Editing IPCC:AR6/WGII/Chapter-6 (section)

=== 6.2.2 Dynamic Interaction of Urban Systems with Climate ===

<div id="h2-7-siblings" class="h2-siblings"></div>

Urban systems interact with climate systems in multiple, dynamic and complex ways ( [[#6.1.1|Section 6.1.1]] , Doblas-Reyes et al., 2021 Box 10.3). Climate change can have direct impacts on the functioning of urban systems, while the nature of those systems plays a substantial role in modifying the effects of climate change ( ''high confidence'' ) (Frank, Delano and Caniglia, 2017; [[#Smid--2018|Smid and Costa, 2018]] ). An example of this urban system climate nexus is the urban heat island effect (discussed in [[#6.2.3.1|Section 6.2.3.1]] ) ( [[#Susca--2020|Susca and Pomponi, 2020]] ). Assessing the inter-relationships between multiple systems and a range of hazards is particularly important as many cities are presently exposed to multiple climate-related hazards: more than 100 cities analysed as part of a 571 city study in Europe were deemed vulnerable to two or more climate impacts (Guerreiro et al., 2018). Rapid expansion of urban areas increases the exposure of urban populations to various hazards independent of global climate change. [[#Huang--2019|Huang et al. (2019)]] project that urban land areas will expand by 0.6–1.3 million km 2 between 2015 and 2050, an increase of 78–171% over the urban footprint in 2015. Specifically in relation to floods and droughts, Güneralp et al. (2015) calculate that even without accounting for climate change, the extent of urban areas exposed to flood hazards will increase 2.7 times between 2000 and 2030, the extent exposed to drought hazards will approximately double during this period, and urban land exposed to both floods and droughts will increase more than 2.5 times.

This section assesses observed and expected impacts from the main hazards identified for cities, settlements and infrastructure; temperature extremes (and the urban heat island), flooding (including sea level rise), water scarcity and security, as well as other hazards that are either less well-studied and/or likely to affect only a limited number of locations. The data assessed in this section are limited by uneven coverage. Despite improvements since AR5, data continue to be more complete for extreme events than for chronic hazards and everyday risks, which may have high aggregate impacts and disproportionately erode the well-being of urban poor households, especially for the most vulnerable, including women, children, the aged, disabled and homeless (van Wesenbeeck, Sonneveld and Voortman, 2016; Kinay et al., 2019; Connelly et al., 2018). Data coverage is also less comprehensive for smaller settlements in poorer countries, the locations where urban growth is often high and adaptive capacities are often low (e.g., Rufat et al., 2015). Thus, data gaps frequently coincide with highly vulnerable populations (Rufat et al., 2015; [[#Satterthwaite--2017|Satterthwaite and Bartlett, 2017]] ). Here, even small changes in livelihoods, health, or representation and voice can rapidly bring households into positions of risk, even when hazard conditions are relatively stable (Ziervogel et al., 2017). These structural limits in available data are discussed also in Section 7 (Health, Well-being and the Changing Structure of Communities) and Section 8 (Poverty, Livelihoods and Sustainable Development), and in Doblas-Reyes et al. (2021) Box 10.3. There are implications also for adaptation ( [[#6.3|Section 6.3]] ), where the greater availability of evidence on exposure-driven risk can limit resilience-building interventions that focus on the reduction of vulnerability.

<div id="6.2.2.1" class="h3-container"></div>

<span id="temperatures-and-the-urban-heat-island"></span>
==== 6.2.2.1 Temperatures and the Urban Heat Island ====

<div id="h3-1-siblings" class="h3-siblings"></div>

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|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--2021|Klein and Anderegg, 2021]] ).

<div id="_idContainer012" class="Figure"></div>

[[File:95ebfce822c99664f5ae5dda0a9e14f0 IPCC_AR6_WGII_Figure_6_003.png]]

'''Figure 6.3 |''' '''Global distribution of population exposed to hyperthermia from extreme heat for (a) the present, and projections from selected Representative Concentration Pathways in (b) the mid-21st century and (c) the end of the 21st century.''' Shading indicates projected number of days in a year in which conditions of air temperature and humidity surpass a common threshold beyond which climate conditions turned deadly and pose a risk of death (Mora et al ''.'' , 2017). Named cities are the top 15 urban areas by population size during 2020, 2050 and 2100, respectively, as projected by [[#Hoornweg--2017|Hoornweg and Pope (2017)]]

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--2020|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--2017|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 [https://www.ipcc.ch/chapter/6#CCP2.2 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%20Milan--2015|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|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|Vanos, 2015]] ; [[#Hyndman--2017|Hyndman, 2017]] ). Young children in cities are particularly sensitive to heatwaves, and may have little experience or capacity to cope with heat extremes ( [[#Norwegian%20Red%20Cross--2019|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|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--2015|Liu and Coley, 2015]] ; van Hooff et al., 2014; Vardoulakis et al., 2015; [[#Dodoo--2016|Dodoo and Gustavsson, 2016]] ; [[#Invidiata--2016|Invidiata and Ghisi, 2016]] ; [[#Makantasi--2016|Makantasi and Mavrogianni, 2016]] ; [[#Mulville--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--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--2015|Liu and Coley, 2015]] ; van Hooff et al., 2014; Vardoulakis et al., 2015; [[#Dodoo--2016|Dodoo and Gustavsson, 2016]] ; [[#Invidiata--2016|Invidiata and Ghisi, 2016]] ; [[#Makantasi--2016|Makantasi and Mavrogianni, 2016]] ; [[#Mulville--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--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--2016|Makantasi and Mavrogianni, 2016]] ; [[#Mulville--2016|Mulville and Stravoravdis, 2016]] ; Salthammer et al., 2018; [[#Fisk--2015|Fisk, 2015]] ; Hamdy et al., 2017; Fosas et al., 2018; [[#Ozarisoy--2019|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%20Zivin--2014|Graff Zivin and Neidell, 2014]] ; [[#Yi--2017|Yi and Chan, 2017]] ; Houser et al., 2015; [[#Stevens--2017|Stevens, 2017]] ; see [[IPCC:Wg2:Chapter:Chapter-8#8.2.1|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--2015|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 2013–14 [[#footnote-002|2]] ; 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) [[#footnote-001|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--2014|Soebarto and Bennetts, 2014]] ; [[#Zander--2019|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 [[#6.2.2.4|Section 6.2.2.4]] ).

Thermal inequity can also be seen as a distributive justice risk ( [[#Mitchell--2018|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--2018|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--2017|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).

<div id="6.2.2.2" class="h3-container"></div>

<span id="urban-flooding"></span>
==== 6.2.2.2 Urban Flooding ====

<div id="h3-2-siblings" class="h3-siblings"></div>

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|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'' [[#footnote-000|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); [[IPCC:Wg2:Chapter:Chapter-4#4.2.4|Section 4.2.4.2]] ). [[#Mishra--2015|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 [[IPCC:Wg2:Chapter:Chapter-9#9.3|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|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--2017|Koop and van Leeuwen, 2017]] ; Hoegh-Guldberg et al., 2018; see also Section [https://www.ipcc.ch/chapter/6#CCP2.2 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|IPCC, 2019b]] ; Winsemius et al., 2015; [[#Kulp--2019|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--2019|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); [[IPCC:Wg2:Chapter:Chapter-4#4.2.4|Section 4.2.4.5]] ; [[#6.2.4.2|Section 6.2.4.2]] ; Box 7.2; [[IPCC:Wg2:Chapter:Chapter-8#8.4.5.2|Section 8.4.5.2]] ).

<div id="6.2.2.3" class="h3-container"></div>

<span id="urban-water-scarcity-and-security"></span>
==== 6.2.2.3 Urban Water Scarcity and Security ====

<div id="h3-3-siblings" class="h3-siblings"></div>

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--2019|Romero-Lankao and Gnatz, 2019]] ; Allan, Kenway and Head, 2018; Huang, Xu and Yin, 2015; [[#Chen--2016|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; [[IPCC:Wg2:Chapter:Chapter-4#4.2.4|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--2018|Jones and van Vliet, 2018]] ; [[#Leddin--2020|Leddin and Macrae, 2020]] ; [[#Lorenzo--2020|Lorenzo and Kinzig, 2020]] ; Ma et al., 2020; [[#Mosley--2015|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--2017|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--2019|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|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--2015|Fulton and Cooley, 2015]] ; [[#Rushforth--2016|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--2015|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|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--2015|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; [[IPCC:Wg2:Chapter:Chapter-4#4.2.4|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|]] [[#Döring--2020|Döring, 2020]] ; Ide et al., 2021), particularly when road infrastructures and access to water are limited ( [[#Detges--2016|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--2020|Akhtar and Shah, 2020]] ; [[#Singh--2019|Singh and Sharma, 2019]] ).

<div id="6.2.2.4" class="h3-container"></div>

<span id="other-dynamic-interactions"></span>
==== 6.2.2.4 Other Dynamic Interactions ====

<div id="h3-4-siblings" class="h3-siblings"></div>

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|Jendritzky, 1999]] ), as well as failures in overhead power lines and towers leading to outages in Moscow and Bucharest ( [[#Panteli--2015|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|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--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--2016|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--2018|Froude and Petley, 2018]] ). Although the spatial footprint of an individual landslide might be small (i.e., &lt; 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--2014|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--2016|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--2015|Rosencrants and Ashley, 2015]] ; [[#Ashley--2016|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|Okunola, 2019]] ), and on urban trees that may be uprooted by strong wind gusts from downbursts ( [[#Ordóñez--2015|Ordóñez and Duinker, 2015]] ; [[#Pita--2016|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|IPCC, 2018]] ), ''likely'' enable weather conditions driving fire events impacting cities within these regions ( [[IPCC:Wg2:Chapter:Chapter-2#2.4.4.2|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--2020|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--2017|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--2016|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 [[IPCC:Wg2:Chapter:Chapter-12|Chapter 12]] Box 12.1]) and anthropogenic sources that are influenced by climate change (e.g., fine particulate matter such as PM 2.5 , 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|Leung, 2015]] ; Peduzzi et al., 2020; Cross-Chapter Box HEALTH in Chapter 7).

The mortality risk for several pollutants, for example PM 2.5, is considerable ( ''high confidence'' ). Current estimates indicate that 95% of global population live in areas where ambient PM 2.5 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 PM 2.5 concentrations are generally highest in cities in Africa, South Asia, the Middle East and East Asia; PM 2.5 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 PM 2.5 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 PM 2.5 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 PM 2.5 due to climate change are less clear than for ozone and may be relatively smaller (Westervelt et al., 2019) as climate change can affect PM 2.5 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 PM 2.5 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--2017|Silva et al. (2017)]] found that global premature mortality attributable to climate change (and not from urbanisation) from ozone and PM 2.5 will increase by about 260,000 deaths per year in 2100 under RCP8.5, but substantial variance in results exists between individual models.

<div id="6.2.3" class="h2-container"></div>

<span id="differentiated-human-vulnerability"></span>