos et al., 2016).

Fuels Extraction and Distribution. Non-electric energy infrastructure is susceptible to many of the same impacts as electric infrastructure. Extreme weather events impact extraction (onshore and offshore) and refining operations of petroleum, oil, coal, gas and biofuels. Disruption of road, rail and shipping routes (see Section 6.2.5.2) interrupts fuel supply chains. However, there are a number of risks that are specific to these sectors. Heat can lead to expansion in oil and gas pipes, increasing the risk of rupture (Sieber, 2013), whilst heatwaves and droughts can reduce the availability of biofuel (Moiseyev et al., 2011; Schaeffer et al., 2012). Subsidence and shrinkage of soils damages underground assets such as pipes intakes (Cruz and Krausmann, 2013), while additional human activity such as extractive drilling may induce earthquakes, as observed in the northern Dutch province of Groningen (Van der Voort and Vanclay, 2015). In Alaska, USA, the thaw of permafrost and subsequent ground instability is estimated to lead to USD 33 million damages to fuel pipelines in an end-of-century RCP8.5 scenario (Melvin et al., 2017), with low-lying coastal deltas particularly vulnerable (Schmidt, 2015).

Since AR5, research has highlighted the implications for disruption to global supply chains (Becker et al., 2018; Shughrue and Seto, 2018; Pató, 2015), and has made advancements in quantifying costs of climate risks to transportation infrastructure. Climate risks to transport infrastructure (from heat- and cold waves, droughts, wildfires, river and coastal floods, and windstorms) in Europe could rise from €0.5 billion to over €10 billion by the 2080s (Forzieri et al., 2018). Across the Arctic, nearly four million people and 70% of all current infrastructure, including resource extraction and transportation routes, will be at risk by 2050 (Hjort et al., 2018), although the design of specific infrastructure may also affect the degree of infrastructure damage, depending on local geological and ecological conditions. Globally, Koks et al. (2019) calculated that approximately 7.5% of road and railway assets are exposed to a 1-in-100 year flood events, and total global expected annual damages (EAD) of USD 3.1–22 billion (mean USD 14.6 billion) due to direct damage from cyclone winds, surface and river flooding, and coastal flooding. The majority of this is caused by surface water and fluvial flooding (mean USD 10.7 billion). Although twice as much infrastructure is exposed to cyclone winds compared with flooding, a mean EAD of USD 0.5 billion is significantly less than for coastal flooding (USD 2.3 billion), as cyclone damages are largely limited to bridge damage and the cost of removing trees fallen on road carriageways and railway tracks. This is small relative to global gross domestic product (GDP; ~0.02%). However, in some countries EAD equates to 0.5–1% of GDP, which is the same order of magnitude as typical national transport infrastructure budgets, but especially significant for countries such as Fiji that already spend 30% of their government budget on transport (World Bank Group, 2017). Koks et al. (2019) did not assess future climate change impacts, but comparable studies calculating changes in EAD from flooding based upon land use show increases of 170–1370%, depending on global greenhouse gas emissions levels (Alfieri et al., 2017; Winsemius et al., 2015). Moreover, Schweikert et al., (2014) report that climate risks to transport infrastructure could cost as much as 5% of annual road infrastructure budgets by 2100, with disproportionate impacts in some low and lower middle-income countries.

Changes in rainfall and temperature patterns are expected to increase geotechnical failures of embankments and earthworks (Briggs, Loveridge and Glendinning, 2017; Tang et al., 2018; Powrie and Smethurst, 2018) from landslides, subsidence, sinkholes, desiccation and freeze-thaw action. For instance, Pk et al. (2018) show this could lead to a 30% reduction in the engineering factor of safety of earth embankments in Southern Ontario (Canada). Increased river flows in many catchments will also increase failures from bridge scours (Forzieri et al., 2018). HR Wallingford (2014) calculate that the projected 8% increase in scouring from high river flows in the UK will lead to 1 in 20 bridges being at high risk of failure by the 2080s, whilst in the USA the 129,000 bridges currently deficient could increase by 100,000 (Wright et al., 2012). With respect to temperature, analysis by Forzieri et al. (2018) concludes that heatwaves will be the most significant risk to EU transport infrastructure in the 2080s, as a result of buckling of roads and railways due to thermal expansion, melting of road asphalt and softening of pavement material. In the USA, over 50% more roads will require rehabilitation (Mallick et al., 2018), whilst USD 596 million will be required through 2050 to maintain and repair roads in Malawi, Mozambique and Zambia (Chinowsky, Price and Neumann, 2013).

In addition to direct damages from flooding and heatwaves, disruption caused by road blockages will be increased by more frequent flood events. For example, in the city of Newcastle upon Tyne (UK), road travel disruption across the city from a 1-in-50 year surface water flood event could increase by 66% by the 2080s (Pregnolato et al., 2017), whilst heatwaves could treble railway speed restrictions in parts of the UK (Palin et al., 2013). Knott et al. (2017) highlighted risks to coastal infrastructure where ~30 cm sea level rise sea level rise would also push up groundwater and reduce design life by 5–17% in New Hampshire (USA). Heavy rain and flooding can also inundate underground transport systems (Forero-Ortiz, Martínez-Gomariz and Canas Porcuna, 2020).

Many airports, and by their nature ports, are in the low elevation coastal zone, making them especially vulnerable to flooding and sea level rise. Under a 2oC scenario, the number of airports at risk of storm surge flooding increases from 269 to 338 or as many as 572 in an RCP8.5 scenario; these airports are disproportionately busy and account for up to 20% of the world’s passenger routes (Yesudian and Dawson, 2021). Airport and port operations could be disrupted by icing of aircraft wings, vessels, decks, riggings and docks (Doll, Klug and Enei, 2014; Chhetri et al., 2015). Warming will increase microbiological corrosion of steel marine structures (Chaves et al., 2016). Fog, high winds and waves can disrupt port and airport activity, but changes are uncertain and with regional variation (Mosvold Larsen, 2015; Izaguirre et al., 2021; Becker, 2020; León-Mateos et al., 2021; Taszarek, Kendzierski and Pilguj, 2020; Danielson, Zhang and Perrie, 2020; Kawai et al., 2016).

Waterways are still important transport routes for goods in many parts of the world, although they are mostly expected to benefit from reduced closure from ice (Jonkeren et al., 2014; Schweighofer, 2014), low flows will likely lead to reduced navigability and increased closures; van Slobbe et al. (2016) estimate the Rhine may reach a turning point for waterway transportation between 2070–2095. Obstruction due to debris and fallen vegetation of roads and rails and to inland and marine shipping from high winds are expected to increase (Koks et al., 2019; Kawai et al., 2018; Karagiannis et al., 2019)..

Information and communication technology (ICT) comprises the integrated networks, systems and components enabling the transmission, receipt, capture, storage and manipulation of information by users on and across electronic devices (Fu, Horrocks and Winne, 2016). ICT infrastructure faces a number of climate risks. Increased frequency of coastal, fluvial or pluvial flooding will damage key ICT assets such as cables, masts, pylons, data centres, telephone exchanges, base stations or switching centres (Fu, Horrocks and Winne, 2016). This leads to loss of voice communications, inability to process financial transactions and interruption to control and clock synchronisation signals. Insufficient information about the location and nature of many ICT assets limits detailed quantitative assessment of climate change risks.

Fixed-line ICT networks that sprawl over large areas are especially susceptible to increases in the frequency or intensity of storms that would increase the risk of wind, ice and snow damage to overhead cables and damage from wind-blown debris. More intense or longer droughts and heatwaves can cause ground shrinkage and damage underground ICT infrastructure (Fu, Horrocks and Winne, 2016). In mountain and northern permafrost regions, communications and other infrastructure networks are subject to subsidence because of warming of ice-rich permafrost (Shiklomanov et al., 2017; Li et al., 2016; Melvin et al., 2017).

For the urban housing sector, climate impacts such as flooding, heat, fire and wind assessed in Section 6.2.3 will likely have detrimental effects on housing stock (including physical damage and loss of property value) and on residents exposed to climate risks (robust evidence, high agreement ).

In the USA, for example, 15.4 million housing units fall within a 1-in-100-year floodplain (Wing et al., 2018). Assessment of the Miami-Dade area in Florida noted that coastal inundation caused by tidal flooding (and to a lesser extent sea level rise) resulted in over USD 465 million in lost real-estate market value between 2005 and 2016 (McAlpine and Porter, 2018), although property values have increased from high-end housing construction and climate adaptation measures (Kim, 2020). Emergent risk reflecting novel research include aggravated moisture problems in buildings from wind driven rain (Nik et al., 2015). Future risks from future sea level rise are elaborated in Section CCP2.2.1. Housing infrastructure are also susceptible to extreme heat and wind events (Stewart et al., 2018). These risks are further elaborated on in Section 6.2.3, although it is important to note that heat risks, in particular, tend to be concentrated within communities with a higher proportion of social housing (Mavrogianni et al., 2015; Sameni et al., 2015) or low-cost government-built houses and informal settlements.

Apart from land subsidence from urbanisation (e.g., Case Study 6.2), substantial climate risks to urban sanitation arise from droughts, flooding and storm surges. Low flows from drought can lead to sedimentation, increase pollutant concentration and block sewer infrastructure networks (Campos and Darch, 2015). Flooding poses a greater risk for urban sanitation in low- and middle-income settings (Burgin et al., 2019) where onsite systems are more common. Floodwater may wash out pits and tanks, mobilising faecal sludges and other hazardous materials leading to both direct and indirect exposure via food and contaminated objects and surfaces, and pollute streams and waterbodies (Howard et al., 2016; Braks and de Roda Husman, 2013; Bornemann et al., 2019). Floods also damage infrastructure; toilets, pits, tanks and treatment systems are all vulnerable (Sherpa et al., 2014; UNICEF and WHO 2019).

Sanitation systems coupled with floodwater management are at risk of damage and capacity exceedance from high rainfall (Thakali, Kalra and Ahmad, 2016; Kirshen et al., 2015; Dong, Guo and Zeng, 2017). In England, the number of water and wastewater treatment plants at risk of flooding is projected to increase by 33% under a 4oC scenario (Sayers et al., 2015), but risks are generally increasing for both formal and informal urban sanitation systems (Howard et al., 2016).

Urban ecological infrastructure includes green (i.e., vegetated), blue (i.e., water-based) and grey (i.e., non-living) components of urban ecosystems (Li et al., 2017). While land cover change from urbanisation directly reduces the extent of natural and ecological infrastructure (e.g., Lin, Meyers and Barnett, 2015), notable risks arise from climate drivers. Recent research particularly highlights future climate impacts on coastal natural infrastructure, including beaches, wetlands and mangroves, which cause significant economic losses from property damage and decreasing tourism income, as well as loss of natural capital and ecosystem services. Research on climate risks to urban trees and forests is comparatively limited. Instead, urban vegetation and green infrastructure are most often cast as adaptation strategies to reduce urban heat, mitigate drought and provide other ecosystem benefits (see Section 6.3.2).

Coastal natural infrastructure is exposed to sea level rise, wave action and inundation from increasing storm events (See also Section CCP 2.2.1). Beaches, in particular, are highly exposed to climate-induced coastal erosion (Toimil et al., 2018; Section CCP2). Research from settlements across coastal Southern California, USA, show that 67% of all beaches may completely erode by 2100 (Vitousek, Barnard and Limber, 2017). Coastal zones across Cancún, Mexico, are exposed to a combination of sea level rise and tropical hurricanes, further exacerbated by urban development patterns blocking natural sediment replenishment to beaches (Escudero-Castillo et al., 2018). In another case, beach erosion along the heavily urbanised Valparaíso Bay, Chile, is heightened by El Niño Southern Oscillation (ENSO) events, which in the past have caused an additional 15–20 cm in mean sea level rise (Martínez et al., 2018).

Wetlands, mangroves and estuaries, which tend to be heavily urbanised areas, are highly at risk from sea level rise and changing precipitation (Green et al., 2017; Feller et al., 2017; Alongi, 2015; Osland et al., 2017; Chow, 2018; Godoy and Lacerda, 2015). Sea level rise is a concern for wetlands and mangroves across coastal urban Asia, the Mississippi Delta (US) and low lying small island states (Ward et al., 2016b). Research on the highly urbanised Yangtze River estuary in China shows that soil submersion and erosion from sea level rise, compounded by land conversation to agriculture and urban development, will cause all tidal flats to disappear by 2100 (Wu, Zhou and Tian, 2017). In another example, sea level rise and high rates of tidal inundation have increased overall salinity in the San Francisco Bay-Delta estuary, threatening the ecosystem’s ability to support biodiversity (Parker and Boyer, 2019).

Research on climate risks to urban trees and forests highlight direct impacts from extreme temperatures, precipitation, wind events and sea level rise, as well as exposure to other hazards such as air pollution, fires, invasive species and disease (Ordóñez and Duinker, 2014). Since the 1960s, climate change has enabled growth of urban trees, supported by longer growing seasons, higher atmospheric CO2 concentrations and reduced diurnal temperature range (Pretzsch et al., 2017), as well as increased fertilisation through urban-enhanced nitrogen deposition (Decina, Hutyra and Templer, 2020). However, these trends may change in the future as further warming and decreasing water supply may depress tree fitness, thus enabling more pests (Dale and Frank, 2017).

Climate risks to urban natural and ecosystem infrastructure entail significant economic costs. For example, in 2012, Hurricane Sandy led to total losses of up to USD 6.5 million to the New York City region’s low-lying salt marshes and beaches (Meixler, 2017). Research from coastal settlements across Catalonia, Spain, shows significant levels of tourism loss (which contribute to 11.1% of the region’s GDP), infrastructure damage and natural capital loss attributed to inundation and erosion of beaches, which are projected to retreat by −0.7 m yr −1 given current sea level rise projections of 0.53–1.75 m by 2100 (Jiménez et al., 2017).

Healthcare facilities (hospitals, clinics, residential homes) will suffer increasing shocks and stresses related to climate variability and change (Corvalan et al., 2020). Some may be sudden shocks from extreme weather events, which both threaten the facility, staff and patients and increase the number of people seeking health care. There are extensive reports of health facilities being damaged after major floods and windstorms (e.g., 2010 floods in Pakistan, Hurricane Sandy in the USA) which can be further exacerbated by power and water supply failures (Powell, Hanfling and Gostin, 2012). Disruption to services may persist for many months because of damage to buildings, loss of drugs and equipment, and damaged transport infrastructure significantly increasing travel time for patients (Hierink et al., 2020). The impacts of climate change on the health of residents of ‘slum’ settlements will also compound the existing health burdens faced by these individuals, including infectious disease and other environmental public health concerns (Lilford et al., 2016; Mberu et al., 2016).

Land use planning plays a major role in the siting of settlements and infrastructure. In relation to climate change, it affects whether development takes place in locations that are exposed to hazards; similarly, it shapes the potential effects that the built environment can have on natural systems. Despite this, generally speaking, there is limited implementation of zoning and land use measures for climate adaptation from cities across diverse contexts (robust evidence, high agreement ), see for example Maputo (Castán Broto, 2014), sub-Saharan cities (Dodman et al., 2017) and Amman, Moscow and Delhi (Jabareen, 2015). Certain countries, such as South Korea, have, however, recently begun to address disaster risk reduction within their land use planning systems (Han et al., 2019).

Conventional zoning regulations (in which only one kind of use is permitted in a given area) and land use planning range in scale from the regional to the local and can be deployed to minimise risks through protection, accommodation or retreat. Protection entails, in addition to allocating zones for protective urban infrastructure (such as seawalls, levees and dykes, and slope revetments), avoidance measures that restrict or prevent urban development (e.g., through growth containment and/or no-build zones). Accommodation involves land use modifications and/or conversions while retreat requires either compulsory or voluntary relocations and may entail buyouts (Butler, Deyle and Mutnansky, 2016; León and March, 2016; Lyles, Berke and Overstreet, 2018). Risk eliminating retreat measures are less widely adopted than other risk reducing zoning and land use measures (Anguelovski et al., 2016; Butler, Deyle and Mutnansky, 2016; Lyles, Berke and Overstreet, 2018). This is attributed to the controversies of relocation and to the complexities of buyouts (Butler, Deyle and Mutnansky, 2016; King et al., 2016).

Evidence from both richer countries and the Global South reveals that conventional zoning is more effective when governance systems facilitate the implementation of land use policies for climate adaptation that preclude negative human-nature interactions and that curb spatial inequity, both of which can trigger climate gentrification and increase the vulnerability of economically disadvantaged groups to climate-related risk (high confidence) (Marks, 2015; Liotta et al., 2020; Keenan, Hill and Gumber, 2018). Cascading benefits of zoning and land use planning for climate adaptation are associated with the use of soft land cover, green infrastructure and improvement of livability through better conditions for walkability and cycling. This decreases auto-dependency and contributes to health and economic development (by attracting businesses and retail that stimulate economic prosperity and increase property values) (Larsen, 2015; Carter et al., 2015). Such increases in property values have also been observed in zones and areas protected from risks (such as flooding), where it may trigger spatial inequity leading to climate gentrification (Marks, 2015; Votsis, 2017; Votsis and Perrels, 2016; Keenan, Hill and Gumber, 2018).

Adaptation actions through zoning and land use are more effective when combined with other planning measures (high confidence), for example with ecosystem-based adaptations (e.g., for flood management and curbing the urban heat island effect) (Larsen, 2015; Nalau and Becken, 2018; Perera and Emmanuel, 2018; Anguelovski et al., 2016; Carter et al., 2015; Tsuda and Duarte, 2018; Nolon, 2016); with community-based adaptations (trade-offs and valuations, i.e., which land uses are valued more) (Larsen, 2015; Nalau and Becken, 2018; Perera and Emmanuel, 2018; Anguelovski et al., 2016; Carter et al., 2015; McPhearson et al., 2018; Nolon, 2016); and with built form regulations and codes (León and March, 2016; Yiannakou and Salata, 2017; Perera and Emmanuel, 2018; Straka and Sodoudi, 2019; Larsen, 2015; Nolon, 2016). The imposition of planning-based tools such as scenario planning, flexible zoning and development incentivisation (among others) has the capacity to influence and encourage these adaptations (United States Environmental Protection Agency, 2017). Local risk-reduction inputs can inform land use adaptation policies (accommodation and/or avoidance, specifically growth containment and no-build zones) that are better integrated within larger urban plans (Lyles, Berke and Overstreet, 2018; Nalau and Becken, 2018; Tsuda and Duarte, 2018) (limited evidence, high agreement ).

Implementation of zoning and land use measures for climate adaptation from cities across diverse contexts remains limited (high agreement , robust evidence) owing to a range of challenges. A range of evidence from multiple locations indicates the challenges of mainstreaming land use planning for climate adaptation, including in Bangkok, Thailand (Marks, 2015), Legazpi City and Camalig Municipality in the Philippines (Cuevas et al., 2016; Cuevas, 2016), the USA (Cuevas et al., 2016; Cuevas, 2016), British Columbia, Canada (Stevens and Senbel, 2017), and Australia (Serrao-Neumann et al., 2017). Mainstreaming is hindered by a lack of clarity of implementation strategies for climate adaptation, insufficient funding, competing priorities (especially among professional planners and politicians), institutional challenges (see Jabareen’s [2015] study of 20 cities globally) and the need to fill data gaps and continuously update weather statistics (Oberlack and Eisenack, 2018) (medium evidence, high agreement ). At the same time, however, limited evidence from cities around the world such as the urban regions of Stuttgart and Berlin in Germany (Larsen, 2015), Greater Manchester in the UK (Carter et al., 2015), and Colombo in Sri Lanka (Perera and Emmanuel, 2018) reveals that risk reduction through zoning and land use can effectively protect and expand green infrastructure and soft land cover to alleviate pluvial flooding and decrease the urban heat island effect. This evidence points that one of the primary roles of land use planning is to guide the development of the urban form. As such, it underpins and establishes the basis for other infrastructure systems such as physical infrastructure and nature-based solutions (Morrissey, Moloney and Moore, 2018).

Understanding how livelihoods, particularly of the urban poor, are both impacted by climate risk and how they might be strengthened is central to understanding climate adaptation in cities and settlements (Dobson et al. 2015). Rapid urbanisation and expanding physical infrastructure do not have a clear relationship with improved outcomes for urban livelihoods of low-income residents (Soltesova et al., 2014). Municipal and national efforts need to be closely aligned with building adaptive capacity of residents themselves, often through community-based adaptation (Soltesova et al., 2014; Dobson, Nyamweru and Dodman, 2015). Social safety nets protect individuals or households from falling below a defined standard of living by providing cash, in kind and other social transfers to fight vulnerabilities (Islam and Hasan, 2019) including those associated with climate change impacts including food shocks. Strengthening the financial and social infrastructure of poor households is a critical component of adaptive and transformative capacity (Haque, Dodman and Hossain, 2014; Ziervogel, Cowen and Ziniades, 2016). Social safety nets are one mechanism for strengthening this capacity.

Social protection, or social security, is defined as the set of policies and programmes designed to reduce and prevent poverty and vulnerability throughout the lifecycle (ILO, 2017). Safety nets are intended to protect vulnerable households from impacts of economic shocks, natural hazards and disasters, and other crises. The UN policy frameworks for sustainable development, including the Sendai Framework for Disaster Risk Reduction 2015–2030, the new Strategic Framework 2018–2030 of the United Nations Convention to Combat Desertification (UNCCD) and UNFCCC, highlight the essential role of social protection in promoting comprehensive risk management (Aleksandrova, 2019). Since the term Adaptive Aocial Protection was introduced by the World Bank (2015) and the IPCC (2014), it has been an emerging strategic tool to integrate poverty reduction, disaster risk reduction and humanitarian development into adaptation to climate change (Béné, Cornelius and Howland, 2018; Aleksandrova, 2019; Watson et al., 2016).

Adaptive social protection (ASP) is defined as a resilience-building approach by combining elements of social protection, disaster risk reduction and climate change adaptation, so as to break the cycle of poverty and vulnerability of household by investing in their capacity to prepare for, cope with and adapt to all types of shocks, especially under climate change and other global challenges (Bowen et al., 2020; Ivaschenko et al., 2018). ASP has been justified as an effective instrument to build household and community resilience to climate extremes and slow-onset climate events such as sea level rise and environmental degradation (Schwan and Yu, 2018; Aleksandrova, 2019). In contexts of extreme poverty or climatic extremes, international development organisations, national provisions and market charities are complementary where family and kinship networks are weak and inadequate. To deal with short-term vulnerability to climate shocks, ASP can act as a crucial complement to risk management tools provided by communities and markets, tools which tend to be insufficient in the face of large or systemic shocks, by providing predictable transfers, developing human capital and diversifying livelihoods (Hallegatte et al., 2016). ASP can also facilitate long-term change and adaptation by improving education and health levels, as well as providing a proactive approach to managing climate-induced migration in both rural and urban areas (Schwan and Yu, 2018; Adger et al., 2014).

Many national ASP programmes are established to cover both rural and urban areas, however, only a small number of researchers pay attention to urban cases (Aleksandrova, 2019). ASP instruments can be classified into four major types as presented in Table 6.5 (Ivaschenko et al., 2018; ILO, 2017). ASP can contribute to both incremental and transformative interventions both at the system level (short-term and long-term coping strategies from communities) and at the beneficiaries’ level (vulnerable populations) (Béné, Cornelius and Howland, 2018; World Bank, 2015; Aleksandrova, 2019; Ivaschenko et al., 2018).

Table 6.5 | Four categories and examples of adaptive social protection.

ASP may be very good at reducing extreme poverty by helping to meet individual or household needs but not collective needs to mitigate long-term climate shocks. For example, few programmes consider risk assessment and climate-proof infrastructures as anticipatory measures to foster early action and preparedness (Aleksandrova, 2019; Costella et al., 2017). They therefore need to enable the adoption of forward-looking strategies for long-lasting adaptation (Tenzing, 2020). Some examples from China show social protection can improve adaptive capacity of urban communities with social medical insurance, housing subsidies, weather-index insurance, post-disaster construction, relocation planning, livelihood shift strategies, and so on. (Pan et al., 2015; Zheng et al., 2018b; Rao and Li, 2019; Song, Zheng and Lin, 2021). However, social protection may lead to maladaptation in urban policy when social security, or similar tools (for example insurance) to compensate for exposure deincentivise risk reduction (Grove, 2021). In many developing countries, high concentrations of poor and vulnerable groups living in disaster-prone zones of urban centres, new urban dwellers and informal residents are often excluded from community-based networks and social services (Aleksandrova, 2019). Risk transfer tools (such as insurance) and risk retention measures (such as social safety nets) can avoid and minimise the burden of loss and damage and limit secondary and indirect effects (Aleksandrova, 2019; Roberts and Pelling, 2018).

Inclusive, targeted, responsive and equitable social protection can support long-term transition toward more sustainable, adaptive and resilient societies (Hallegatte et al., 2016; Shi et al., 2018; Béné, Cornelius and Howland, 2018; Carter and Janzen, 2018; Adger et al., 2014). ASP systems can be cost effective and equitable when targeting accuracy, timely risk sharing (disaster assistance) and improved policy coherence. Carter and Janzen (2018) find that the long-term level and depth of poverty can be improved by incorporating vulnerability targeted social protection into a conventional social protection system. Countries at all income levels can set up ASP systems that increase resilience to natural hazards, but the systems need to identify cost–benefits and be scalable and flexible to adjust to future, increasing climate risk. Bastagli (2014) suggested a new design for effective social protection including: (i) increasing the amount or value of transfer; (ii) extending the coverage of beneficiaries; and (iii) introducing payments or new programmes of social protections. For social protection programmes to contribute more effectively to adaptation, they need to be better coordinated across a range of agencies; better integrated with climate data to anticipate times of need for vulnerable groups; and better aligned with other risk management instruments such as insurance (Agrawal et al., 2019).

There is growing evidence of the benefits of early warning systems for urban preparedness decision making and action for climate and weather-related hazards such as cyclones, hurricanes and floods (medium evidence; high agreement ) (Lumbroso, Brown and Ranger, 2016; Zia and Wagner, 2015; Marchezini et al., 2017). Climate forecasting is constantly evolving and becoming increasingly accurate. Global organisations such as the World Meteorological Organizations are increasingly focusing on new and emerging technologies such as crowdsourced data collection to support integrated city services and early warning systems (Baklanov et al., 2018). However, while climate forecasting is an increasingly central tool for risk management agencies, a focus on urban areas or key infrastructure is still considerably rare (Lourenço et al., 2015; Nissan et al., 2019; Harvey et al., 2019). The significant rise in urban risks poses significant challenges to humanitarian agencies. Humanitarian responses and local emergency management are vital for disaster risk reduction yet are compromised in urban contexts where it is difficult to confirm property ownership and where renters and informal dwellers are often excluded from decision-making and planning (Parker and Maynard, 2015; Maynard et al., 2017). Disaster survivors and growing urban refugee populations are often displaced across the city thereby complicating efforts to track and provide support (Maynard et al., 2017).

Existing early warning systems remain insufficient and the complexity of urban landforms makes accurate and detailed early warning difficult (medium evidence; high agreement ) (Jones et al., 2015). This is particularly the case in low- and middle-income countries (LMICs) where urban centres are often characterised by rapid expansion of interlinked formal and informal human settlements and land use zones. In such contexts, early warning services vary in effectiveness within the same urban centre (Allen et al., 2020c; Rangwala et al., 2018). Often, forecast-based action follows linear structures where forecast information is applied mainly for responding to negative impacts rather than anticipatory decision-making and preparation to avoid such impacts (Marchezini et al., 2017). Early warning systems are effective for warning of threshold breaching events including cyclonic activity and riverine flooding but less able to provide localised warning, though capability is rapidly increasing. Probabilistic risk forecasting and forecast based early action are only beginning to be applied to urban contexts and often those that are most vulnerable do not receive warnings regarding hazardous events (Nissan et al., 2019). There is less capacity for early warning systems in LMICs with key challenges linked to a lack of well-established risk baseline information; accessibility, communication and understanding of forecast information, as well as political and institutional barriers and limited resources and capacities to act on such information (Jones et al., 2015; Mustafa et al., 2015; Zia and Wagner, 2015; Marchezini et al., 2017; Gotgelf, Roggero and Eisenack, 2020). Political and institutional barriers to the incorporation of climate information to decision-making are not limited to LMICs (Harvey et al., 2019). For example, comprehensive studies on sectoral use of climate information in Europe revealed that, despite climate services becoming increasingly accessible and well resourced, there is limited organisational uptake of seasonal climate forecasts across key sectors (e.g., energy, transport, water and infrastructure) in informing their decision making processes (Soares and Dessai, 2016; Soares, Alexander and Dessai, 2018). This is due both to technical and non-technical barriers such as lack of awareness and knowledge of climate information and forecasting (Soares and Dessai, 2016; Soares, Alexander and Dessai, 2018).

Globally, a considerable diversity of tools and frameworks for urban resilience assessments are being developed at multiple scales (Arup and Rockefeller, 2015; Elias-Trostmann et al., 2018). These include hybrids such as ecosystem-based disaster risk reduction (Eco-DRR) (Begum et al., 2014).While important advances have been made in assessing urban resilience, much debate remains around such tools and assessment approaches regarding issues such as validation, dynamics in exposure and vulnerability, and appropriateness of generic methods in high-density urban settlements (Leitner et al., 2018; Cardoso et al., 2020; Rufat et al., 2019). Disaster impact and recovery time are strongly influenced by the behaviour and actions of individuals, communities, businesses, and government organisations (Meriläinen, 2020; Räsänen et al., 2020). For example, the review by Aaerts et al. (2018) shows how the limitations of existing flood risk assessment methods (which tend to account for human behaviour in limited terms) can be addressed through innovative flood-risk assessments that integrate behavioural adaptation dynamics. The study by Moghadas et al. (2019) highlights the importance of hybrid multi-criteria approaches for assessing urban flood resilience in Tehran, Iran. A growing literature shows how multidisciplinary and inclusive approaches that include Local knowledge can achieve greater accuracy in risk characterisation and support lasting impact of investments into more robust climate services (Aerts et al., 2018; Lourenço et al., 2015; Sword-Daniels et al., 2018; Singh et al., 2018; Nissan et al., 2019; Harvey et al., 2019; Simon and Palmer, 2020). This literature highlights the need for innovative approaches in urban contexts that transcend traditional approaches of local knowledge inclusion widely applied in rural contexts, such as participatory rural appraisal.

The inclusion of Local knowledge and Indigenous knowledge in urban vulnerability and risk assessments can strongly enhance local resilience, but its effectiveness is constrained by wider decision making and policy contexts dominated by top-down approaches (medium evidence; high agreement ) (Jones et al., 2015; Sword-Daniels et al., 2018; Nissan et al., 2019). Established non-state actors such as Shack and Slum Dwellers International are particularly effective at implementing inclusive approaches for local knowledge incorporation into urban decision-making. Climate change and disaster risk exacerbate existing problems of economic development, yet macro-economic planning seldom incorporates adaptation. Recent evidence also confirms the role of Indigenous knowledge and local knowledge in management practices to reduce climate risks through early warning preparedness and response (see also Section 6.3.2.3). These practices are particularly important where alternative early warning methods are absent. For instance, Abudu Kasei et al. (2019) show that Indigenous knowledge gathered through observations on changes in natural indicators (such as links between rainfall patterns, certain flora and fauna, and temperature changes) could be applied to develop early warning of climate hazards (floods and droughts) in informal urban settlements in African countries such as Ghana. Similarly, Hiwaski et al. (2015) show that observations of changes in the environment and celestial bodies are used to predict climate-related hazards in Indonesia, the Philippines and Timor-Leste where communities in turn use local materials and methods, and customary practices to respond to the impacts of climate change.

Insurance is a risk transfer mechanism for middle- and high-income countries, yet is less widely available in LMICs (Surminski and Thieken, 2017). Additionally, where insurance options do exist in LMICs, these are not usually available to large populations living or operating in the informal sector. Flood insurance is widely available in many Organisation for Economic Co-operation and Development (OECD) countries but the demand and uptake differ significantly across countries (Hanger et al., 2018). This financial tool is subject to increasing pressure under the changing climate, with growing concerns around affordability and availability. More integrative approaches are required, such as where changes in the insurance industry are closely linked to adaptation strategies, building standards and land use planning and their application (Cremades et al., 2018). This is particularly important in LMICs and of central concern for all insurance schemes is ensuring access, fairness and affordability for the most poor and vulnerable. However, there are some notable examples of low-income communities setting up their own disaster insurance mechanisms. For example, the Community Development Funds for the Baan Mankong upgrading programme in Thailand include disaster funds as insurance against housing damage (Archer, 2012). Such approaches also need to be more closely linked to existing urban risk management planning approaches where urban livelihoods are seldom integrated and informed by more dynamic risk reduction frameworks that consider adaptive cycles and how resilience changes over time (Beringer and Kaewsuk, 2018; Cremades et al., 2018).

Disaster risk management systems face increasing challenges in adapting to evolving risk profiles, shaped by expanding urban areas and changing environmental conditions associated with climate change. In addition to flooding, risk monitoring and management systems have recently shown considerable shortfalls in planning for and responding to increased fire risk such as the devastating Californian wildfires in October 2019 (Morley, 2020) and Australia’s unprecedented and catastrophic 2019–2020 wildfire season. Risk management has also been challenged by new risk experiences including wild/bush fires encroaching on expanding urban areas and fire outbreaks in densely populated informal settlements pose increasing threats to livelihoods, human health and habitats globally (see also Sections 2.4.4.2 and 2.5.5.2).

Climate resilient health systems are a vital part of adaptation to protect the most vulnerable from climate change (WHO, 2020). Cardiovascular fitness for example is a root cause of morbidity and mortality form heat stress (Schuster et al., 2017). The World Health Organization has developed a framework of climate-resilient health systems that addresses both mitigation and adaptation goals (WHO, 2015). Universal health coverage (UHC) is an essential component of climate-resilient health systems. In most countries, access to health services is better in urban than in rural areas. However, there remain large urban populations with insufficient coverage of health services (WHO and WB, 2015) and UHC tracking needs to take better account of inequalities in coverage, including differences in access within cities and further disaggregation of urban populations by income. Thus, health sector investment is an important tool in adaptive action and capacity. Analyses of health survey data shows that, globally, access to health care is increasing toward UHC targets (Lozano et al., 2020). Financing for global health has increased steadily in the last two decades and modelling shows this trend is likely to continue to 2050, but at a slower pace of growth and the current disparities in per-capita health spending persist between high and low/middle income countries, leading to insufficient health service coverage for the poorest populations (Chang et al., 2019a). Out-of-pocket spending is projected to remain substantial in LMIC and will remain the only means to access health care for many poor urban populations.

The WHO Operational Framework highlights the components that can be strengthened to adapt to extreme weather (e.g., health care workforce, information systems, etc.). The evidence is greatest for impacts on larger health facilities (such as hospitals) and there is less evidence regarding impacts on health service delivery outside these settings (smaller health facilities, pharmacies, first responders, public health inspectors, etc.). Improved building design and spatial urban planning (where facilities are located) are essential to increase resilience for higher temperature and flood risk (medium evidence; high agreement ) (WHO, 2021; Codjoe et al., 2020; Korah and Cobbinah, 2017). Public health systems rely on information systems (including disease and vector surveillance and monitoring) to identify new and emergent public health risks. Improvements to health surveillance will increase resilience, particularly for populations in informal settlements that are absent from health and vital registration systems.

City-level and local government adaptation planning is facilitated by information on health impacts (Reckien et al., 2015), highlighting the need for monitoring and surveillance and the need for local evidence-based risk assessments. Adaptation in the health sector can be limited by lack of collaboration between health and other sectors, although this is often easier to facilitate at the local level (Woodhall, Landeg and Kovats, 2021).

Since AR5, there has been significant growth in research about climate education and activism (Simpson, Napawan and Snyder, 2019; O’Brien, Selboe and Hayward, 2018; Hayward, 2021). Access to knowledge is an important determinant of well-being, inclusivity and livelihood mobility and of driving human behaviour. Knowledge systems include formal educational provision (capital assets, syllabus and human capital), informal learning based in social interaction and customary institutions (including through social media) and public communication (news media, government and other information systems including commercial messaging). There is a growing body of literature addressing the role of information and communication technology in shaping behaviour in disaster response and recovery and climate action, with particular focus on social media use and serious gaming (Houston et al., 2015; Carson et al., 2018) (see Section 6.3.4.3)

Given the amount of time that children spend in school settings, adapting educational infrastructure and programmes to climate change is highly important. This includes not only making physical structures safe, but also providing students with the knowledge and confidence to support individual and family-based adaptation. Several UN agencies (e.g., UNICEF and UNDRR) and international non-governmental agencies (e.g., Plan International) have prioritised safer schools and child-centred risk management that often focus on schools as places that should be prioritised for retrofitting and safe construction, but also as focal points for knowledge dissemination and community organising where impacts can extend beyond the school to reduce risk among students’ families. Universities and think tanks, as well as the third and private sector are key support mechanisms, particularly at the local level and when working in collaboration with local government and communities. They can support the development of critical educational resources and innovative communication methods, as well as facilitate the design and implementation of climate policies and related action plans.

Youth, adult communities, social media and the commercial media can have a significant impact on advancing climate awareness and the legitimacy of adaptive action, particularly in large urban areas (medium evidence, high agreement ). Climate change education in urban settlements has increasingly focused on enhancing children and young people’s political agency in schools, universities, and in formal and informal media settings (Cutter-Mackenzie and Rousell, 2019). However, an ambiguous framing of climate impacts and adaptation, for example around the science of urban heat islands by the media, can also exacerbate local community confusion and uncertainty (Iping et al., 2019) and further training and capacity building opportunities such as for vocational qualifications is still required across diverse settings (Simmons, 2021). Communication strategies deployed in formal education and social media can be highly influential in exchanging information and establishing narratives and viewpoints that frame what adaptive action is legitimate, especially in large cities (Simpson, Napawan and Snyder, 2019). However, the effectiveness of communication strategies for change, for example from Mayoral offices, can also be influenced by wider political and structural drivers including community literacy or political partisanship (Boussalis, Coan and Holman, 2019). Recent research (e.g., Macintyre et al., 2018) highlights the need for new learning approaches to climate education from school age to adult education. Emphasis is on inclusivity in learning and recognising diverse perspectives across multiple levels and settings, from formal and informal education to wider social learning. Informal learning that takes place outside of school settings, such as in libraries and botanical gardens, in everyday life is increasingly recognised as a key arena for climate education, life-long learning and nurturing environmental citizenship and activism (Paraskeva-Hadjichambi et al., 2020).

The integration of culture into urban policy and planning is increasingly recognised as critical to developing sustainable and resilient cities, and features in international agreements such as the SDGs (limited evidence; high agreement ) (Sitas, 2020). However, urban cultural policies are still limited, for example, Cape Town is the only African city to have developed a city-level cultural policy (Sitas, 2020). Cultural heritage refers to both tangible (e.g., historic buildings and sites) and intangible (e.g., oral traditions and social practices) resources inherited from the past (Fatorić and Egberts, 2020; Jackson, Dugmore and Riede, 2018). Learning about past societal and environment changes through heritage offers opportunity for reflection and transfer of knowledge and skills. This takes place in multiple contexts such as museums and cultural landscapes, and in everyday life (Fatorić and Egberts, 2020; Jackson, Dugmore and Riede, 2018). Cultural heritage is primarily associated with identity and is closely intertwined with the complexities of history, politics, economics and memory. Climate change adds another layer of complexity to cultural heritage and resource management (Fatorić and Seekamp, 2017b). Changing climatic conditions are already negatively impacting World Heritage Sites such as the Cordilleras’ Rice Terraces of the Philippines and earthen architecture sites, for example the Djenné mosque in Mali, are particularly vulnerable to changes in temperature and water interactions (UNESCO, 2021). Climate change impacts intangible cultural heritage across diverse settings such as in the Caribbean and Pacific Small Island Developing States (SIDS) where traditional ways of life and related aspects such as oral traditions and performing arts are under threat from extreme weather events (UNESCO, 2021).

The climate change adaptation options for built cultural heritage fall into seven categories (Rockman et al., 2016; Fatorić and Seekamp, 2017b). Financial constraints are the primary barriers that underpin the first four adaptation options: no action at all, merely monitoring and/or documenting, or annual maintenance (Xiao et al., 2019; Sesana et al., 2019; Fatorić and Seekamp, 2017a; Fatorić and Seekamp, 2017b; Fatorić and Seekamp, 2018). Core and shell preservation, the fifth and sixth categories, are cost effective when they improve the condition of built cultural heritage (BCH) (Bertolin and Loli, 2018; Loli and Bertolin, 2018a; Loli and Bertolin, 2018b), while elevation and/or relocation, the final adaptation options, are extremely costly and might jeopardise the historic value (Xiao et al., 2019). To date, however, evidence indicates that adaptation actions prioritise archaeological sites (Carmichael et al., 2017; Fatorić and Seekamp, 2018; Pollard et al., 2014; Dawson, 2013). The efficacy of adaptation of historic buildings can be increased through increased and stable funding, incentives, stakeholder engagement, and legal and political frameworks (Dutra et al., 2017; Fatorić and Seekamp, 2018; Fatorić and Seekamp, 2017b; Fatorić and Seekamp, 2017a; Leijonhufvud, 2016; Phillips, 2015; Sesana et al., 2019; Sesana et al., 2018; Sitas, 2020).

Other barriers to implementation include harnessing expert and local knowledge (of individuals and organisations) to identify both quantitative and qualitative methods and indicators that connect cultural significance and local values vis-à-vis climatic change over time and that move beyond the prevalent high-risk or high-vulnerability centred approaches (Carmichael et al., 2017; Fatorić and Seekamp, 2018; Haugen et al., 2018; Leijonhufvud, 2016; Pollard et al., 2014; Puente-Rodríguez et al., 2016; Richards et al., 2018; Dawson, 2013; Filipe, Renedo and Marston, 2017; Kotova et al., 2019). This is particularly important given that the significance of cultural heritage is often intangible, and its value cannot be determined solely through quantitative indicators. Accessing local resources (craftsmanship and materials compatible with the originals) can also improve built cultural heritage’s adaptation capacity (Phillips, 2015).

Effective decision-making and practice for adapting built and intangible cultural heritage requires open dialogue and exchange of cultural, historical and technical information between diverse stakeholders and decision makers (Fatorić and Seekamp, 2017b; Benson, Lorenzoni and Cook, 2016). As noted in Section 6.2.6, human behaviour can be a driving force for adaptation impacts on BCH at risk. Despite challenges associated with intangibility, socio-cultural heritage such as Indigenous knowledge (e.g., food security and water management practices) presents important opportunities for climate adaptation and resilience building. More research is needed across diverse contexts to understand feasible climate adaptation measures, and barriers and opportunities for building the resilience of both built and intangible cultural heritage, as well as to increase awareness of cultural heritage benefits among climate change policymakers (Fatorić and Egberts, 2020).

Nature-based strategies, including street trees, green roofs, green walls and other urban vegetation, can reduce heat and extreme heat by cooling private and public spaces (robust evidence, high agreement ). Shading and evapotranspiration are the primary mechanisms for vegetation-induced urban cooling (Coutts et al., 2016). Shading reduces mean radiant temperature, which is the dominant influence on outdoor human thermal comfort under warm, sunny conditions (Thorsson et al., 2014; Viguié et al., 2020). Outdoor green space and parks may also slightly reduce indoor heat hazard (Viguié et al., 2020). Apart from lowering temperature, NBS may also contribute to lower energy costs by reducing extra demand for conventional sources of cooling (e.g., air conditioning) (Viguié et al., 2020; Foustalieraki et al., 2017), especially during peak demand periods. Homes with shade trees that are located in cities where air conditioning systems are common can save over 30% of residential peak cooling demand (Zardo et al., 2017; Wang et al., 2015). Green roofs have been shown to significantly lower surface temperatures on buildings (Bevilacqua et al., 2017) and modelling suggests that green roofs, if employed widely throughout urban areas, have the potential to impact the regional heat profile of cities (Bevilacqua et al., 2017; Rosenzweig, Gaffin and Parshall, 2006). Community or allotment gardens, backyard greening and other types of low vegetation, as well as lakes, ponds, rivers and streams, can also provide local cooling benefits to nearby residents (Gunawardena, Wells and Kershaw, 2017; Larondelle et al., 2014; Santamouris, 2020).

Urban climate models show that increased vegetation cover results in reducing both mean air temperatures and extreme temperatures during heatwaves (Heaviside, Cai and Vardoulakis, 2015; Ferreira and Duarte, 2019; Schubert and Grossman-Clarke, 2013). Greater density and more canopy coverage relative to other built and paved surfaces increases shade provision and evapotranspiration (Hamstead et al., 2016; Grilo et al., 2020; Herath, Halwatura and Jayasinghe, 2018; Knight et al., 2021). However, local cooling by vegetation depends on regional climate context, geographic setting of the city, urban form, the density and placement of the trees, in addition to a variety of other ecological, technical, and social factors, such as local stewardship (Salmond et al., 2016). Green spaces less than 0.5–2.0 ha may have negligible cooling effects at regional scales, but impacts of shading can have microscale cooling benefits (Gunawardena, Wells and Kershaw, 2017; Zardo et al., 2017). Vegetation impacts on day versus night-time cooling varies (Imran et al., 2019) as does cooling potential in temperate versus tropical climates. The supply of cooler air from surrounding peri-urban and rural areas can impact cooling in the urban core suggesting that regional adaptation planning for NBS is important to maintain or extend ventilation paths from the urban fringe into the city centre (Schau-Noppel, Kossmann and Buchholz, 2020).

To maximize the adaptation benefits of NBS for regulating urban heat, it can be helpful to prioritise tree planting and other urban greening investments in areas where heat vulnerability and risk are the highest, especially communities that lack urban tree canopy or accessibility to parks to cool off during hot days or heatwaves (Ziter et al., 2019). Planting trees closely together or in partly permeable vegetated barriers along streets can improve local cooling benefits. Additionally, choosing tree species with leaves that have the greatest leaf area index or the largest leaves can improve cooling performance, as those trees have the greatest shading and evapotranspiration benefits that, in turn, provide the greatest cooling effects (Keeler et al., 2019). Drought-resistant trees, often native trees, are ideal to avoid high watering costs, though dry or water scarce areas may limit adoption of urban vegetation as an NBS strategy (Coutts et al., 2013). Native trees and permaculture can provide additional benefits for local biodiversity as shown in study in Melbourne, Australia which found that increasing vegetation from 10% to 30% increased occupancy of bats, birds, bees, beetles and bugs by up to 130% (Threlfall et al., 2017), with particularly high impact on native species.. Additionally, planting fruit or nut trees can provide co-benefits for local food production, and yet choice of species and placement is important to consider with respect to local cultural needs and norms (Adegun, 2018; Adegun, 2017).

NBS in cities can help regulate air quality by absorbing air pollutants (medium evidence, medium agreement ). For example, planting trees or vegetated barriers along streets or in urban forests can reduce particulate matter, the ambient air pollutant with the largest global health burden (Janhäll, 2015; Tiwary, Reff and Colls, 2008; Matos et al., 2019; McDonald et al., 2016). However, findings show that trees can also positively affect ground-level ozone (Calfapietra et al., 2013; Kroeger et al., 2014), airborne pollen concentrations (Willis and Petrokofsky, 2017) and indirectly affect air quality through reduced emissions from energy production offset by shade provision (Keeler et al., 2019). Certain tree species however can also be detrimental to urban ozone formation by emitting significant amounts of reactive biogenic volatile organic compounds (VOCs). Decreasing urban emissions of VOCs is an increasingly important ozone mitigation strategy in urban areas (Fitzky et al., 2019).

Trees can also have negative effects by increasing pedestrian exposure to pollution if they are introduced in heavily travelled street canyons where air pollutants can be trapped (Vos et al., 2013; Gromke and Blocken, 2015). To maximise the adaptation benefits of NBS for improving air quality, planners and managers can target tree selection for species with low VOC emissions, low allergen emissions and high pollutant deposition potential (Keeler et al., 2019), and combine with low pollution transportation policies. Studies suggest sensitive planting of roadside tree canopies can have positive effects on air pollutants (Beckett, Freer Smith and Taylor, 2000; Yang, Chang and Yan, 2015). For example, Xue et al. (2021) found that the PM2.5 reduction between 2013 and 2017 in China was associated with a saving of approximately USD 111 billion yr -1 nationally. Tree planting near schools, nursing homes and hospitals can ensure that benefits provided by trees are delivered to the local populations that stand to benefit the most from improved air quality, but species need to be adapted to regional climate to provide benefits over time (Donovan, 2017; Nowak et al., 2018).

Urban parks and open spaces, forests, wetlands, green roofs and engineered stormwater treatment devices help manage stormwater and wastewater by reducing the volume of stormwater runoff, reducing surface flooding, and reducing contamination of runoff by pollutants (robust evidence, high agreement ). Engineered devices include bioswales, rain gardens, and detention and retention ponds, and are becoming common and standard approaches to mitigate the negative effects of impervious surfaces on stormwater quality and surface flooding in cities (Zhou, 2014; McPhillips et al., 2020). Allotment gardens, street trees, green roofs and urban forests may also help reduce runoff and provide a stormwater retention service (Pennino, McDonald and Jaffe, 2016; Berland et al., 2017; Gittleman et al., 2017). Modelling and empirical studies show that NBS at small spatial scales lead to improvements in water quality and reduction of peak flows (Moore et al., 2016; Keeler et al., 2019; Webber et al., 2020). Peak flow reductions are greatest for small rain events. For example, D-Ville et al. (2018) observed a 30–70% reduction in peak flow for the 1-in-30 year storm, but performance reduces for more intense rainfall or if saturated (Garofalo et al., 2016). Employing NBS to reduce flooding on roads can be an important adaptation mechanism for reducing the impact of flooding events on traffic flows (Pregnolato et al., 2016).

During periods with intense precipitation, low-lying urban parks and open space, engineered devices and wetlands can play an important role in reducing stormwater runoff volumes by providing places for water to be stored and infiltrate during heavy storms (Moore et al., 2016). However, the magnitude of the runoff reduction service will depend on the total area of green infrastructure, vegetation type and its position on the landscape. There is less evidence of the effectiveness of NBS at larger temporal and spatial scales (Pregnolato et al., 2017; Jefferson et al., 2017). The performance of NBS depends on the degree to which their extent and spatial configuration in the city are optimised to capture runoff (Fry and Maxwell, 2017). Investing in a diversity of NBS types may be important to maximise stormwater management and flood regulation, as different types of engineered NBS have different strengths and weaknesses.

Overall, NBS are attractive adaptation options for stormwater management and to reduce impacts of pluvial and fluvial flooding in cities (Rosenzweig et al., 2018a) compared, and in combination, with grey infrastructure. Cities with combined sewer infrastructure are likely to see benefits from NBS due to reductions in stormwater quantity and reduced sewage overflows. Cities where a large proportion of residents lack access to piped infrastructure and drink surface water may see large benefits, especially to human health, from NBS investments (Keeler et al., 2019). Where future large-scale upgrades or installation of grey infrastructure will be necessary, new and growing cities may have more opportunity to realise large net benefits from investments in NBS. Older cities, and new, rapidly urbanising areas that lack large-scale water infrastructure may see the greatest benefits from enhanced NBS, relative to cities where heavy investments infrastructure upgrades have already been made. Cities facing climate changes that include more frequent or extreme precipitation may also see large water quality benefits from investment in NBS (Keeler et al., 2019). Overall, there is increasing evidence that NBS for addressing stormwater is cost effective (Bixler et al., 2020; Kozak et al., 2020; Mguni, Herslund and Jensen, 2016), especially in cities facing a need to update current infrastructures.

Coastal ecosystems including coral and oyster reefs, coastal forests including mangroves and other tree species, salt marshes and other types of wetland habitat, seagrass, dunes and barrier islands can reduce impacts of coastal flooding and storms (robust evidence, high agreement ) (Zhao, Roberts and Ludy, 2014; Boutwell and Westra, 2016; Narayan et al., 2017; Yang, Kerger and Nepf, 2015; Bridges et al., 2015; World Bank, 2016) (see also Section CCP2 Cities and Settlements by the Sea). Recent literature highlights the value of nature-based approaches for coastal protection in terms of avoided damages and human well-being (Narayan et al., 2017; Silva et al., 2016a). NBS can protect coasts from flooding through reducing the wave energy by drag friction, reducing wave overtopping by eliminating vertical barriers, and absorbing floodwaters in soil (Arkema, Scyphers and Shepard, 2017; Dasgupta et al., 2019; Zhu et al., 2020). For example, coastal and marine vegetation and reefs can dissipate wave energy, attenuate wave heights and nearshore currents, decrease the extent of wave run-up on beaches, and trap sediments (Ferrario et al., 2014; Bridges et al., 2015). These effects result in lower water levels and reduce shoreline erosion, which in turn has potential to save lives and prevent expensive property damages (Narayan et al., 2017).

Researchers, practitioners and policy-makers are increasingly calling for the use of nature-based approaches to protect urban shorelines from coastal hazards (Cunniff and Schwartz, 2015; Bilkovic et al., 2017). The expectation is that coastal ecosystems can help stabilise shorelines, protect communities against storm surge and from tidal-influenced flooding, while providing other co-benefits for people and ecosystems. However, vegetation along protected coastlines with higher frequency, lower intensity coastal hazards (National Research Council, 2014) may be more effective for stabilising shorelines and reducing risk to coastal communities and properties, and benefits will depend on local hydrology of the coastal region. Narayan et al. (2017) estimate that coastal wetlands alone reduced direct flood damages by USD 625 million during Hurricane Sandy in the USA in 2012. Similarly, researchers found that villages with wider mangroves between them and the coast experienced significantly fewer deaths than villages with narrow or no mangroves during a 1999 cyclone in India (World Bank, 2016). Recently, Arkema et al. (2017) noted that the number of people, poor families, elderly and total value of residential property most exposed to hazards along the entire coast of the USA can be reduced by half if existing coastal habitats remain fully intact.

Coastal habitats also have limitations in their ability to protect coasts from extreme events. Some studies suggest reduced effectiveness of vegetation and reefs for coastal protection from large storm waves and surge (Möller et al., 2014; Guannel et al., 2016) and there is active debate in the literature about the ability of ecosystems to mitigate the impact of tsunamis (Gillis et al., 2017). Further research is needed to understand and quantify coastal protection services provided by these hybrid green-grey solutions, especially in urban areas (Bilkovic et al., 2017). Additionally, in some coastlines, water may be too deep or waves too high for some species such as mangroves to grow, thrive and provided needed NBS.

Maximising the adaptation benefits of NBS for improving coastal flood protection research requires that cities seek to restore and conserve the vegetation and reef types that are appropriate for the exposure setting and in sufficient abundance to be effective. In particular, planners and managers can use vegetation in protected bays as alternatives to hard infrastructure for shoreline stabilisation. However, the influence of ecosystems on flooding and erosion is variable and depends on a suite of social, ecological and infrastructural factors that vary within and among urban areas (Narayan et al., 2017; Ruckelshaus et al., 2016; Bridges et al., 2015). Additionally, long-term planning to restore or ensure resilience of individual species and ecosystems that may themselves be damaged or destroyed during extreme events is needed in order for urban green and blue infrastructure to continue providing NBS over the longer term.

NBS reduce both the volume of floodwater and the impact of floods (medium evidence, medium agreement ). NBS reduce the volume of runoff by increasing infiltration and water storage (Shuster et al., 2005; Salvadore, Bronders and Batelaan, 2015), and affect the production and impact of flood waters through reducing river energy and flow speed through physical blockage, stabilising riverbanks during flood events, creating space for floodwaters to expand and combating land subsidence (Palmer, Filoso and Fanelli, 2014; Ahilan et al., 2018). Installing NBS to increase infiltration on low slopes and high-permeability soils can reduce the impacts of potential increases in urban flooding driven by climate change, especially for small- to medium-scale flood events (lower than 20% mean annual flood) (Moftakhari et al., 2018).

Source reduction strategies include creating permeable areas such as parks and open spaces, as well as engineered devices such as raingardens, bioswales and retention ponds that help retain stormwater runoff from impervious areas. River restoration can reduce flood peak flow and provide space for floodwaters to expand. Planting and maintaining vegetation along riverbanks, often in the form of parks or river restoration, maintains structural integrity during flood events. Wetland construction and improved connectivity to floodplains also reduces flood peaks. Efforts to restore floodplains are important to create space for floodwaters and reduce exposure by moving people out of the hazard zone. Floodplain restoration also provides access to the river that has multiple benefits including recreation, access to water for domestic use and other cultural ecosystem services. A key adaptation strategy is to reduce streambank erosion (a result of high peak flow) using riparian vegetation to stabilise riverbanks during flood events.

Cities manage flood risk using different types of adaptation and regulatory mechanisms (Naturally Resilient Communities, 2017). Built flood-control infrastructure, such as levees and stream channelisation, reduces the demand for nature-based flood impact reduction. Cities facing flood risk that do not currently have extensive grey flood-mitigation infrastructure may find NBS to be an appealing, lower cost solution (Keeler et al., 2019). In cities where flood-control grey infrastructure already exists, there is less demand for NBS of flood protection, but NBS may provide important back up, especially in a changing climate that may increase flood hazards (City of Los Angeles, 2017; Elmqvist et al., 2019). Overall, city and basin-wide NBS for riverine flood impact reduction can reduce the generation of new hazards by making space for water which can reduce the potential for a false sense of security provided by traditional flood management approaches (Ruangpan et al., 2020; Turkelboom et al., 2021).

The role of NBS has been increasingly recognised for improving urban water management, emphasising it’s contribution for climate-adapted development and sustainable urbanisation (robust evidence, high agreement ) (Wong and Brown, 2009). NBS that protect or restore the natural infiltration capacity of a watershed can increase the water supply service to various extents, improving drought protection and providing resilient water supply (Drosou et al., 2019; Krauze and Wagner, 2019), although different forms of NBS (e.g., street trees, parks and open space, community gardens, and engineered devices such as rain gardens, bioswales or retention ponds) contribute in different ways to increasing stormwater infiltration. Additional sources of water may be available to replace the water supplied by NBS, such as rainwater harvesting, inter-basin transfers or desalination plants. Reliance on naturally sourced, locally available surface water and groundwater is more energy efficient and economical than desalination or water reuse for potable use (Boelee et al., 2017), while rainwater harvesting is even more economical. Increasing the amount of green space in urban areas can secure and regulate water supplies, improving water security (Liu and Jensen, 2018; Bichai and Cabrera Flamini, 2018). However, Bhaskar et al. (2016) reviewed the effect of urbanisation and NBS on baseflow and suggest that the confounded effects of infiltration and evapotranspiration losses, combined with the subsurface infrastructure (sewer systems) and geology, makes it difficult to predict the magnitude of baseflow enhancement resulting from the implementation of NBS in cities.

To maximise the adaptation benefits of NBS for urban water supply research suggests that managers and planners consider NBS as alternatives to traditional stormwater management techniques, where possible, since these solutions can promote groundwater recharge. As green infrastructure is increasingly being used for stormwater absorption in cities (McPhillips et al., 2020), rain gardens, wetlands, or engineered infiltration ponds and bioswales are the NBS most likely to promote recharge, reduce evapotranspiration and contribute to water provisioning.

Urban agriculture can serve as a NBS for food security (medium evidence, medium agreement ) across a range of urban contexts (Lwasa and Dubbeling, 2015; Nogeire-McRae et al., 2018; Pourias, Aubry and Duchemin, 2016) by contributing to food provisioning as well as providing co-benefits including for recreation, place making and mental health (Petrovic et al., 2019; Soga, Gaston and Yamaura, 2017; Goldstein et al., 2016b).

Urban agriculture among poorer communities in lower income areas is already an important source of food supply for those communities, contributing to food security and health (Orsini et al., 2013). However, potential for expanding open air urban food production may be practically constrained by land availability (Badami and Ramankutty, 2015; Martellozzo et al., 2014). This is particularly true in some lower-income countries where rapid urbanisation is occurring, which compounds existing food insecurity (Satterthwaite, McGranahan and Tacoli, 2010; Vermeiren et al., 2013). Land availability and suitability for gardens can be further constrained by land use history, including past industrial uses that can contaminate soils with pollutants such as lead.

At the same time, investments in vertical agriculture continue to expand, such as in Singapore where private investment in food production is occurring in high rise buildings (Wong, Wood and Paturi, 2020). Not all cities can benefit similarly from vertical agriculture since higher heating costs to produce vegetables indoors during northern winters consumes considerable amounts of energy and may generate fossil fuel emissions depending on the energy source (Goldstein et al., 2016a; Mohareb et al., 2017). Some regions can benefit from more traditional outdoor urban farming, such as in South and Southeast Asia, which can support multiple growing cycles per year for some crops, particularly in tropical areas where irrigation is available. Light availability, soil health and water availability will impact food production in urban areas. For example, a study conducted in Vancouver, Canada, demonstrated that light attenuation from buildings and trees can reduce both crop yield and water demand for crop growth (Johnson et al., 2015).

Climate change may have important impacts on urban food production and food security. While urban agriculture may provide benefits in terms of stability of food access in low-income households in some regions of the Global South where the climate is warmer, the shorter growing seasons in colder climates will reduce the role of outdoor urban agriculture in year-round food supply and diets. Though urban agriculture constitutes a small fraction of total food consumption in some urban areas, several studies have attempted to estimate the extent to which urban agriculture could theoretically meet urban total food or vegetable demand (Badami and Ramankutty, 2015; McClintock, 2014; Hara et al., 2018). Maximising the adaptation and resilience benefits of NBS for food production and security suggests the need to embrace the multi-functionality of urban agriculture, rather than viewing it as solely concerning food production (Barthel, Parker and Ernstson, 2015).

Urban morphology describes the overall status of cities as physical, environmental and cultural entities. Cities interact with surrounding environmental processes, for example, as documented in Section 6.2 by influencing urban temperature, but also precipitation and through coastal and riverine development fluvial and coastal sedimentary regimes of erosion and deposition that impact on flood risk. Rapid, increased urbanisation has contributed to observed flood risks in recent decades (see Section 5 4.2.4; Tramblay et al., 2019). The design process for physical infrastructure projects and significant construction (e.g., residential or industrial estates and large industrial development) typically includes risk assessments and social and environmental impact assessments that consider neighbouring land uses and connected infrastructure. Land use planning can consider diverse land uses and their interactions at the neighbourhood level (Section 6.3.2.1). Resilience planning aims to bring together integrated, systemic views and enable joined-up planning at the city level (as well as lower scales) (Section 6.3.2.1). There is however a lack of long-term studies that assess the climate change impacts on urban form, including informal settlements (Bai et al., 2018; Ramyar, Zarghami and Bryant, 2019), leading to impact assessments that often overlook urban form (Ramyar, Zarghami and Bryant, 2019). Additionally, context-specific spatial tools and community based approaches lack a precise connection to urban morphology. For example, there is a need for further studies that connect solar radiation, urban morphology (e.g., aspect and plot ratio), and the urban heat island spatio-temporal variability (Giridharan and Emmanuel, 2018; Li et al., 2019c).

Several tools and models have emerged in response to recommendations from AR5, including models that assess the impacts of urban heat island (Ramyar, Zarghami and Bryant, 2019), climatic uncertainty (Dhar and Khirfan, 2017), flood vulnerability (Abebe, Kabir and Tesfamariam, 2018) and inundation (Barau et al., 2015; Ford et al., 2019). For example, findings from Kano, Nigeria, reveal that a lack of distribution of certain urban morphological features, including open spaces and streets (both pervious and impervious), roof and building materials (e.g., concrete and metallic) and urban ecological features (e.g., urban ponds and ecological basin), exacerbates inundations and their associated impacts (Barau et al., 2015). Also, findings about the urban forms of coastal settlements, particularly in small islands, reveal that they often experience severe beach erosion due to wave action, sea level rise and storm surge that leads to landward retreat of coastline which threatens their social and economic activities (Dhar and Khirfan, 2016; Lane et al., 2015; Khirfan and El-Shayeb, 2019). Despite these examples, very limited research is available to offer assessments of different urban scale morphologies and urban scale adaptation planning, including planning adaptation across supply chains and networked relationships with distant urban and rural places connected through trade and resource (financial, human and material) or waste flows.

Interventions in the morphology and built form of cities can contribute to the reduction of the urban heat island effect and reduce the consequences of urban heatwaves. These can include installing air conditioning, establishing public cooling centres (i.e., for use during heatwaves), pavement watering (Parison et al., 2020a) and increasing surface albedo through ‘cool roofs’ (i.e., with high-reflectance materials) and walls. Air conditioning can significantly increase the local urban heat island (Salamanca et al., 2014; Wang et al., 2019a) and the choice of refrigerant has a significant impact on global warming potential (McLinden et al., 2017). The relative efficiency of cool roofs compared with green roofs is variable, because while white roofs have similar potential to reduce the urban heat island (Li, Bou-Zeid and Oppenheimer, 2014), they can quickly turn grey due to dust and air pollution, losing their effectiveness (Gunawardena, Wells and Kershaw, 2017), although these effects are now well studied and newer performance standards should account for ageing and soiling effects on reflectivity (Paolini et al., 2014). Ageing of ‘cool pavements’ is more complex, which makes their long-term performance less reliable to predict (Lontorfos, Efthymiou and Santamouris, 2018). The cooling performance of green roofs is highly variable and depends on the actual water content of the green roof substrate, with dry vegetation performing poorly in terms of cooling (Parison et al., 2020b). This holds true for regular vegetation and NBS in general (Daniel, Lemonsu and Viguie, 2018). For all built environment adaptations, changes are locked-in for a long time, and are likely to be expensive so that care is needed to avoid potential negative impacts on social equity (Cabrera and Najarian, 2015; Romero-Lankao et al., 2018; Fried et al., 2020; Rode et al., 2017) and carbon-intensive construction (Bai et al., 2018; Seto et al., 2016).

Architectural and urban design regulations at the single-building scale (building codes and guidelines) facilitate climate responsive buildings that adapt to a changing climate and have the potential to collectively change user behaviour during extreme weather events (Osman and Sevinc, 2019). They include buildings that are adaptive to ensure user comfort during extremes of hot and cold as well as to floods (e.g., building on stilts and amphibian architecture). Changes to design standards can scale quickly and widely, but retrofit of existing buildings is expensive, so care must be taken to avoid potential negative impacts on social equity (Schünemann et al., 2020; Matopoulos, Kovács and Hayes, 2014; Ajibade and McBean, 2014; Bastidas-Arteaga and Stewart, 2019). Buildings can be adapted to the negative consequences of climate change by altering their characteristics, for example increasing the insulation values (e.g., van Hooff et al., 2014; Makantasi and Mavrogianni, 2016; Fisk, 2015; Fosas et al., 2018; Barbosa, Vicente and Santos, 2015; Invidiata and Ghisi, 2016; Pérez-Andreu et al., 2018; Taylor et al., 2018; Triana, Lamberts and Sassi, 2018), adding solar shading (e.g., van Hooff et al., 2014; Makantasi and Mavrogianni, 2016; Barbosa, Vicente and Santos, 2015; Invidiata and Ghisi, 2016; Pérez-Andreu et al., 2018; Taylor et al., 2018; Triana, Lamberts and Sassi, 2018; Dodoo and Gustavsson, 2016; Osman and Sevinc, 2019), increasing natural ventilation, preferably during the night (e.g., van Hooff et al., 2014; Makantasi and Mavrogianni, 2016; Pérez-Andreu et al., 2018; Triana, Lamberts and Sassi, 2018; Dodoo and Gustavsson, 2016; Osman and Sevinc, 2019; Mulville and Stravoravdis, 2016; Cellura et al., 2017; Fosas et al., 2018; Dino and Meral Akgül, 2019), solar orientation of bedroom windows (Schuster et al., 2017), applying high-albedo materials for the building envelope (van Hooff et al., 2014; Invidiata and Ghisi, 2016; Baniassadi et al., 2018; Triana, Lamberts and Sassi, 2018), altering the thermal mass (van Hooff et al., 2014; Mulville and Stravoravdis, 2016; Din and Brotas, 2017), adding green roofs/facades to poorly insulated buildings (Geneletti and Zardo, 2016; Skelhorn, Lindley and Levermore, 2014; van Hooff et al., 2014; de Munck et al., 2018; Feitosa and Wilkinson, 2018) and for water harvesting (Sepehri et al., 2018).

In general, the most promising adaptation measures are a combination of solar shading with increased levels of insulation and ample possibilities to apply natural ventilation to cool down a building (e.g., van Hooff et al., 2014; Makantasi and Mavrogianni, 2016; Fosas et al., 2018; Barbosa, Vicente and Santos, 2015; Taylor et al., 2018; Triana, Lamberts and Sassi, 2018; Dodoo and Gustavsson, 2016). However, it must be noted that the cooling potential of natural ventilation will decrease in the future because of increasing outdoor air temperatures (Gilani and O’Brien, 2020). Increased insulation (including through green solutions) without shading and ventilation can also lead to adverse impacts through the lowering of nighttime cooling (Reder et al., 2018). Similarly, air conditioning performance also decreases with increasing outdoor temperatures, in addition to being maladaptive where use increases anthropogenic heat emissions into the urban area, and global greenhouse gas emissions if powered by carbon intensive energy systems (Wang et al., 2018c).

Passive cooling is a design-based, widely used strategy to create naturally ventilated buildings, making it an important alternative to address the urban heat island for residential and commercial buildings (Al-Obaidi, Ismail and Rahman, 2014). Generally, passive cooling is achieved by controlling the interactions between the building envelope and the natural elements. Façade fixes such as overhangs, louvres and insulated walls are effective at shading buildings from solar radiation, while complex ones such as texture walls, diode roofs and roof ponds are effective at minimising heat gains from solar radiation and ambient heat (Oropeza-Perez and Østergaard, 2018). Passive cooling is inspired also by traditional design forms, for example from Mediterranean, Islamic and Mughal architecture in the Indian sub-continent (Di Turi and Ruggiero, 2017; Izadpanahi, Farahani and Nikpey, 2021).

In addition, wind towers, solar chimneys and air vents are features that facilitate cool air circulation within buildings while dissipating heat (Bhamare, Rathod and Banerjee, 2019). These features may be arranged to address hotspots or highly frequented spaces within buildings. Similar to NBS, the effectiveness of passive cooling to ameliorate the urban heat island varies widely depending on the location of the sun, wind direction and the type of strategy used. For instance, natural ventilation strategies (e.g., wind towers, solar chimneys, etc.) have shown temperature reductions of up to 14°C (Bhamare, Rathod and Banerjee, 2019; Calautit and Hughes, 2016; Rabani et al., 2014). Shading strategies alone can reduce indoor temperatures by 3°C, while heat sinks (in which heat is directed at a medium such as water) may result in indoor temperatures up to 6°C lower than the outdoor temperature (Oropeza-Perez and Østergaard, 2018). More systemic interventions, such as altering urban form through urban planning, can mitigate the urban heat island across suburbs and cities (Lee and Levermore, 2019; Takkanon and Chantarangul, 2019; Yin et al., 2018; Liang and Keener, 2015; Emmanuel and Steemers, 2018). Experience in Kano (Nigeria) has shown that incorporating Indigenous knowledge into building design and urban planning can increase resilience to heat and flood risks (Barau et al., 2015). A review by Lemi (2019) suggests that traditional ecological knowledge can provide wider climate change adaptation benefits.

Limits on housing and building adaptation include failure of regulatory systems so that formal design standards are not followed even when legally required (Arku et al., 2016; Durst and Wegmann, 2017; Pan and Garmston, 2012; Awuah and Hammond, 2014). This can be a result of pressures from clients for cheaper structures, developers illegally cutting costs or regulators lacking capacity for enforcement. Technological innovation can also be slow to embed itself in building norms and standards. Innovation also lies outside the formal sector and can include artisanal building techniques that may have adaptive value. Examples from Latin America demonstrate how initiatives in informal settlement improvement associated with housing policy, guaranteeing access to land and decent housing, show the opportunity for overarching policies encompassing development, poverty reduction, disaster-risk reduction, climate-change adaptation and climate-change mitigation (see Section 12.5.5).

Information and communication technologies (ICTs) are deeply intertwined with the functioning of urban and infrastructure systems, and are at the core of the ‘smart city’ concept (Angelidou, 2015). ICT is more flexible than other physical infrastructure, although as other sectors are increasingly reliant on ICT, it is creating new climate-related failure mechanisms (Norman, 2018; Maki et al., 2019). ICT assets and networks in urban, national and international communications systems will need to be strengthened to enable ICT infrastructure to better cope with climate change, and to enable ICT infrastructure to support the resilience of cities, settlements and other infrastructure. The increased pervasiveness of ICT in smart cities, smart infrastructure and day-to-day living, will evidently have long-term implications for exposure to climate change risks and how cities manage those risks (Norman, 2018; Maki et al., 2019). For example, even if the ICT network is resilient to heatwaves, it is dependent on the electricity network to power it. Conversely, other networks are dependent upon ICT for control systems, for example smart grids for energy. There is limited information on how these interdependencies, and associated risks, will evolve.

Although networked like many other infrastructure systems, ICT components have some distinctive properties. They are relatively cheap, and the advent of wireless communications has enabled ICT to have the widest reach of all infrastructures. Components can be rapidly deployed or repaired, and generally ICT networks are therefore built with inherent redundancy and flexibility (Sakano et al., 2016). Components have a wide range of expected lifetimes which leads to faster cycles of innovation. There is therefore greater potential to accelerate uptake of climate resilience in this infrastructure sector, but conversely, this can increase waste and (energy intensive) resource consumption. For example, mobile phones and computers may last as little as a year, cables and switching units may be moved and upgraded to improve bandwidth every few years, poles and masts are typically designed to last several decades, whilst exchanges and other critical nodes can be in use for over half a century.

ICTs are playing an increasing role in resilience building and enabling climate change adaptation. They are enabling access to information needed for decision making, facilitating learning and coordination among stakeholders, and building social capital, as well as helping to monitor, visualise and disseminate current and future climate impacts (Eakin et al., 2015; Heeks and Ospina, 2019; Haworth et al., 2018; Imam, Hossain and Saha, 2017). Advocacy and awareness raising through ICTs, such social media applications, can influence behaviours and attitudes in support of adaptive pathways (Laspidou, 2014).

ICTs play a role in adaptive responses to both short-term shocks and long-term trends associated with climate change. Timely access to information (e.g., early warning, temperature and rainfall, agricultural advice) through ICTs (e.g., mobile devices, SMS, radio, social media) can be crucial to respond and mitigate the impact of emergencies such as floods and drought, for identifying pest and disease prevalence, and for informing livelihood options, key in adaptation pathways of vulnerable communities (Devkota and Phuyal, 2018; Panda et al., 2019).

In addition to contributing to the robustness and stability of the critical infrastructure in the event of disasters, ICTs can strengthen other attributes of resilient urban systems by enabling learning and community self-organisation, cross-scale networks and flexibility, helping vulnerable stakeholders, in particular, to adjust to change and uncertainty (Heeks and Ospina, 2015; Heeks and Ospina, 2019). Big data is being used to inform responses to humanitarian emergencies (Pham et al., 2014; Ali et al., 2016), as well as to generate new forms of citizen engagement and reporting (e.g., community-based maps of flood-prone areas) that can help to inform coping and adaptive responses (Ogie et al., 2019).

The selection and use of ICTs for adaptation needs to be fairly grounded in the broader socio-cultural, economic, political and institutional context, to ensure that these tools effectively help address existing, emerging and future adaptive needs. Typically, ICT is inadequate on its own to make a significant difference (Toya and Skidmore, 2015). The role of ICTs in adaptive pathways is influenced by the availability of locally relevant information (e.g., weather-based advisory messages, local market prices), the accessibility of information by all members of the community (e.g., using various text, audio and visual content, local languages, addressing gender-related exclusion, cost and digital competencies) and the applicability of information at the appropriate scale (local, regional or national), including data quality and verification (Namukombo, 2016; Haworth et al., 2018).

Information privacy and security, as well as the unintended impacts of ICTs on inequality, spread of misinformation and on widening existing gaps (e.g., due to poverty, gender and power differentials), can also constrain the contribution of ICTs to urban adaptation (Haworth et al., 2018; Coletta and Kitchin, 2017; Leszczynski, 2016) and are among the key challenges that need to be addressed in order to fully realise their potential.

A number of measures are available to adapt existing energy infrastructure to climate change. These typically involve changing engineering design codes and upgrading facilities to cope with new climatic conditions, building redundancy and robustness into systems, and preparation to ensure continued operation following extreme events. Adapting low carbon energy infrastructure improves its climate resilience whilst simultaneously delivering mitigation goals (Kemp, 2017; Feldpausch-Parker et al., 2018), benefitting all other sectors (Dawson et al., 2018; Pescaroli and Alexander, 2018; Kong, Simonovic and Zhang, 2019).

Hall et al. (2019) identified 4223 GW of global power generation at risk of flooding. If these assets were protected by 0.5 m flood protection, ~700 GW would be at risk from the 1-in-100 year flood. Many assets can be strengthened, relocated or replaced with new equipment built to higher standards. An example of this is in the UK where a total of £172 million is being invested between 2011 and 2023 to raise flood protection of substations to be resilient to the 1-in-1000 year flood (ENA, 2015). Electricity cables can be upgraded in anticipation of reduced efficiency in a warmer climate, although in many locations this may be achieved autonomously to meet growth in electricity demand (Fu et al., 2017).

Fuels, including oil, natural gas, hydrogen, biomass and CO2 prior to sequestration are delivered and distributed by pipeline or transportation by road, rail and shipping. In addition to engineering improvements, adaptation measures also include planning and preparation for service disruption by changing transport patterns, increasing local storage capacities and identifying and prioritising protection of critical transport nodes (Wang et al., 2019b; Panahi, Ng and Pang, 2020).

Several options are available to reduce the impacts of reduced cooling water for thermoelectric power generation, increases in water temperature and lower flows for hydropower generation. These include (i) switching from freshwater to seawater (if available) or air cooling; (ii) replacing once-through cooling systems with recirculation systems; (iii) replacing fuel sources for thermoelectric power generation; (iv) increasing the efficiency of hydro and thermoelectric power plants; (v) relaxing discharge temperature rules to allow warmer water to enter rivers; (vi) installation of screens to stop algae or jellyfish blooms clogging intakes; (vii) reducing power production and managing demand; and (viii) changing reservoir operation rules (where available). Shreshta et al. (2021) show that changing reservoir operation rules can offset reduced water availability under RCP8.5 until 2050, but is insufficient by the 2080s. van Vliet et al., (2016) showed that a 10% increase in hydroelectric generation efficiency can compensate for reduced water availability in most regions. Higher efficiency thermoelectric plans offset impacts under lower climate change scenarios but are shown to be inadequate under RCP8.5 by the 2080s; whereas a switch to seawater and dry (air) cooling provides a net increase under this scenario. However, these technologies can increase costs. Increasing the temperature of water discharged from the power station can have negative environmental impacts (Thome et al., 2016; Yang et al., 2015).

Longer term systemic strategies could include a combination of increased network redundancy and decentralisation of generation locations (Fu et al., 2017), or the use of ‘defensive islanding’ which involves splitting the network into stable islands to isolate components susceptible to failure and subsequently trigger a cascading event (Panteli et al., 2016). Smart grids are being increasingly deployed within municipalities to provide more efficient management of supply and demand and mitigate greenhouse gas emissions, however, there is limited understanding of their performance and reliability during floods and other extreme weather events (Vasenev, Montoya and Ceccarelli, 2016; Feldpausch-Parker et al., 2018).

Adaptation and preparedness at the household level can minimise impacts during power outages, but neighbourhood-level assistance may be more appropriate to ensure support for vulnerable households and coordination of action and information (Ghanem, Mander and Gough, 2016). More generally, it is important for responder organisations to integrate energy needs in disaster preparedness and response plans. Whilst over the longer term, reducing household and industrial demand for energy supply will reduce the need for capital investments and upgrades (Fu et al., 2017).

Providing a reliable and resilient power supply is crucial to economic and social development (Fankhauser and Stern, 2016). Furthermore, there are co-benefits from the use of low carbon energy systems (Chapter 8, WGIII AR6). For example, solar-charged street lamps and household lighting gives reliable nighttime lighting, providing safety, security and resilience to disruption of network power supplies (Burgess et al., 2017). At larger scales, deploying solar power on building roofs reduces energy demand for cooling by 12% and lowers the urban heat island, and thereby has health benefits (Masson et al., 2014a). In the USA, construction of solar panels over 200 million parking spaces would generate a quarter of the country’s electricity supply (Erickson and Jennings, 2017).

As presented in Table 6.3, access to energy supply varies considerably. In particular, many African countries require substantial energy infrastructure to support their economic development. The combination of smart technologies with solar and other renewable generation provides a huge opportunity (Anderson et al., 2017; Kolokotsa, 2017). However, care must be taken in rapidly developing cities, as failure to ensure energy access during urbanisation can reduce resilience (Ürge-Vorsatz et al., 2018).

A wide range of adaptation options are available for transport infrastructure and most provide a good benefit cost ratio (Doll, Klug and Enei, 2014; Forzieri et al., 2018). Options include upgrading infrastructure (which can often be achieved autonomously as part of standard repair and replacement schedules) and strengthening or relocating (critical) assets. Adaptation of road and rail networks in Australasia includes re-routing, coastal protection, improved drainage and upgrading of rails (Table 11.7.) In areas with substantial infrastructure deficits, such as much of Africa, investments in public transport and transit-oriented development are highlighted as desired mitigation-adaptation interventions within cities of South Africa, Ethiopia and Burkina Faso (Section 9.8.5.3). Adapting low carbon transport infrastructure will be crucial to ensure resilience to climate change impacts whilst simultaneously delivering mitigation goals (Shaheen, Martin and Hoffman-Stapleton, 2019; Costa et al., 2018).

Wright et al. (2012) calculated that strengthening bridges in the USA would cost USD 140–250 billion by 2090 (or several billion dollars a year), but costs are reduced by 30% if interventions are made proactively. Koks et al. (2019) calculate a benefit–cost ratio of greater than one for over 60% of the world’s roads exposed to flooding. The greatest benefits from adaptation of the global road network are in LMICs where reductions in flood risk are typically between 40% and 80%. Pregnolato et al. (2017) showed that in the city of Newcastle upon Tyne (UK), two carefully targeted interventions at key locations to manage surface water flooding reduced the impacts of the 1-in-50 year event in 2050 by 32%. In permafrost regions, geo-reinforcement, foundation and piles can be strengthened (Trofimenko, Evgenev and Shashina, 2017), whilst passive cooling methods, including high-albedo surfacing, sun-sheds and heat drains can cool infrastructure (Doré, Niu and Brooks, 2016).

Hanson and Nicholls (2020) calculate the total global investment costs for port adaptation to sea level rise and provision of new areas at USD 223–768 billion by 2050. However, adaptation of existing ports is only 6% of this. Yesudian and Dawson (2021) estimate the cost of maintaining present levels of flood risk in 2100 for the global air network will cost up to USD 57 billion (Monioudi et al., 2018; Esteban et al., 2020b).

New technologies and design innovations can improve the resilience of cars, trains, boats and other vehicles to cope with more extreme weather. Mobility transitions have the potential to improve mobility and accessibility, to influence urban form and to reduce vehicular use (and thereby infrastructure degradation), vehicle miles travelled and vehicle-based emissions (Sperling, Pike and Chase, 2018). For example, use of electric vehicles, hydrogen vehicles and greater uptake of public transport and other vehicles that reduce exhaust head emissions reduces the urban heat island (Kolbe, 2019). Carsharing can reduce carbon emissions by over 50% (Shaheen, Martin and Hoffman-Stapleton, 2019). Ride hailing, matching non-professional drivers of private vehicles with paying passengers, positively impacts low-income, low-car ownership households in Los Angeles (Brown, 2018), and fills market gaps in cities where public transit infrastructure is inadequate, unreliable or unsafe (Suatmadi, Creutzig and Otto, 2019; Vanderschuren and Baufeldt, 2018), but can also create a precarious and insecure job market that impacts well-being (Fleming, 2017). Whether the resulting impacts are positive or negative, largely depends on local, national and international policy and practices.

Safe and convenient walking and cycling (and public transport) infrastructure in cities reduces carbon emissions and urban heat island intensity, but also improve cardiovascular capacity which reduces heat stress (Schuster et al., 2017). In some regions, warmer weather may bring opportunities for increased uptake of cycling and walking, though precipitation or thermal discomfort caused by high temperature and humidity can reduce the use of active travel modes for commuting and recreation (Chapman, 2015). Shaded pavements and lanes, and measures to mitigate the urban heat island can reduce risks to disruption of active travel thereby also enhancing mitigation (Wong et al., 2017).

Full system re-design may enable the greatest resilience but it does not usually have a good benefit–cost ratio (Doll, Klug and Enei, 2014). Moreover, Caparros-Midwood et al. (2019) show that transport infrastructure planners will not always be able to resolve trade-offs between managing climate risks and mitigating greenhouse gases without tackling other sectors. However, infrastructure planners should continually seek opportunities for positive infrastructure lock in where available (Ürge-Vorsatz et al., 2018).

Adaptation to water scarcity can be through measures to increase supply (e.g., water storage, rainwater harvesting, desalination, river basin transfers, increased abstraction, reduced pollution of water sources), or manage demand (e.g., reduce leakage lower consumption, use of water efficiency devices, greywater reuse, behaviour change). A combination of these measures is usually required (e.g., Ives, Simpson and Hall, 2018; Dirwai et al., 2021; Wang et al., 2018a). Reliable and well-adapted water and sanitation services support economic growth, public health, reduce marginalisation and poverty, and can lower energy use and improve water quality (Campos and Darch, 2015; Miller and Hutchins, 2017; Jeppesen et al., 2015; Hamiche, Stambouli and Flazi, 2016).

Globally, water sector adaptation costs are estimated to be USD 20 billion yr -1 by 2050 (Fletcher, Lickley and Strzepek, 2019). Globally, the budget required by 2030 for water infrastructure (new and refurbishment) is more than half of the budget required for all infrastructure (Koop and van Leeuwen, 2017). For OECD countries, water adaptation increases costs by 2%, but this proportion is far higher for developing nations (Olmstead, 2014).

A number of adaptation actions are available to reduce the impacts of floods on water and sanitation infrastructure. Active management reduces blockages in water infrastructure and protects related services such as roads and culverts which are essential to ensure the operation of onsite sanitation infrastructure (Capone et al., 2020). The impact of floods for onsite or sewerage systems can be lowered by reducing or eliminating excreta from the environment through regular maintenance, cleaning and clearing of blockages (O’Donnell and Thorne, 2020; Borges Pedro et al., 2020).

Infrastructure to protect key assets such as water and wastewater treatment plants or pumping stations has a high cost but benefits all connected households and reduces pollution from flood events. In well-regulated water sectors, there has been an increasing focus on such investments (Campos and Darch, 2015). Whereas more diffused cheaper interventions can reduce flood water ingress to domestic toilets (Irwin et al., 2018). Luh et al. (2017) found that protected dug wells were one of the least resilient technologies, whereas piped, treated, utility managed surface water systems had higher resilience.

Protecting water sources from pollution is even more important in a warmer climate that increases the frequency of algal blooms. Individual assets such as water intake pipes can be protected using screens (Kim et al., 2020a), whereas basin-scale land management is required to reduce nutrient load from runoff (Me et al., 2018), whilst injecting water or installing barriers can protect coastal aquifers from salinisation (Siegel, 2020).

More radical structural interventions may be needed in the longer term, but would need to be planned and delivered in coordination with investments in other sectors, particularly housing (Lüthi, Willetts and Hoffmann, 2020). As an interim measure, sanitation services with a lower reliance on fixed infrastructure, or container-based sanitation could be appropriate in many urban areas that are badly affected by flooding (Mills et al., 2020).

Other actions include use of adaptive planning (Evans, Rowell and Semazzi, 2020), integration of measures of climate resilience into water safety plans (Prats et al., 2017), as well as improved accounting and management of water resources (Lasage et al., 2015). Policy prescriptions on technologies for service delivery and changes in management models offer potential to reduce risks, particularly in low-income settings (Howard et al., 2016). Where formal sewerage provision is lacking, community based adaptation that incorporates both the function of the sanitation system and the vulnerability of users (e.g., women, children, elderly, ill or disabled) into the design is essential (Duncker, 2019).

Cities are deploying a broad range of strategies to adapt infrastructure to flooding, with hard engineering approaches (e.g., dikes and seawalls) increasingly complementing soft approaches, including planning and use of nature-based solutions, that emphasise natural and social capital (Jongman, 2018; Sovacool, 2011). The infrastructure can alter downstream risks and lead to increased residual risk by encouraging more floodplain construction (Miller, Gabe and Sklarz, 2019; Ludy and Kondolf, 2012). Physical infrastructure is highly cost effective for large settlements, but not always for small settlements (Tiggeloven et al., 2020) and can be inaccessible to poorer communities (Sayers, Penning-Rowsell and Horritt, 2018; Van Bavel, Curtis and Soens, 2018). It is often inflexible once installed but new designs and adaptive pathways are emerging (Anvarifar et al., 2016; Kapetas and Fenner, 2020).

As urban areas have expanded, so too have the number of vulnerable assets, and efforts may now emphasise reducing construction in high-risk regions (Paprotny et al., 2018a). The National Flood and Coastal Erosion Risk Management Strategy for England, for example, calls for reductions in inappropriate developments in floodplains (Kuklicke and Demeritt, 2016; UK Environment Agency, 2020). Because climate change increases the flood risk profile of certain regions, reconsideration of design criteria has become more common (Ayyub, 2018). New York City now requires the sewer system currently designed for hydraulic capacity in 5-year design life should be designed for 50-year design life, taking into account climate changes over that period (NYC, 2019).

Adaptation strategies are diverse and often involve hybrid physical and NBS, and increasingly integrated management plans that consider both flood prevention and designing infrastructure and supporting people to cope with floods when they occur. Adaptation typically focuses on (i) increasing the standard of protection to compensate for the increased magnitude of extreme events; (ii) increased maintenance to cope with increased frequency of extremes and changes in ambient conditions; (iii) changed maintenance regimes from narrower maintenance windows, for example as assets are used more frequently (Sayers, Walsh and Dawson, 2015); (iv) land use planning and management to reduce exposure and manage hydrological flows; and (v) raising awareness, preparedness and incident management. In high population areas, hard interventions such as dikes and levees are generally cost effective (Jongman, 2018; Ward et al., 2017).

Prevention or attenuation solutions include: rooftop detention, reservoirs, bioretention, permeable paving, infiltration techniques, open drainage, floating structures, wet-proofing, raised structures, coastal defences, barriers and levees, and have been deployed in diverse configurations and environments around the world (Matos Silva and Costa, 2016). Barcelona (Spain) reached 90% impermeable surface cover by the 1980s, and has recently begun implementing artificial detention, underground reservoirs and permeable pavement technologies (Favaro and Chelleri, 2018; Matos Silva and Costa, 2016). Florida Power and Light (USA), which provides service to approximately 10 million people, is investing USD 3 billion in flood protection and the hardening of assets (for example, upgrading wooden poles to steel and concrete) (Brody, Rogers and Siccardo, 2019). The City of Seattle recommends increasing preventative maintenance activities, the regular review of appropriate pavement technologies and modifications to subgrades and drainage facilities for high-risk areas (City of Seattle, 2017), whilst also providing benefits to transport disruption (Arrighi et al., 2019). Adaptation in African cities is often dominated by informal responses (Owusu-Daaku and Diko, 2018). In the absence of centralised responses, low-income residents in Nairobi (Kenya) dig trenches and construct temporary dikes to protect homes, and in Accra (Ghana) the community has developed a range of social responses, including communal drains and local evacuation teams, to help protect people and critical valuables, although these innovations require connection to city-wide infrastructure to effectively reduce widespread risk (Amoako, 2018).

More recent developments include sensor arrays to catalogue a river’s reach and how changing hydraulics interact with roadways (Forbes et al., 2019). Kuala Lumpur’s (Malaysia) stormwater management and road tunnel (SMART) during extreme rain events transitions the motorway to a stormwater conduit, an example of multifunctionality enabling agility (Isah, 2016; Markolf et al., 2019). Smart stormwater control systems are starting to use real-time control to dynamically manage the retention and movement of water during storms, though uptake at large scales which provide the greatest improvements in performance have been limited (Xu et al., 2020b).

In contrast to a ‘fail-safe’ approach to design which emphasises strengthening infrastructure against more intense environmental conditions, ‘safe-to-fail’ flood strategies allow infrastructure to fail in its ability to carry out its primary function but control the consequences of the failure. Examples include the use of a bioretention basin in Scottsdale (Arizona, USA) to accommodate excess runoff and help drain the city; a subsidy for affected farmers for lost crop production as part of the Netherlands’ Room for the River programme; targeted destruction of a levee to control flooding in the Mississippi River Valley in 2011 (Kim et al., 2019). Water-sensitive urban design, low-impact development, sponge cities, sustainable urban drainage and natural flood management involve deployment of systems and practices that use or mimic natural processes that result in the infiltration, evapotranspiration or use of stormwater to protect water quality and associated aquatic habitat. These are being designed and implemented at increasingly ambitious scales. For example, China’s Sponge City initiative sets a goal of 80% of urban land able to absorb or reuse 70% of stormwater through underground storage tanks and tunnels, and use of pervious pavements, in addition to NBS (Chan et al., 2018; Muggah, 2019). Similarly, several thousand water-sensitive urban design interventions have been implemented across the city of Melbourne (Kuller et al., 2018).

Physical coastal management infrastructure has significant benefits in reducing flood and erosion losses and damage from storms. Physical infrastructure includes seawalls, dikes, breakwaters, revetments, groynes and tidal barriers. Adapted infrastructure can alter risks in morphologically connected areas, and lead to increased residual risk by encouraging more construction in the coastal zone (Miller, Gabe and Sklarz, 2019; Ludy and Kondolf, 2012). The infrastructure is highly cost effective for large settlements, but not always for small settlements (Tiggeloven et al., 2020) and can be inaccessible to poorer communities (Fletcher et al., 2016; Pelling and Garschagen, 2019).

Anticipated costs for this vary widely. For example, Hinkel et al. (2014) calculate that adaptation costs to maintain current global levels of coastal flood protection would be 1.2–9.3% of gross world product but protect assets in human settlements of USD 21–210 billion; Tiggeloven et al. (2020) calculate the cost of adaptation to be USD 176 billion (although this would provide a benefit–cost ratio of 106 under RCP8.5); while Nicholls et al. (2019) estimate that global coastal protection would cost substantially more, up to USD 18.3 trillion between 2015 and 2100 for RCP8.5 (this includes ranges of unit costs and maintenance costs which have often been ignored).

Coastal protection infrastructure such as dikes and sluice gates can inhibit salinity intrusion through careful management of water levels, this can provide co-benefits for flood risk reduction and agricultural productivity, but can also have negative impacts on ecosystems (Renaud et al., 2015). Managed aquifer recharge can be effective if the objective is to secure freshwater drinking supply (Hossain, Ludwig and Leemans, 2018).

Physical infrastructure can provide substantial benefits, be constructed quickly and has enabled coastal cities and settlements around the world to flourish and grow. Multifunctional physical infrastructure can also provide economic and social co-benefits. These include integration of transport, recreation, agriculture (e.g., cattle pasture), founding for wind turbines, housing, office or industry into the coastal management infrastructure (Anvarifar et al., 2017; Kothuis and Kok, 2017). However, physical infrastructures can also disrupt natural processes, often leading to undesirable impacts such as pollution, degradation of ecosystems and displacement of erosion and flood risk to other locations (Wang et al., 2018b; Dawson, 2015; Nicholls, Dawson and Day, 2015). Coastal management strategies that take a hybrid approach, integrating physical and natural infrastructure, provide the best opportunities for managing risk and achieving wider socioeconomic and environmental benefits (Depietri and McPhearson, 2017; Morris et al., 2018; Schoonees et al., 2019; Powell et al., 2019).

Questions of equity and justice influence adaptation pathways for cities, settlements and infrastructure (see also Chapter 8). Although infrastructure, ranging from social to ecological and physical to digital, can help to reduce the impacts of climate change (Stewart and Deng, 2014; Baró Porras et al., 2021), there is limited evidence of how infrastructures, implemented to reduce climate risk, also reduce inequality. Rather, there is more evidence to suggest that both adaptation plans and associated infrastructure implementation pathways are increasing inequality in cities and settlements (Chu, Anguelovski and Carmin, 2016; Anguelovski et al., 2016; Romero-Lankao and Gnatz, 2019). Social, economic and cultural structures that marginalise people by race, class, ethnicity and gender all contribute in complex ways to climate injustices and need to be urgently addressed for adaptation options to shift to benefit those most vulnerable, rather than mainly benefitting the already privileged and maintaining the status quo (Thomas et al., 2019; Porter et al., 2020; Ranganathan and Bratman, 2019). Innovation and imagination are needed in adaptation responses to ensure that cities and settlements shift from perpetuating structural domination and inequality to fairer cities (Porter et al., 2020; Henrique and Tschakert, 2019; Parnell, 2016b). To support these possibilities, this section explores adaptation through the lens of distributive and procedural justice. Although not expanded on here, spatial and recognition injustices are equally important (Fisher, 2015; Chu and Michael, 2018; Campello Torres et al., 2020). Recognition can be supported through a capabilities approach that helps to bring attention to past cultural domination and enable citizens to develop the functioning life they choose (Schlosberg, Collins and Niemeyer, 2017). This brings a focus on local action, emphasising the relevance to vulnerability reduction and resilience building of individual and local/community capacities and supporting structures. This blurs the distinction between climate change adaptation and community development, with the former firmly embedded in the latter. Struggles for recognition are deeply political and central to adaptation responses which requires increased focus on power to support more equitable and just adaptation (Nightingale, 2017). Justice questions are not static, Box 6.4 overviews the implications of COVID-19 for urban justice and vulnerability.

Distributive justice calls attention to unequal access to urban services, land, capital and technology. Related to this, exposure to health, flooding and drought risks of people living in low-income and informal settlements is a growing concern, as is disaster preparedness and the ability to support the needs of vulnerable groups such as the elderly, children and disabled, where data is often lacking (Lilford et al., 2016; Castro et al., 2017). There are also differences in who benefits from infrastructures, as they are inherently political, embedded in social contexts, politics and cultural norms (McFarlane and Silver, 2017) and often tend to benefit those already privileged (Henrique and Tschakert, 2019). As an example, fixing water leaks can depend as much on the politics of who is involved and whose knowledge is prioritised as on the technical aspects (Anand, 2015).

The quality and maintenance of infrastructure is often unequal across cities, benefiting some and increasing vulnerability of others. Some property is seen as dangerous and of lower value if highly exposed to risk (Wamsley et al., 2015). Similarly, areas suffering from disinvestment in infrastructure might have a high risk of flooding (Haddock and Edwards, 2013). Zoning and land use trade-offs have been seen to be unequally skewed in favour of prime real estate and economically valuable assets (e.g., protecting factories and refineries from flooding) (Anguelovski et al., 2016; Carter et al., 2015). Urban planning reforms are therefore central to building a fairer urban adaptation response (Parnell, 2016b).

Infrastructure is often not adequately implemented in low-income urban areas and not equally accessible to all (Meller et al., 2017). For example, low-income neighbourhoods often have less green space and therefore less associated cooling benefits. Even in high-income areas, there is often unequal access to services. For example, an assessment of sustainable urban mobility plans in Portugal showed that some areas have considered equity in their plans and increased access for disadvantaged users including the elderly and disabled, but in other cities this is lacking (Arsenio, Martens and Di Ciommo, 2016). Understanding who has access to what infrastructure can help to redress the drivers of social vulnerability that are central to just urban adaptation (Michael, Deshpande and Ziervogel, 2018; Shi et al., 2016).

Changing land use and increasing green spaces to reduce climate risks and attract investments and job opportunities has increased real estate values, triggered climate gentrification in some areas (Keenan, Hill and Gumber, 2018) and decreased access to affordable housing in other areas (Larsen, 2015; Carter et al., 2015). Displacement through evictions and relocations linked to land use conversion and resettlement in the name of adaptation has also increased people’s vulnerability (Anguelovski et al., 2016; Henrique and Tschakert, 2019).

Understanding social and economic elites and their investment in infrastructure has implications for distributive justice, particularly when there is secession from public infrastructure services that has financial implications for viability (Romero-Lankao, Gnatz and Sperling, 2016). In the case of the 2015–17 Cape Town drought, wealthy households secured their water needs through off-grid technologies such as rainwater tanks and boreholes. Although this resulted in more water being available in the dams, it also led to less revenue being collected for municipal water and less ability to cross-subsidise water for poor households (Ziervogel, 2019b; Simpson, 2019; Bigger and Millington, 2019). More attention needs to be paid to how shifts in infrastructure are serving the interests of urban elites, often driven by the state, and failing to adequately consider the needs of the disadvantaged (Bulkeley, Castán Broto and Edwards, 2014; Ajibade, 2017; Shi et al., 2016). Equally, more risk-reducing infrastructure is needed across all urban areas (Reckien et al., 2018a).

Procedural justice, which focuses on the institutional processes by which adaptation decisions are made, brings attention to the lack of opportunity for engaging in political decision making and limited representation of diverse voices in cities and settlements, and in relation to investment in infrastructure (Coates and Nygren, 2020; Henrique and Tschakert, 2019). Even when inclusive adaptation processes are run, they seldom produce procedurally just outcomes (Malloy and Ashcraft, 2020). Understanding who is excluded and included is important (Sara, Pfeffer and Baud, 2017). One example is the increasing numbers of migrants who are confronted with lack of access to citizenship rights and housing tenure (Romero-Lankao and Norton, 2018). Often, migrants are not allowed to formally claim public provisions in health, finance and shelter (Chu and Michael, 2018). Further, migrants and their settlements are likely unrecognised in spatial or infrastructure development plans. In this context, social infrastructure, zoning and land use planning for climate adaptation has triggered inequity through omission, as some planning process have been racialised and excluded groups such as migrants and ethnic minorities (Anguelovski et al., 2016). Urban adaptation policy-making processes that explicitly integrate multiple stakeholder interests can help to balance top-down solutions (Reckien et al., 2018a).

Identifying who is least able to adapt to climate risks sufficiently is important (Thomas et al., 2019). Some people may have few opportunities to relocate away from flooded areas in the long term or to evacuate in the short term. It is also harder for many from low-income areas to rebuild after an extreme event. Lack of housing tenure and sub-standard housing has been shown to limit the ability of residents to improve and manage their landscapes and therefore it is hard for them to enhance energy efficiency (Dempsey et al., 2011). Access to information is critical for adapting to climate risk and reducing vulnerability to hazards, yet access to this information is often not equally available (Ma et al., 2014). For example, low literacy can hamper ability to respond to early warning information (Dugan et al., 2011). In other instances, racial violence has surfaced during disasters, with Black victims’ lives being seen as less important than others (Anderson et al., 2020).

When looking at justice issues in urban adaptation, it is important to recognise that the adaptation of one individual or household may lead to maladaptation and negative impacts elsewhere (Holland, 2017; Limthongsakul, Nitivattananon and Arifwidodo, 2017; Atteridge and Remling, 2018). For example, the case of an area of peri-urban Bangkok experiencing localised flooding due to unregulated private sector development saw households take both individual (building flood walls around homes, digging temporary drainage swales in the carriageway) and collective action (petitioning authorities, pumping water into vacant land). These actions, to a certain extent, merely displaced the flood water to other areas, or created new problems by damaging the carriageway, creating negative impacts on other households and the wider community. However, ultimately, it was the actions of improperly regulated private sector developers driving the need for this autonomous adaptation (Limthongsakul, Nitivattananon and Arifwidodo, 2017).

One of the tensions that emerge when addressing injustice is that the global provision of modern infrastructure is increasingly seen as unfeasible. It is unfeasible, both in terms of the current high emissions associated with infrastructure (World Bank, 2017 ) and the centralised, high standard ideal (Lawhon, Nilsson and Silver, 2018; Coutard and Rutherford, 2015). Decentralisation is increasingly needed, which the urban poor already engage in through their use of ‘informal’ infrastructure technologies, given their limited access to infrastructure networks. Transformative adaptation pathways that reduce climate risk whilst reducing inequity require an approach that sees infrastructure as inherently social and political.

As analytical concepts, mitigation and adaptation have helped, over the years, to structure thinking and action around climate change. However, since AR5 there has been a growing debate about the adequacy of a neat separation between adaptation and mitigation (Castán Broto, 2017).

The delivery of climate change action has revealed numerous co-benefits between adaptation and mitigation, around diverse areas such as implementing NBS and delivering health and development benefits (Ürge-Vorsatz et al., 2014; Suckall, Stringer and Tompkins, 2015; Puppim de Oliveira and Doll, 2016; Spencer et al., 2017). There has been a strong interest in delivering development benefits alongside climate mitigation, thus benefiting the overall infrastructure base (Suckall, Stringer and Tompkins, 2015). Some of these co-benefits have also emerged in experiences of urban planning, pointing toward the dilemma of separating adaptation and mitigation in a context in which integration, rather than an analytical differentiation, was seen as being required to transcend work in silos (Aylett, 2015). Because urban planning needs to carefully consider long time scales, the neat separation between mitigation and adaptation runs counter to integrated forms of planning that can consider scales (time and space) carefully and that are aimed to deliver the sustainable city as a whole (Solecki et al., 2015; Grafakos et al., 2020).

For example, the ideas of climate resilient development and climate compatible development help planners to consider the simultaneous wins that emerge between adaptation, mitigation and development, requiring institutional building and partnerships to deliver triple win solutions (Stringer et al., 2014; Seo, Jaber and Srinivasan, 2017; Mitchell and Maxwell, 2010). While the evidence base for the actual possibility of achieving such triple wins remains scarce (Tompkins et al., 2013; Sharifi, 2020), emerging examples show important developments. For example, establishing safe and convenient walking and cycling infrastructure can lead to improvements in population health, thereby highlighting the close interaction between urban land use, infrastructure and population health (Schuster et al., 2017), while clean cooking has the potential to deliver positive health outcomes alongside improvements in air quality and emissions reductions and through reducing pressure on woodland as a fuel source for expanding urban populations (Msoffe, 2017). Furthermore, active transport infrastructure reduces air pollution and related health risks, and helps to mitigate further climate change (Schuster et al., 2017). These are supported by city networks such as the C40 Clean Air Cities Declaration and the Clean Air Coalition that complements WHO guidelines and standards, for example through the Breathe Life Campaign. In conclusion, in both urban environments and infrastructural sectors, triple wins are only realisable through broader perspectives that link climate compatible development to institutional change or the achievements of wider welfare objectives such as those enshrined in the United Nations 2030 Agenda of Development (Castán Broto et al., 2015; England et al., 2018) (medium evidence, high agreement ).

The aspiration to deliver climate change action within a broader agenda of transformative change, introduced in the SREX report, received renewed attention after the publication of IPCC Special Report on Global Warming of 1.5°C, which argues for a focus on urban transformations and highlighted that informal settlements were vital for understanding the delivery of these transformations. Deep decarbonisation has emerged as a new idea that regards the development of low or zero carbon pathways as a condition for good adaptation in the long term. Decarbonisation becomes urgent in the face of growing impacts attributable to climate change (Ribera et al., 2015; Bataille et al., 2016; Wesseling et al., 2017). Urbanisation opens opportunities for deep mitigation in low-impact developments, and hence, it is imperative to understand the implications of those opportunities for climate action (Mulugetta and Broto, 2018). These gains are not limited to urban areas. The reliance on connected urban–rural systems for water, food and fuel has led to city government and urban-based businesses supporting landscape adaptations in rural hinterlands with strong potential for mitigation and rural development co-benefits. Water Funds bring downstream urban public and private finance to support upstream rural residents to make land use and agricultural management decisions to avoid damaging runoff, soil erosion and downstream sedimentation with reduction in water quality and increased flood risk. There are more than 30 Water Funds in Latin America and sub-Saharan Africa. These operate at landscape scale; the Upper Tana-Nairobi Water Fund, Kenya (Vogl et al., 2017), planned for a USD 10 million investment in Water Fund-led conservation interventions, with a projected return of USD 21.5 million in economic benefits over a 30-year timeframe (Apse and Bryant, 2015). However, these investments do not occur where communities lack funding or the institutions to direct funding from downstream beneficiaries to upstream residents (Brauman et al., 2019).

The assessment of cases of local adaptation demonstrates that most urban adaptation is led by local governments (although the local government is also a heterogeneous category and local governance arrangements may vary across administrative and political contexts) (high confidence) (Amundsen et al., 2018; Lesnikowski et al., 2021). Local government reform at different levels can improve local adaptation, whether this is by strengthening specific teams or building cross-departmental linkages (high confidence) (Paterson et al., 2017; Shi, 2019; Wamsler and Raggers, 2018). Adaptation success often depends on having political champions driving the adaptation agenda alongside measures such as access to a knowledge base, resources at hand, political stability and the presence of dense social networks that can be supported through local government reform (Pasquini et al., 2015). Aligning adaptation objectives with other potential benefits of sustainable development also supports adaptation. Specifically, policies and plans that link adaptation to the objectives of Agenda 2030 supports action at the local level (UN-Habitat, 2016b). Showing the economic benefits of adaptation is a strategy for local institutions to gain support for adaptation action. For example, local governments in Surat, Indore and Bhubaneswar in India linked adaptation to local development needs in experiments that facilitated accessing human and finance resources, at the local, national and international levels (Chu, 2016). However, linking adaptation to co-benefits may also divide efforts and reduce the effectiveness of adaptation actions. For example, urban land use planning and management in Ambo town, Ethiopia, resulted in the implementation of urban greening projects, but these projects did not directly address the climate-related disaster risks affecting the settlement, including urban flooding, water stress, water shortages, increased urban heat, wind and dust storms (Ogato et al., 2017).

Multi-level governance measures that support local governments can foster robust adaptation approaches and address risks and vulnerabilities across scales (high confidence) (Westman, Broto and Huang, 2019; Hardoy et al., 2014; Romero-Lankao and Hardoy, 2015). Effective action by local government requires national government’s support (medium confidence). For example, Araos et al. (2017) documents the case of Dhaka, Bangladesh, where a national plan prioritises measures for protecting coasts and agricultural production. In this context, the local government has minimal access to human and financial resources. Without national support, the local government struggles to coordinate action among different stakeholders. National urban adaptation directives can influence municipal governments’ action and planning, but evidence suggests that national policy alone is not sufficient to deliver action on the ground without understanding local conditions (high confidence) (Archer et al., 2014; Lehmann et al., 2015).

There are barriers for municipal adaptation plans to deliver effective adaptation outcomes and implemented actions often diverge from plans (see Section 6.4.6). For example, a comparison of adaptation plans and budget expenditures of six metropolitan cities in South Korea between 2012 and 2016 showed that the implementation of adaptation programmes diverged substantially from the original plans, both in terms of total and sector-specific spending (Lee and Kim, 2018). Often, a focus on institutional change and reform limits attention to more practical aspects of adaptation that improve communities’ resilience (Castán Broto and Westman, 2020). Adaptation actions, even where financed effectively, do not always deliver positive outcomes (high confidence) (Reckien et al., 2015; Woodruff and Stults, 2016; Uittenbroek, 2016; Aguiar et al., 2018; Reckien et al., 2018a; Olazabal et al., 2019b; Campello Torres et al., 2021) (see also Section 6.4.7).

There are multiple actors, other than local governments, that can deliver adaptation action, including businesses, not for profit organisations and trade unions (high confidence) (Giordano et al., 2020; Eakin et al., 2021). Empirical evidence since the AR5 highlights the role of communities, universities, the private sector and transnational networks in adaptation (Hunter et al., 2020; Bäckstrand et al., 2017). Non-state actors are particularly important in enabling adaptation by linking government agencies with low-income and marginalised communities, including those living in informal settlements (Kuyper, Linnér and Schroeder, 2018; Khosla and Bhardwaj, 2019).

Since AR5, civil society and private actors have emerged as core knowledge holders and drivers of experimentation, even succeeding in changing public policy in the process (Klein, Juhola and Landauer, 2017; McKnight and Linnenluecke, 2016; Mees, 2017). Previous IPCC Assessment Reports noted that civil society actors enable local risk awareness, sensitisation and adaptive capacity, and generate locally based innovation (e.g., through community based adaptation programmes).

Community based adaptation includes a range of initiatives that put communities at the centre of planning for adaptation, often led by communities themselves (Reid, 2016). Community based adaptation is a comprehensive and effective strategy to deliver resilience at a human scale (Trogal et al., 2018; Greenwalt et al., 2020). Many community based responses to climate impacts represent coping strategies developed within households with a small effect on adaptation capacities beyond incremental improvements. Residents adopt private coping strategies to reduce exposure to and the impacts of heat, floods, flash floods, landslides, storms and diseases on their lives (Hambati and Yengoh, 2018). These coping strategies include the construction of physical protection against flooding, reforestation, the construction of terraces, flood diversion measures and interventions to protect houses (such as raised doorsteps or use of sandbags and adoption of building techniques for making homes resilient to storms and landslides), ventilation of houses, urban agriculture and redefinition of daily practices and livelihoods (Navarro et al., 2020; Malabayabas and Baconguis, 2017; Apreda, 2016; de Andrade and Szlafsztein, 2020; Sahay, 2018; Bausch, Eakin and Lerner, 2018).

Individual coping strategies are generally ineffective in reducing extreme risks and they rarely address the underlying structural causes of vulnerability (high confidence) (Sahay, 2018; Rözer et al., 2016; Jay et al., 2021). Expending resources on private coping strategies in some cases may divert resources and capacity for wider community adaptation efforts (de Andrade and Szlafsztein, 2020). However, individual coping strategies can provide foundations for the implementation of collaborative action in communities, building on people’s experiences, in ways which may have a longer-term, durable impact on developing resilience (high confidence) (McEwen et al., 2018). Community based adaptation can be effective at different scales, whether this is to manage transboundary issues (Limthongsakul, Nitivattananon and Arifwidodo, 2017), support the replication of local solutions (Danière et al., 2016), increase the uptake of adaptation measures (Liang et al., 2017) or inform the design of more effective policies for resilience (Berquist, Daniere and Drummond, 2015; Odemerho, 2015). Community action may be mediated by NGOs or third sector organisations who play a coordinating or enabling role, particularly where other local government mechanisms are absent.

There is weak evidence of private sector involvement in urban adaptation (Pauw, 2015; Heurkens, 2016). The absence of private sector investment in adaptation is particularly visible in rapidly urbanising countries (Nagendra et al., 2018). Business continuity describing private sector preparedness notes that firms underestimate the impacts of climate risks on their business models (Goldstein et al., 2019; Forino and von Meding, 2021; Korber and McNaughton, 2017; Crick et al., 2018b). There is little research on how businesses can play a leading role in urban adaptation (Klein et al., 2018). A global assessment of the private sector’s role in urban adaptation using data from 402 cities shows that most adaptation projects focus on the public sector and do not address private sector concerns or local people’s participation (Klein et al., 2018). Recorded private sector action is recognised through partnerships and participation (Peterson and Hughes, 2017; Hughes and Peterson, 2018). There are a few examples of studies of private sector-led adaptation action which adopts a national focus (Crick et al., 2018a; Crick et al., 2018b). This lack of evidence contrasts with a well-developed body of literature on private sector-led mitigation (Averchenkova et al., 2016).

Businesses have an essential role in urban adaptation actions, through the collective formulation of adaptation strategies, the provision of critical adaptive interventions and collaboration in partnerships. Businesses in the property sector, such as real estate developers, are on the frontline of climate change impacts but display differing attitudes toward climate adaptation. A study of property businesses in cities in Australia (Taylor et al., 2012) showed that speeding up planning approval processes facilitated adaptation actions, and joint private–public decision-making was the preferred mode of governance for responding to climate concerns. Property businesses in cities in Sweden had a limited and reactive engagement in climate issues and resisted regulation (Storbjörk et al., 2018). Corporate, private sector interventions in urban risk reduction more broadly remain limited, with a mix of public and private responsibility for planning, implementing and maintaining adaptations in the built environment, and yet limited engagement of private sector actors in providing healthcare measures for heat prevention (medium confidence) (Mees, 2017).

There is little published literature documenting the heterogeneity of business and the private sector’s responses to climate impacts (Linnenluecke, Birt and Griffiths, 2015; Doh, Tashman and Benischke, 2019). Firms have varying abilities to introduce climate adaptation measures related to staff availability, levels of awareness, perceptions of responsibility and duration of contracts (short-term projects implies less interest in adaptation outcomes) (Shearer et al., 2016). The impact of COVID-19 has serious but uncertain implications for both access to finances for sustainable development by LMICs and sub-national governments, and the possibility of stimulating maladaptive infrastructure and policy responses (OECD, 2020; Sovacool, Del Rio and Griffiths, 2020). The response of businesses to disasters influences the resilience in the communities in which they operate (McKnight and Linnenluecke, 2016; Linnenluecke and McKnight, 2017). However, at the same time there is a growing literature that warns against the conflict interests that businesses may have in their adaptation strategies. For example, real estate responses to flooding have led to processes of climate gentrification, whereby lower income populations are displaced toward higher risk areas which stablishes racialised and class-based patterns of inequality of exposure to risk, with hard evidence rapidly growing specially in US cities (Keenan, Hill and Gumber, 2018a; Shokry, Connolly and Anguelovski, 2020; De Koning and Filatova, 2020; Aune, Gesch and Smith, 2020). Private-sector participation in adaptation solutions depend on having mechanisms to enable transparency and open reporting on the nature of support and the solutions proposed. For example, businesses adopting ‘community-centric’ disaster management strategies can assist local recovery efforts by protecting employment, provision of emergency supplies and participation in reparations (McKnight and Linnenluecke, 2016). Private sector actors engaged in community climate responses can play a role in funding and managing programmes that address public health and education concerns. The potential of ecopreneurship, social enterprises, cooperatives and other sustainability-oriented business models (Schaltegger, Hansen and Lüdeke-Freund, 2016; Lopes et al., 2020; Battaglia, Gragnani and Annesi, 2020) for urban adaptation remains under-explored in the literature on urban climate governance.

The private sector also constitutes a key stakeholder group involved in collaborative processes to develop adaptation strategies. The inclusion of private sector actors in deliberative policy-making processes in urban adaptation can lead to higher procedural legitimacy levels, as witnessed in Rotterdam’s case (Mees, Driessen and Runhaar, 2014). Rotterdam has created an institutional environment that favours eco-innovation (Huang-Lachmann and Lovett, 2016). The municipal government works directly with the private sector to enhance protection against flooding constructing a marketing strategy around a ‘floating city’ concept. A ‘floating housing’ market has expanded, with benefits for the local real estate and construction industries and knowledge-exporting businesses that provide consultation expertise, delta technologies and architectural models. Nevertheless, these new trends raise new governance challenges to deliver adaptation.

There are obstacles associated with reconciling private sector interests with public priorities and justice agendas in local climate programmes. The involvement of the private sector in adaptation actions may lead to the appropriation of land and natural resources, and to the exclusion of vulnerable populations (Anguelovski et al., 2016; Rumbach, 2017; Scoppetta, 2016) (see also Section 6.4.4.2). Navigating the inclusion of businesses in urban planning processes requires local authorities to engage in ongoing negotiations, to reflect on constantly shifting power balances and to move delicately between the role of regulator and facilitator in the process of defining and maintaining long-term objectives (Storbjörk, Hjerpe and Glaas, 2019b; Storbjörk, Hjerpe and Glaas, 2019a).

Multi-level governance remains an influential paradigm that recognises government institutions’ influence at different scales and the diversification of actors intervening in public issues from the private sector and civil society (robust evidence, high agreement ). Establishing linkages between multiple organisations can help deliver coordinated action. Multi-level governance includes mechanisms for multiple actors to engage in local adaptation strategies through collaborative processes of planning, learning, experimentation, capacity building, construction of coalitions and communication channels (Barton, 2013; Jaglin, 2013; Reed et al., 2015; Restemeyer, van den Brink and Woltjer, 2017; Melica et al., 2018). Many of these studies directly focus on institutional arrangements that facilitate interaction between communities and civil society, experts, government representatives, firms and international organisations. Box 6.5 demonstrates the decisive role that community activists can play in building resilience over long periods.

Institutional fragmentation reduces the capacity to deliver adaptation (Den Uyl and Russel, 2018) Multi-level governance shows a commitment to tackling fragmented and complex policy issues through collaboration between national governments and non-state actors, as explained in the 2030 Development Agenda, especially SDG17 (‘Revitalize the global partnership for sustainable development’). Multi-level governance is particularly important to deliver adaptation at the metropolitan scale, that require coordinating actions across different institutions in inter-municipal institutions (Lundqvist, 2016). Gaps in knowledge remain regarding the effectiveness of multi-level governance actions in different contexts and the extent to which multi-level governance strategies transfer the brunt of responsibility for adaptation action to less-resourced local governments (Hale et al., 2021).

Public–private partnerships are increasingly relevant for collaborative development of urban adaptation (Klein et al., 2018). Partnerships can deliver infrastructure, coordinate policy and support learning. The main limitation of partnerships is scale, as partnership action is usually limited to discrete projects or objectives. Partnerships tend to be linked to reactive (rather than proactive) adaptation projects and the deviation of objectives away from adaptation concerns (Harman, Taylor and Lane, 2015). Partnerships can support capacity building in public and private organisations and facilitate networking efforts that extend beyond the private sector to communities and NGOs (Bauer and Steurer, 2014; Castán Broto et al., 2015). Public actors can benefit from the private sector’s innovation and implementation capacity, and businesses can de-risk investments. Still, partnerships can also strengthen the ideologies of growth and managerialism within the operations of the local government (Taylor et al., 2012). Reconciling divergent norms and routines within public and private organisations remains one of the challenges to establishing successful public–private partnerships for adaptation (Lund, 2018). Administrative and political culture influences the nature of interactions between public and private sector actors in urban adaptation agendas (Bauer and Steurer, 2014), with negative consequences such as the imposition of vertical chains of commands on horizontal collaborations, and the need to formalise contractual relations (Klein and Juhola, 2018).

Local authorities are an important enabling actor that can guide the private sector and communities to take responsibility for creating policy and regulatory environments that encourage private sector participation aligned with the SDGs’ equity and ecological sustainability principles (high confidence). For example, Frantzeskaki et al. (2014) report a port relocation project in the Netherlands where sustainability principles drove private sector participation. Klein et al. (2017) cite examples from two cities—Helsinki and Copenhagen, where local authorities have shifted adaptation responsibilities to private actors through regulation and public problem ownership. In Mombasa, private companies provide green infrastructure to match local government requirements, in what has frequently been cited as an example of NBS (Kithiia and Dowling, 2010; Kitha and Lyth, 2011).

Since the late 1990s, transnational municipal networks (TMNs) have increased awareness of climate change and served as a bridge for cities to access critical financial resources from private and philanthropic sources (Rashidi and Patt, 2018; Fünfgeld, 2015). Recently, TMNs have taken on more programmatic functions, working with cities to strategise, plan and incrementally improve their organisation functions in the face of climate change. For example, the Rockefeller Foundation’s 100 Resilient Cities program (2014–2019) provided a two-year salary for a Chief Resilience Officer (CRO) to be situated in a municipal authority to bridge silos, incentivise change and develop development strategies for resilience (Bellinson and Chu, 2019; Spaans and Waterhout, 2017). In these cases, external actors have enabled broad organisation change, resource mobilisation pathways and alternative forms of agenda-setting in cities (Chu, 2018; Hakelberg, 2014) (see also Case Study 6.2, Semarang).

A range of TMNs also support and encourage cities and settlements to plan and implement adaptation actions. ICLEI-Local Governments for Sustainability has developed protocols and implemented projects for member cities. The C40 Climate Leadership Group has facilitated the coordination of both local governments and business actors at a global scale (Gordon, 2020). Policy coordination has been central to the signatories of the Covenant of Mayors (Domorenok et al., 2020). Such networks can encourage the sharing of information about appropriate practices between urban areas; contribute to goal setting; support experimentation and development of new policy instruments; enhance stakeholder engagement; institutionalise climate agendas; and encourage policy integration across governance levels and sectors (Bellinson and Chu, 2019; Busch, Bendlin and Fenton, 2018; Fünfgeld, 2015; Busch, 2015; Papin, 2019; Rashidi and Patt, 2018). However, participation in TMNs is biased toward cities in the Global North (Bansard, Pattberg and Widerberg, 2017; Haupt and Coppola, 2019). A recent comparative study of 337 cities found out that cities that participation in TNMs are more likely to take adaptation action and that being part of multiple networks leads to higher levels of adaptation planning (Heikkinen et al., 2020).

Input-driven institutional change creates incentives to deliver adaptation action. An input view focuses on the intrinsic capacities of a given organisation. Input indicators are often referred to as political capital (Rosenzweig and Solecki, 2018; Diederichs and Roberts, 2016), existing or endogenous resources (Moloney and Fünfgeld, 2015; Wamsler and Brink, 2014), or local drivers for adaptation (Dilling et al., 2017). ReResearch conducted across two municipalities in Western Cape, South Africa, showed the importance of a dedicated environmental champion, access to a knowledge base, the availability of resources, political stability and the presence of dense social networks (Pasquini et al., 2015). Research from São Paulo, Brazil, showed how intrinsic political capacities and contextual factors, such as the political ideology of elected officials, shaped opportunities for embedding adaptation into ongoing urban agendas (Di Giulio et al., 2018).

Networks, interactions and actor coalitions shape options for institutional change. Aylett (2015) noted the importance of internal networks between municipal departments, including informal communication channels, cultivating personal contacts and trust between the person or team responsible for climate planning and staff within other local government agencies. Internal networks can facilitate the commitment of local elected officials (Hughes, 2015), support higher municipal expenditures per capita and foster perceptions that climate adaptation is needed (Shi, Chu and Debats, 2015). Collective decision-making can integrate multiple types of information with moral concerns and provide key rationales that enable adaptation action (Carlson and McCormick, 2015). In urban areas in Africa, research on internal networks has also investigated how informal arrangements shape action possibilities (Satterthwaite et al., 2020). For example, in Zimbabwe, informal, traditional and civil society institutions are core arenas for issue discussion because of lower public sector capacities (Mubaya and Mafongoya, 2017). In Durban, South Africa, local governments rely considerably on shadow systems and informal spaces of information and knowledge exchange across their operations to introduce and sustain new ideas (Leck and Roberts, 2015). In the metropolitan area of Styria, Austria, informal cooperation has supported the development of rural–urban partnerships for the formulation of common goals (Oedl-Wieser et al., 2020). In Arkansas, USA, informal governance structures support planning to manage wildfires (Miller, Vos and Lindquist, 2017).

Cities can leverage input-driven institutional change even without national support for climate change adaptation or mitigation. For example, where cities have defined policymaking and budget raising powers, city level political leadership can support adaptation action going beyond national policy (Hamin, Gurran and Emlinger, 2014; Shi, Chu and Debats, 2015; Carlson and McCormick, 2015). Examples include the Surat Climate Change Trust in Surat, India (Chu, 2016) and Initiative for Urban Climate Change and Environment in Semarang, Indonesia (Taylor and Lassa, 2015). In Saint Louis, Senegal, support from national and state-level actors enabled local institutional change (Vedeld et al., 2016). Processual levers may be also mobilised in situations of political instability (which disrupts patterns in champions and networks), clientelism (which can cause environmental projects to be discontinued) (Pasquini et al., 2015) or in contexts where there are high political and socioeconomic inequalities (Harris, Chu and Ziervogel, 2018; Chu, Anguelovski and Carmin, 2016).

Output-driven institutional change is shaped by organisational products such as strategies, plans, policies and evaluative metrics (Patterson and Huitema, 2019; Bellinson and Chu, 2019) (See Table 6.8). There are numerous examples of institutional change through planning outcomes. For example, Manizales, Colombia has included climate adaptation into long-established environmental policy (Biomanizales) and a local environmental action plan (Bioplan), which follows on from a long coherent trajectory of climate change policy (Hardoy and Velásquez Barrero, 2014). A significant number of North American cities have integrated adaptation into long-range plans, while fewer cities integrate adaptation in sustainable development plans or sectoral plans (Aylett, 2015). Canadian cities are more likely to have a plan specifically focused on adaptation rather than having adaptation integrated into municipal long-range planning (Aylett, 2015). In the European Union, adaptation plans depended on national climate legislation or, in fewer cases, the influence of an international climate network (Reckien et al., 2018b). A comparative report from the Covenant of Mayors, however, suggests that the adaptation pillar needs development to demonstrate the effectiveness of adaptation responses and their integration with mitigation goals (Bertoldi et al., 2020). Municipalities in Sweden have been called ‘pre-reactive’ because adequate strategic guidelines are in place to frame the accessibility, aesthetics and adaptability of waterfront developments (Storbjörk and Uggla, 2015). Some Asian cities also report high output effectiveness, where they are more likely to indicate senior local government officials’ performance management contracts, the budgeting procedures of local government agencies and the procedures that local government agencies use for budgeting infrastructure spending (Aylett, 2015). Despite this evidence, there is a gap in understanding the general trends of planning and institutional change in Africa, Asia, East Europe and the Middle East.

Institutional change processes are complex, contested and sporadic (Patterson, de Voogt and Sapiains, 2019). Such processes are often inhibited by unclear planning mandates, conflicting development priorities, lack of leadership and resource and capacity shortfalls (Anguelovski et al. 2014). There is no one size fits all approach to institutional change, which works in situ, and benefits from clearly defined plans and an incremental approach to revising new elements and addressing gaps or failures (Beunen et al., 2017). A longitudinal view of institutional change allows for assessing actors and dynamics involved in integrating adaptation into the sectoral agendas or governance arrangements mentioned above (Patterson and Huitema, 2019).

Table 6.8 | Examples of institutional and policy instruments to enable adaptation

Success in urban adaptation is most often understood as requiring measurable outcomes and evaluation (see also Section 6.4.6). However, many adaptation outcomes are not measurable (medium evidence, medium agreement ) (Béné et al., 2018). Adaptation action solely focused on action tends to ignore areas in the city for which there is no existing data even though actions in these areas may play an essential role in shaping resilience and its limits. Informal settlements and informal economies, which are integral in managing urban resources for effective climate adaptation, are not routinely included in formal urban and national monitoring(Guibrunet and Castán Broto, 2016). The resulting understanding and monitoring of city needs, capacities and actions that feed into policy is incomplete. The innovation, as well as particular concerns and capacities of the informal sector, which is often highly gendered, are not always measured (Brown and McGranahan, 2016). An emphasis on measurable adaptation outcomes may lead to prioritising techno-economic measures to adaptation at the local level. Technocratic approaches to environmental policy continue to shape local sustainability politics (Bulkeley, 2015). The deployment of such technocratic approaches at the local scale is detrimental for democratic and collaborative practices (Metzger and Lindblad, 2020). For example, while China has received praise in terms of delivering urban policies that put climate change at its core, thus suggesting its role providing leadership in climate change debates (Liu et al., 2014; Wang and He, 2015; Fu and Zhang, 2017), other analyses suggest that processes of planning should take greater account of certain groups and interests (Westman and Broto, 2018). Urban sustainability policy may, as a result, fail to deliver collaborative social and environmental objectives, and this is maladaptive in terms of CRD.

Two forms of mainstreaming are usually found in urban policy: incorporating climate adaptation into different sectors or incorporating climate adaptation in holistic sustainability or resilience plans, linking climate adaptation objectives with other social and development objectives (Reckien et al., 2019; Fainstein, 2018). The integration of climate adaptation in local policies in cities and settlements has often been seen as maintaining business-as-usual and not always aligned with transformative efforts to address structural drivers of vulnerability (high confidence). For example, mainstream actions that seek to advance other development objectives, as explained above, may reduce adaptation to ‘low-hanging fruits’, which may maintain business-as-usual practices without any fundamental transformation of the social, institutional and economic systems that drive vulnerabilities (Aylett, 2014). However, as explained above, mainstreaming can also generate wider processes of institutional change (Section 6.4.2). Mainstream strategies may help to demonstrate how policy and frameworks can produce practical outcomes on the ground (Biesbroek and Delaney, 2020). However, previous experiences in other sectors, such as gender mainstreaming, have shown the limitations of the mainstreaming approach, particularly in terms of addressing the structural drivers of inequality and vulnerability, and in achieving justice for those who suffer most (Moser, 2017). Local governments in particular, can link mainstreaming efforts with specific strategies that support justice in adaptation, including redistribution efforts to address vulnerabilities (see Section 6.3.2), representation in local institution and deliberative processes, and recognition of the conditions for self-realisation, including personal and collective safety (Agyeman et al., 2016; Castán Broto and Westman, 2017; Castán Broto and Westman, 2019; Hess and McKane, 2021).

Coordination of adaptation policy goals cuts across cities to integrate them into international processes of climate policy formulation; coordination in cities produces effective collective outcomes, cementation of common standards and methodologies for climate action (e.g., emission inventories) (high agreement , medium evidence) (Gordon and Johnson, 2017; Hsu and Rauber, 2021). A collective global response has become a significant concern in international climate policy (Chan et al., 2015a). The UNFCCC has adopted a role as an orchestrator, including providing framework for city governments (Bäckstrand and Kuyper, 2017). Within cities, coordination can arise from active programming; for example, in Rotterdam and New York City, local authorities adopted long-term objectives and conditions for action, bringing together a multiplicity of actors across sectors to orient contributions, share knowledge and coordinate actions (Hölscher et al., 2019). Where national politics is supportive, coordination between city and national government is an asset (Chan and Amling, 2019; Inch, 2019). The use of social media and digital mechanisms for coordination with public interest is ambiguous: in China, Weibo has facilitated an expansion of public engagement, although it remains top down and dominated by a few influencial actors (Liu and Zhao, 2017; Yang and Stoddart, 2021). The pilot project #OurChangingClimate is one example of engaging youth with an understanding of their communities and their resilience or vulnerability to climate change (Napawan, Simpson and Snyder, 2017).

Co-production can advance urban sustainability and social justice in cities and settlements to provide infrastructure adapted to the human scale and advancing SDGs (medium confidence) (McGranahan, 2015; McGranahan and Mitlin, 2016; Chowdhury, Jahan and Rahman, 2017; Moretto and Ranzato, 2017; Nastiti et al., 2017). Co-production involves the active involvement of citizens and citizens’ organisation in iterative public service planning and delivery, and has become increasingly central in climate change responses alongside other bottom-up, community-led strategies (Bremer et al., 2019; Vasconcelos, Santos and Pacheco, 2013).

Co-production builds on public participation that brings together diverse sets of citizen interests, values and ideas to inform change and solve problems relating to a collective adaptation challenge (Archer et al., 2014; Bisaro, Roggero and Villamayor-Tomas, 2018; Sarzynski, 2015), and is increasingly important in environmental policy more widely (McGranahan, 2015; Moretto and Ranzato, 2017). For example, in three cities across the Czech Republic, stakeholder participation exercises were used to prioritise climate change risks, provide impetus and opportunity for knowledge co-production, and support adaptation planning (Krkoška Lorencová et al., 2018). In municipalities in Malaysia, stakeholders and citizens are active in the adaptation policy cycle (Palermo and Hernandez, 2020). In Quebec, Canada, citizens collaborated with the municipal authority to bring together climate science and ‘ordinary’ urban management and design solutions (Cloutier et al., 2015). Service co-production enables integrating multiple actors in the management and delivery of public services (Pestoff and Brandsen, 2013; Pestoff, Brandsen and Verschuere, 2013). Civil society-driven, co-productive approaches can pioneer new forms of institutional relations and practices filling gaps where the public sector is absent or retreating (Frantzeskaki et al., 2016).

A co-production approach to climate change governance addresses the increasing public interest on climate change (Davies, Broto and Hügel, 2021). Youth movements such as Forum for Future have joined forces with other environmental and Indigenous organisations to lobby governments and institutions to action (Kenis, 2021; Fisher and Nasrin, 2021; Davies and Hügel, 2021 ; Hayward, 2021). These movements have built momentum moving local governments and other institutions to declare a climate emergency and have supported the creation of new forums where climate change can be addressed collectively, such as citizens’ assemblies. In the UK, for example, initial scepticism has led to the proliferation of citizen-centric Climate Assemblies at the local level (Sandover, Moseley and Devine-Wright, 2021).

Cooperative governance models provide insights for designing forms of participatory and collaborative planning through which communities and state actors can identify concrete actions and resources to improve services and mitigate structural vulnerabilities to disasters (Castán Broto et al., 2015). Experiences of co-production of sanitation services show how co-production may improve outcomes, while at the same time opening up avenues for grassroots organisations to claim political influence (McGranahan and Mitlin, 2016). Co-production may change institutions in response to external interventions (Das, 2016). Although there are drawbacks in terms of the extent to which co-production can be used to legitimise unfair interventions within a given context, co-production may also be a tool for improving the accountability of dominant groups to vulnerable sectors of the population (Nastiti et al., 2017). There are limitations to co-production. The city of Barcelona, Spain, used co-production methodologies to develop the Barcelona Climate Plan. However, policymakers and civil servants were reluctant to use lay knowledge from participants and political deadlines constrained the time dedicated to deliberation (Satorras et al., 2020).

Inclusive and sustainable adaptation can address the causes of systemic vulnerability (medium evidence, high agreement ). This points to the fundamental requirements of adaptation action in line with the Universal Declaration of Human Rights. Climate justice theories draw on the environmental justice movement experiences at the local level (Bickerstaff, 2012; Bickerstaff, Walker and Bulkeley, 2013; Perez et al., 2015; Hall, Hards and Bulkeley, 2013). Slogans such as ‘leave no one behind’ embedded in international policy for cities and settlements recognise the connection between systems of oppression and exclusion that reproduce and perpetuate urban inequality and the delivery of urban services and security (Kabeer, 2016; Stuart and Woodroffe, 2016).

Intersectional strategies of action seek to consider the multiple forms of structural oppression experienced at the local level (Grunenfelder and Schurr, 2015) and, in the context of adaptation, explain how they produce or exacerbate vulnerabilities. For example, intersectionality ties with the idea of how multiple deprivations shape access to services (from sanitation to health and education) and the exposition to environmental risks (Sicotte, 2014; Lau and Scales, 2016; Van Aelst and Holvoet, 2016; Lievanos and Horne, 2017; Raza, 2017; Yon and Nadimpalli, 2017; European Environment Agency, 2020) (see Box 6.6 on the participation of women in local decision making bodies). For example, fisherwomen in the western coast of India rely on a complex arrangement of relationships around categories of class, caste and gender that shapes their possibilities to draw political resources to maintain their livelihoods and, hence, influence the dynamics of transformation (Thara, 2016). Intersectionality is central to build resilience across communities, rather than in particular areas (Khosla and Masaud, 2010; Reckien et al., 2017). Including intersectionality deliberately in partnerships with communities can empower socially excluded groups and highlight justice issues while aligning agendas with local development priorities (Castán Broto et al., 2015a). Despite the high confidence on the growing importance of intersectionality concerns in the delivery of just environmental policies, there is limited evidence of its explicit inclusion in adaptation policies.

Cities and settlements in higher-income countries typically have access to funding that could be used to enhance resilience and build adaptive capacity; this includes both the private resources of individual households and firms (which varies significantly within and among cities) and public budgets of different government tiers (see Table 6.10).

Depending on fiscal devolution levels within a country, public revenues may be collected and managed primarily at the national, state, metropolitan or local level. In federal countries, sub-national governments collect an average of 49.4% of public revenues, compared with only 20.7% in unitary countries (OECD/UCLG, 2019). For example, sub-national revenues represent over a quarter of total public revenues in Belgium, Canada and Denmark, but less than 5% in Greece, Ireland and New Zealand (OECD/UCLG, 2019). The share of the national revenue transferred to sub-national governments also varies significantly among countries: grants and subsidies account for over three-quarters of sub-national government revenue in Malta, but less than a quarter of sub-national revenue in Iceland (OECD/UCLG, 2019). A local government’s capacity to collect revenues is further mediated by incomes within a city (which dictates the prospective tax base) and the capacity of civil servants to administer taxes, fees and charges. The result is that metropolitan and local governments’ budgets vary dramatically, across and within countries. For example, per capita municipal budgets vary from USD 1114 in Saskatoon and USD 2682 in Peterborough (Canada), USD 2635 in Leipzig and USD 3638 in Freiburg (Germany), to USD 4907 in Bristol and USD 5612 in Aberdeen (UK) (Löffler, 2016).

Revenue streams are often insufficient relative to the scale of adaptation requirements. For example, Kano, Nigeria, is a large urban area that urgently needs investment in human development and climate resilience but where a fragmented local government has little capacity to finance their climate plans (Mohammed, Hassan and Badamasi, 2019). Many local governments are unable to mobilise funds for adaptation as they face competing priorities, meaning that resources for resilience must be allocated by higher levels of government (Hughes, 2015), which also perceive opportunity costs to adaptation investments. Funding from non-state actors is, therefore, proving important. For example, in the USA, private foundations and non-profit organisations account for 17% and 16%, respectively, of adaptation support in urban areas (Carmin, Nadkarni and Rhie, 2012). However, tapping into these funding sources raises complex questions about accountability and ownership of urban adaptation (Chu, Schenk and Patterson, 2018). Land reclamation may foster real estate markets and mobilise finance for adaptation, as shown in Germany, the Netherlands and the Maldives (Bisaro et al., 2019).

City governments need to anticipate climate shocks and stresses, and design their operating models and investment plans accordingly to ensure financial resilience (Clarvis et al, 2015). Climate risks threaten fiscal models, for example, a drought may disrupt water revenues by reducing total water consumption and incentivising households and firms to invest in independent water storage or supply infrastructure (Simpson et al., 2019). Storm surges and sea level rise may threaten sunk investments in revenue-generating infrastructures such as toll roads or electricity generation and transmission systems. .

Common sources of adaptation finance might include donor agencies including the Green Climate Fund, sovereign funds (e.g., the Bangladesh Climate Change Resilience Fund) and private finance from commercial banks, investment companies, pension funds and insurance companies (Floater et al., 2017). These capital sources have different risk–return expectations and investment horizons, so they will suit different types and stages of projects. Many sub-national governments in the Global North have access to well-developed domestic, if not global, capital markets to raise and steer finance for urban investment (Banhalmi-Zakar et al., 2016).

However, investments in ex ante urban climate adaptation may prove less attractive to these financiers than other opportunities because of their long maturities and high risk (Keenan, Chu and Peterson, 2019) (see also Table 6.11). Many generate economic returns primarily through avoided losses from climate impacts, which are difficult to measure and are, in any case, more attractive to funders than financiers (Kaufman, 2014). Ex post , insurance already plays a critical role in protecting urban households, firms and other stakeholders from the full economic costs of high-severity, low-frequency events by sharing risk over time and space. Insurance can also be designed to incentivise risk-reducing behaviours and investments (Banhalmi-Zakar et al., 2016; Paddam and Wong, 2017). Some researchers suggest that, in urban environments, insurance practices are helping to establish adaptation and risk as a new area of public health and public protection. For example, local governments are using new risk transfer instruments, such as re-insurance and catastrophe bonds, to fund investments in resilience projects and disaster recovery (Collier and Cox, 2021). However, the commercial feasibility of private sector insurance depends on more robust estimates of current and future risks, and premiums commensurate with the ability and willingness of consumers to pay. Therefore, ex ante investments must complement insurance schemes to improve climate modelling and reduce climate risk (Surminski, Bouwer and Linnerooth-Bayer, 2016). The private sector also faces practical barriers to invest in adaptation.

Table 6.11 | Barriers to finance adaptation in urban areas (Richmond et al., 2021)

National governments typically determine the fiscal transfers that sub-national governments receive and the taxes, fees and charges they permit to collect (see for example CBO, 2016). Local governments may strengthen their own source revenue collection and management capacities to better exploit these funding streams and improve their balance sheets, but their total budget will be limited to these funding sources (Ahmad et al., 2019). The amount of local public funding available for urban adaptation depends on the relationships across different government levels.

Similarly, mobilising private finance for urban adaptation projects demands robust institutional, fiscal and regulatory frameworks, which are typically national authorities’ responsibility. For local governments to access private finance for adaptation may require national (or in federal countries, state) governments to reform policies and rules governing municipal borrowing, public–private partnerships, land value capture instruments and other financing mechanisms (Ware and Banhalmi-Zakar, 2017). Such fiscal reforms tap into fundamental political and policy issues, such as local governments’ autonomy or the tariff-setting powers of national ministries (Gorelick, 2018; White and Wahba, 2019).

Access to private finance can support infrastructure development through private provisioning, public–private partnerships (PPP) and public debt arrangements (high confidence) (see also Section 6.4.1.2). Private provisioning attracts coastal adaptation investment when returns are high (e.g., when there is a real estate market associated with it) (Bisaro and Hinkel, 2018). Public–private partnerships attract investments from dredging and construction companies that involve a large share of operational costs (Bisaro and Hinkel, 2018). Public debt instruments appear less successful in supporting investment in adaptation infrastructure. Real estate firms focus on adaptation actions if they perceive climate change impacts such as flooding may impact their activity, mostly focusing on adaptation action as a means to gain competitive advantage (Teicher, 2018).

There have been numerous attempts to innovate in climate finance, for example, mobilising community and cooperative forms of finance, or crowdfunding, which have already proven effective in the context of mitigation (De Broeck, 2018). A well-studied instrument in urban environments is land-value capture. Land-value capture refers to communities’ ability to capture the benefit of increased land values that result from public investment or other government actions (Germán and Bernstein, 2020). There is considerable potential to mobilise land-value capture for adaptation (limited evidence, medium agreement ), but its potential remains unexplored (Dunning and Lord, 2020). While there are numerous examples of the mobilisation of land-value capture to finance sustainable development action (Li and Love, 2019; Wang, Samsura and van der Krabben, 2019), there is limited evidence of its use in climate adaptation (see Case Study 6.2). These innovations are particularly important in contexts where resources are very constrained, such as in the financing of adaptation in African cities (See Box 6.7).

Corruption in urban adaptation and disaster risk management finance is a considerable but little researched challenge observed from all world regions (Sanderson et al., 2021). Corruption generates maladaptation, increasing risk, for example where infrastructure is constructed with faulty design, substandard materials and inadequate maintenance (Kabir et al., 2021). More widely, corruption increases vulnerability and reduces capacity by damaging the body politic, distorting markets and reducing economic growth (Alexander, 2017). The construction and infrastructure industries are repeatedly identified as sources of corruption (GIACC, 2020; Chan and Owusu, 2017; Sanderson et al., 2021). Corruption and misuse of climate finance is exacerbated by limited public access to information, political considerations in finance decision making and lack of accountability for decisions and actions (Kabir et al., 2021). In construction, Owusu et al. (2019) found causes included too-close relationships, poor professional ethical standards, negative industrial and working conditions, negative role models and inadequate sanctions throughout the phases of construction. Post-disaster response and reconstruction, and periods of surge funding following international or national policy priorities are especially vulnerable to corruption, with increased funding and pressure to lower norms of financial management (Imperiale and Vanclay, 2021). Mixed delivery mechanisms have been shown to reduce corruption, for example where civil society organisations are involved in project approval stages, although there is also a risk that civil society organisations will themselves become entangled in corruption. International donors have a role to play in working with government and civil society to promote wider scrutiny and transparency of financing processes and project delivery through promoting media and press freedom and legislation for access to information to reduce corruption by enhancing transparency and accountability (Kabir et al., 2021).

Expanding the resource envelope available for adaptation investment is often beyond the authority or competency of city governments. Sovereign and state governments have critical roles to play in providing funding or securing finance for adaptation investments. Such a role is particularly important where the impacts of climate change are distributed inequitably across a country, so that the costs borne by a city may exceed local budgets.