FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1
2 Chapter 16: Key Risks Across Sectors and Regions
3
4 Coordinating Lead Authors: Brian O'Neill (USA), Maarten van Aalst (The Netherlands), Zelina Zaiton
5 Ibrahim (Malaysia)
6
7 Lead Authors: Lea Berrang Ford (United Kingdom/Canada), Suruchi Bhadwal (India), Halvard Buhaug
8 (Norway), Delavane Diaz (USA), Katja Frieler (Germany), Matthias Garschagen (Germany), Alexandre
9 Magnan (France), Guy Midgley (South Africa), Alisher Mirzabaev (Uzbekistan/Germany), Adelle Thomas
10 (Bahamas), Rachel Warren (United Kingdom)
11
12 Contributing Authors: Sharina Abdul Halim (Malaysia), Idowu Ajibade (Nigeria/Canada/USA), Philip
13 Antwi-Agyei (Ghana), Richard Betts (United Kingdom), Rachel Bezner Kerr (USA), Joern Birkmann
14 (Germany), Martina Angela Caretta (USA), Tamma Carleton (USA), Connor Cavanagh (Ireland/Norway),
15 Winston Chow (Singapore), Gueladio Cisse (Mauritania/ Switzerland/France), Andrew Constable
16 (Australia), Mark Costello (New Zealand), Jackie Dawson (Canada), Richard Dawson (United Kingdom),
17 Simon Donner (Canada), Sybren Drijfhout (The Netherlands), Virginie K.E. Duvat (France), Kristie Ebi
18 (USA), Tamsin Edwards (United Kingdom), Francois Engelbrecht (South Africa), Alexandra Paige Fischer
19 (USA), Isabel Fletcher (United Kingdom), James Ford (United Kingdom), Eranga Galappaththi (Sri
20 Lanka/Canada), Francois Gemenne (Belgium), Patrick Gonzalez (USA), Neal Haddaway
21 (Sweden/Germany/United Kingdom), Isabel Hagen (Switzerland), Stephane Hallegatte (France/USA),
22 Toshihiro Hasegawa (Japan), Masahiro Hashizume (Japan), Cullen Hendrix (USA), Kevin Hennessy
23 (Australia), Tom Hertel (USA), Jeremy Hess (USA), Helene Hewitt (United Kingdom), Kirstin Holsman
24 (USA), Veronika Huber (Germany/Spain), Christian Huggel (Switzerland), Bramka Jafino (Indonesia/The
25 Netherlands), Kripa Jagannathan (Canada/India/USA), Rhosanna Jenkins (United Kingdom), Chris Jones
26 (United Kingdom), Vhalinavho Khavhagali (South Africa), Elco Koks (The Netherlands), Gerhard Krinner
27 (France), Judy Lawrence (New Zealand), Gonéri Le Cozannet (France), Alexandra Lesnikowski (Canada),
28 Karen Levy (USA), Tabea Lissner (Germany), Rachel Lowe (United Kingdom), Simone Lucatello (Mexico),
29 Yong Luo (China), Brendan Mackey (Australia), Shobha Maharaj (Germany/Trinidad and Tobago),
30 Custodio Matavel (Mozambique/Germany), Timon McPhearson (USA), Veruska Muccione (Switzerland),
31 Aditi Mukherji (India), Didacus Namanya (Uganda), Gerald Nelson (USA), David Obura (Kenya), Jean
32 Ometto (Brazil), Friederike Otto (United Kingdom/Germany), Camille Parmesan (United Kingdom/USA),
33 Patricia Pinho (Brazil/United Kingdom), Franziska Piontek (Germany), Prajal Pradhan (Nepal/Germany),
34 Jeff Price (United Kingdom), Joacim Rocklov (Sweden), Steven Rose (USA), Alexander Ruane (USA),
35 Daniela Schmidt (Germany/United Kingdom), Alcade Segnon (Republic of Benin), Sonia Seneviratne
36 (Switzerland), Olivia Serdeczny (Poland/Germany), Mohammad Aminur Rahman Shah (Bangladesh/United
37 Kingdom), Yuanyuan Shang (China/Australia), Roopam Shukla (India/Germany), AR Siders (USA),
38 Nicholas P. Simpson (Zimbabwe/South Africa), Chandi Singh (India), Asha Sitati (Kenya), Tom Spencer
39 (United Kingdom), Nicola Stevens (South Africa), Emily Theokritoff (Germany), Maria Cristina Tirado-von
40 der Pahlen (Spain/USA), Christopher Trisos (South Africa), Nicola Ulibarri (USA), Mariana Vale (Brazil),
41 Krispa Vasant (India), Ana Vicedo-Carbrera (Spain/Switzerland), Colette Wabnitz (Canada), Anita Wreford
42 (New Zealand), Gary Yohe (USA), Carol Zavaleta (Peru), Xuebin Zhang (Canada), Zhibin Zhang (China)
43
44 Review Editors: Tong Jiang (China), Michael Oppenheimer (USA)
45
46 Chapter Scientist: Rhosanna Jenkins (United Kingdom)
47
48 Date of Draft: 1 October 2021
49
50 Notes: TSU Compiled Version
51
52
53 Table of Contents
54
55 Executive Summary..........................................................................................................................................3
56 16.1 Introduction and Framing .......................................................................................................................9
57 16.1.1 Objective of the Chapter .................................................................................................................9
Do Not Cite, Quote or Distribute 16-1 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 16.1.2 Risk Framing...................................................................................................................................9
2 16.1.3 Storyline of the Chapter, and What's New Compared to Previous Assessments..........................10
3 16.1.4 Drivers of Exposure and Vulnerability .........................................................................................13
4 16.2 Synthesis of Observed Impacts..............................................................................................................14
5 16.2.1 Definitions.....................................................................................................................................15
6 16.2.2 Methods and Data for Impact Attribution Including Recent Advances ........................................16
7 16.2.3 Observed Impacts..........................................................................................................................17
8 16.3 Synthesis of Observed Adaptation-related Responses ........................................................................25
9 16.3.1 Adaptation-related Responses by Natural Systems.......................................................................26
10 16.3.2 Adaptation-related Responses by Human Systems .......................................................................27
11 16.3.3 Knowledge Gaps in Observed Responses .....................................................................................39
12 Cross-Chapter Box INTEREG: Inter-regional Flows of Risks and Responses to Risk...........................39
13 16.4 Synthesis of Limits to Adaptation Across Natural and Human Systems ..........................................44
14 16.4.1 Definitions and Conceptual Advances Since AR5.........................................................................44
15 Box 16.1: Linking Adaptation Constraints, Soft and Hard Limits............................................................45
16 16.4.2 Insights from Regions and Sectors about Limits to Adaptation....................................................46
17 16.4.3 Regional and Sectoral Synthesis of Limits to Adaptation .............................................................48
18 16.5 Key Risks Across Sectors and Regions .................................................................................................55
19 16.5.1 Defining Key Risks ........................................................................................................................56
20 16.5.2 Identification and Assessment of Key Risks and Representative Key Risks.................................57
21 16.5.3 Variation of Key Risks Across Levels of Global Warming, Exposure and Vulnerability, and
22 Adaptation ............................................................................................................................. 75
23 16.5.4 RKR Interactions...........................................................................................................................80
24 16.6 Reasons for Concern Across Scales ......................................................................................................89
25 16.6.1 Key Risks and Sustainable Development ......................................................................................89
26 16.6.2 Framework and Approach for Assessment of RFCs and Relation to RKRs .................................93
27 16.6.3 Global Reasons for Concern.........................................................................................................98
28 Cross-Working Group Box ECONOMIC: Estimating Global Economic Impacts from Climate Change
29 ................................................................................................................................................................ 111
30 16.6.4 Summary......................................................................................................................................119
31 FAQ16.1: What are key risks in relation to climate change?...................................................................121
32 FAQ16.2 How does adaptation help to manage key risks and what are its limits?................................122
33 FAQ16.3: How do climate scientists differentiate between impacts of climate change and changes in
34 natural or human systems that occur for other reasons? .................................................................123
35 FAQ16.4: What adaptation-related responses to climate change have already been observed, and do
36 they help reduce climate risk?.............................................................................................................124
37 FAQ16.5: How does climate risk vary with temperature? .......................................................................125
38 FAQ16.6: What is the role of extreme weather events in the risks we face from climate change? ......126
39 References......................................................................................................................................................128
40
41
Do Not Cite, Quote or Distribute 16-2 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 Executive Summary
2
3 Introduction and framing
4
5 This chapter synthesizes observed climate change impacts (16.2), adaptation-related responses (16.3), limits
6 to adaptation (16.4), and the key risks identified across sectors and regions (16.5). We consider how these
7 risks accrue with increasing global average temperature; how they depend on future development and
8 adaptation efforts; and what this implies for the Sustainable Development Goals and the five main Reasons
9 for Concern about climate change (16.6).
10
11 Observed impacts
12
13 The impacts of changes in climate-related systems have been identified in a wide range of natural,
14 human, and managed systems (very high confidence1). Compared to the last IPCC AR5 there is more
15 evidence for impacts of long-term changes in climate-related systems (including the atmosphere, ocean and
16 cryosphere) on socio-economic indicators and high confidence in the sensitivity of societies to weather
17 conditions. There is also stronger evidence for impacts of long-term climate change on ecosystems, including
18 the observed widespread mortality of warm water corals, far reaching shifts in phenology in marine and
19 terrestrial ecosystems and the expansion of tropical species into the ranges of temperate species, and boreal
20 species moving into Arctic regions (high confidence). {16.2.3, 16.2.3.1}
21
22 Increased rainfall intensity associated with tropical cyclones and rising sea levels have contributed to
23 observed damages in local coastal systems (medium confidence). However, while the impact is expected
24 to be widespread, formal attribution of damages to long term changes in the climate-related systems is still
25 limited by restricted knowledge about changes in exposure and vulnerability and the missing quantification
26 of the contribution of sea level rise to the extent of flooded areas. {16.2.3.3}
27
28 Due to complex interactions with socio-economic conditions, evidence on the impact of long-term
29 climate change on crop prices and malnutrition is largely lacking while the sensitivity of malnutrition to
30 weather conditions has become more evident in some regions, particularly Africa (medium to high
31 confidence). A negative impact of long-term climate change on crop yields has been identified in some
32 regions (e.g., wheat yields in Europe) (medium confidence) while studies are still inconsistent in other
33 regions. {16.2.3.4}
34
35 Climate change has increased observed heat-related mortality (medium confidence) and contributed to
36 the observed latitudinal or altitudinal range expansion of vector- borne diseases into previously colder
37 areas (medium to high confidence) while evidence on the impact of long-term climate change on water-
38 borne diseases is largely lacking. Overall, there is extensive observational evidence that extreme ambient
39 temperatures increase human mortality (high confidence) and that the occurrence of water- and vector-borne
40 diseases is sensitive to weather conditions (high confidence). {16.2.3.5, 16.2.3.6, 16.2.3.7}
41
42 Extreme weather events not only cause substantial direct economic damage (high confidence), but also
43 reduce economic growth in the short-term (year of, and year after event) (high confidence) as well as
44 in the long-term (up to 15 years after the event) (medium confidence), with more severe impacts in
45 developing than in industrialized economies (high confidence). Evidence has increased for all of these
46 conclusions; however, evidence for impacts of long-term climate change is still limited. {16.2.3.7}
47
48 Climate variability and extremes are associated with increased prevalence of conflict, with more
49 consistent evidence for low-intensity organized violence than for major armed conflict (medium
50 confidence). Compared to other socio-economic drivers, the link is relatively weak (medium confidence) and
51 conditional on high population size, low socioeconomic development, high political marginalization, and
52 high agricultural dependence (medium confidence). Literature also suggests a larger climate-related influence
1 In this Report, the following summary terms are used to describe the available evidence: limited, medium, or robust;
and for the degree of agreement: low, medium, or high. A level of confidence is expressed using five qualifiers: very
low, low, medium, high, and very high, and typeset in italics, e.g., medium confidence. For a given evidence and
agreement statement, different confidence levels can be assigned, but increasing levels of evidence and degrees of
agreement are correlated with increasing confidence.
Do Not Cite, Quote or Distribute 16-3 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 on the dynamics of conflict than on the likelihood of initial conflict outbreak (low confidence). There is
2 insufficient evidence at present to attribute armed conflict to climate change. {16.2.3.8}
3
4 There is high confidence that anthropogenic climate forcing has had an impact on internal
5 displacement, given the observed impact of anthropogenic climate forcing on the occurrence of
6 weather extremes (high confidence, Table SM16.21) and the strong contribution of weather extremes
7 to observed displacement (high confidence). However, the link between long-term changes in the climate-
8 related systems has not been demonstrated systematically and so far there is no attribution of observed trends
9 in displacement to long-term changes in the climate-related systems. Links between weather fluctuations
10 (including extreme events) and human mobility are complex and conditional on socio-economic situations;
11 e.g., poor populations may more often be involuntarily displaced or `trapped' and not be able to migrate.
12 {16.2.3.9}
13
14 Observed adaptation in ecosystems
15
16 While species are increasingly responding to climate change, these responses may not be adaptive or
17 sufficient to cope with the rate of climate changes (high confidence). Responses have been documented in
18 a range of species, including for example changes in the timing of breeding and migration. It is unclear
19 whether these responses reflect long-term evolutionary adaptation or short-term coping mechanisms.
20 Existing assessments indicate that some species' responses will be insufficient to avert extinction. {16.3.1}
21
22 Observed adaptation-related responses in human systems
23
24 Responses across all sectors and regions reported in the scientific literature are dominated by minor
25 modifications to usual practices or measures for dealing with extreme weather events, whilst evidence
26 of transformative adaptation in human systems is low (high confidence). Responses have accelerated in
27 both developed and developing regions since AR5, with some examples of regression. Despite this, there is
28 negligible evidence in the scientific literature documenting responses that are simultaneously widespread,
29 rapid, and that challenge norms and adaptation limits. {16.3.2.3}
30
31 There is negligible evidence that existing responses are adequate to reduce climate risk (high
32 confidence). There is some evidence of global vulnerability reduction, particularly for mortality and
33 economic losses due to flood risk and extreme heat. (16.3.2.4) Evidence on the effectiveness of specific
34 adaptations remains limited. There is negligible robust evidence to assess the overall adequacy of the global
35 adaptation response to address the scale of climate risk. No studies have systematically assessed the
36 adequacy and effectiveness of adaptation at a global scale, across nations or sectors, or for different levels of
37 warming. {16.3.2.3}
38
39 Adaptation responses are showing co-benefits, for mitigation and other societal goals (high
40 confidence). There is increasing evidence of co-benefits of adaptation responses. Co-benefits are most
41 frequently linked to changes in agricultural practices (e.g., conservation agriculture), land use management
42 (e.g., agroforestry), building technologies (e.g., building efficiency standards), and urban design (e.g.,
43 walkable neighbourhoods). {16.3.2.3}
44
45 Evidence of maladaptation is increasing (high confidence), i.e. adaptation that increases climate risk or
46 creates new risks in other systems or for other actors. Globally, maladaptation has been reported most
47 frequently in the context of agriculture and migration in the global south. {16.3.2.6}
48
49 Limits to adaptation across natural and human systems
50
51 There is increasing evidence on limits to adaptation which result from the interaction of adaptation
52 constraints and can be differentiated into soft and hard limits (high confidence). Soft limits may change
53 over time as additional adaptation options become available. Hard limits will not change over time as no
54 additional adaptive actions are possible. Evidence focuses on constraints that may lead to limits at some
55 point of the adaptation process, with less information on how limits may be related to different levels of
56 socio-economic or climatic change (high confidence). {16.4.1, 16.4.2, 16.4.3}
57
Do Not Cite, Quote or Distribute 16-4 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 Limits to adaptation have been identified for terrestrial and aquatic species and ecosystems, coastal
2 communities, water security, agricultural production, and human health and heat (high confidence).
3 Beginning at 1.5°C, autonomous and evolutionary adaptation responses by terrestrial and aquatic species and
4 ecosystems face hard limits, resulting in biodiversity decline, species extinction and loss of related
5 livelihoods (high confidence). Beginning at 3°C, hard limits are projected for water management measures,
6 leading to decreased water quality and availability, negative impacts on health and wellbeing, economic
7 losses in water and energy dependent sectors and potential migration of communities (medium confidence).
8 Adaptation to address risks of heat stress, heat mortality and reduced capacities for outdoor work for humans
9 face soft and hard limits across regions beginning at 1.5°C, and are particularly relevant for regions with
10 warm climates (high confidence). {16.4.2, 16.4.3}
11
12 Soft limits are currently being experienced by individuals and households along the coast and by
13 small-scale farmers (medium confidence). As sea levels rise and extreme events intensify, coastal
14 communities face soft limits due to financial, institutional and socio-economic constraints reducing the
15 efficacy of coastal protection and accommodation approaches and resulting in loss of life and economic
16 damages (medium confidence). {16.4.2, 16.4.3}
17
18 Hard limits for coastal communities reliant on nature-based coastal protection will be experienced
19 beginning at 1.5°C (medium confidence). Soft and hard limits for agricultural production are related to
20 water availability and the uptake and effectiveness of climate-resilient crops which are constrained by socio-
21 economic and political challenges (medium confidence). {16.4.2, 16.4.3}
22
23 Across regions and sectors, the most significant determinants of soft limits are financial, governance,
24 institutional and policy constraints (high confidence). The ability of actors to overcome these socio-
25 economic constraints largely influence whether additional adaptation is able to be implemented and prevent
26 soft limits from becoming hard. While the rate, extent and timing of climate hazards largely determine hard
27 limits of biophysical systems, these factors appear to be less influential in determining soft limits for human
28 systems (medium confidence). {16.4.2, 16.4.3}
29
30 Financial constraints are important determinants of limits to adaptation, particularly in low-to-middle
31 income countries (high confidence). Impacts of climate change may increase financial constraints (high
32 confidence) and contribute to soft limits to adaptation being reached (medium confidence). Global and
33 regional evidence shows that climate impacts may limit the availability of financial resources, stunt national
34 economic growth, result in higher levels of losses and damages and thereby increase financial constraints.
35 {16.4.3.2, 16.4.3.3}
36
37 Key risks across climate and development pathways
38
39 Regional and sectoral chapters of this report identified over 130 Key Risks (KRs) that could become
40 severe under particular conditions of climate hazards, exposure, and vulnerability. These key risk are
41 represented in eight so-called Representative Key Risks (RKRs) clusters of key risks relating to low-
42 lying coastal systems; terrestrial and ocean ecosystems; critical physical infrastructure, networks and
43 services; living standards; human health; food security; water security; and peace and mobility (high
44 confidence). A key risk is defined as a potentially `severe' risk, i.e. that is relevant to the interpretation of
45 dangerous anthropogenic interference (DAI) with the climate system. Key risks cover scales from the local
46 to the global, are especially prominent in particular regions or systems, and are particularly large for
47 vulnerable subgroups, especially low-income populations, and already at-risk ecosystems (high confidence).
48 The conditions under which RKRs would become severe have been assessed along levels for warming,
49 exposure/vulnerability, and adaptation: for warming, high refers to climate outcomes consistent with RCP8.5
50 or higher, low refers to climate outcomes consistent with RCP2.6 or lower, and medium refers to
51 intermediary climate scenarios; exposure/vulnerability levels are relative to the range of future conditions
52 considered in the literature; for adaptation, high refers to near maximum potential and low refers to the
53 continuation of today's trends. (6.5.2.1, 16.5.2.2, Table SM16.4).
54
55 For most Representative Key Risks (RKRs), potentially global and systemically pervasive risks
56 become severe in the case of high warming, combined with high exposure/vulnerability, low
57 adaptation, or both (high confidence). Under these conditions there would be severe and pervasive risks to
Do Not Cite, Quote or Distribute 16-5 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 critical infrastructure and to human health from heat-related mortality (high confidence), to low-lying coastal
2 areas, aggregate economic output, and livelihoods (all medium confidence), of armed conflict (low
3 confidence), and to various aspects of food security (with different levels of confidence). Severe risks
4 interact through cascading effects, potentially causing amplification of RKRs over the course of this century
5 (low evidence, high agreement). {16.5.2.3, 16.5.2.4, 16.5.4, Figure 16.10}
6 For some RKRs, potentially global and systemically pervasive risks would become severe even with
7 medium to low warming (i.e. 1.5-2°C) if exposure/vulnerability is high and/or adaptation is low
8 (medium to high confidence). Under these conditions there would be severe and pervasive risks associated
9 with water scarcity and water-related disasters (high confidence), poverty, involuntary mobility, and insular
10 ecosystems and biodiversity hotspots (all medium confidence). {16.5.2.3, 16.5.2.4}
11
12 All potentially severe risks that apply to particular sectors or groups of people at more specific
13 regional and local levels require high exposure/vulnerability or low adaptation (or both), but do not
14 necessarily require high warming (high confidence). Under these conditions there would be severe,
15 specific risks to low-lying coastal systems, to people and economies from critical infrastructure disruption,
16 economic output in developing countries, livelihoods in climate-sensitive sectors, waterborne diseases
17 especially in children in low- and middle-income countries, water-related impacts on traditional ways of life,
18 and involuntary mobility for example in small islands and low-lying coastal areas (medium to high
19 confidence). {16.5.2.3, 16.5.2.4}
20
21 Some severe impacts are already occurring (high confidence) and will occur in many more systems
22 before mid-century (medium confidence). Tropical and polar low-lying coastal human communities are
23 experiencing severe impacts today (high confidence), and abrupt ecological changes resulting from mass
24 population-level mortality are already observed following climate extreme events. Some systems will
25 experience severe risks before the end of the century (medium confidence), for example critical infrastructure
26 affected by extreme events (medium confidence). Food security for millions of people, particularly low-
27 income populations, also faces significant risks with moderate to high warming or high vulnerability, with a
28 growing challenge by 2050 in terms of providing nutritious and affordable diets (high confidence). {16.5.2.3,
29 16.5.3}
30
31 In specific systems already marked by high exposure and vulnerability, high adaptation efforts will
32 not be sufficient to prevent severe risks from occurring under high warming (low evidence, medium
33 agreement). This is particularly the case for some ecosystems and water-related risks (from water scarcity
34 and to indigenous and traditional cultures and ways of life). {16.5.2.3, 16.5.2.4, 16.5.3}
35
36 Interconnectedness and globalization establish pathways for the transmission of climate-related risks
37 across sectors and borders, for instance through trade, finance, food, and ecosystems (high
38 confidence). Examples include semiconductors, global investments, major food crops like wheat, maize and
39 soybean, and transboundary fish stocks. There are knowledge gaps on the need for, effectiveness of, and
40 limits to adaptation to such interregional risks {Cross-Chapter Box INTERREG in this Chapter}
41
42 Key risks increase the challenges in achieving global sustainability goals (high confidence). The greatest
43 challenges will be from risks to water (RKR-G), living standards (RKR-D), coastal socio-ecological systems
44 (RKR-A) and peace and human mobility (RKR-H). The most relevant goals are Zero hunger (SDG2),
45 Sustainable cities and communities (SDG11), Life below water (SDG14), Decent work and economic
46 growth (SDG8), and No poverty (SDG1). Priority areas for regions are indicated by the intersection of
47 hazards, risks and challenges, where, in the near term, challenges to SDGs indicate probable systemic
48 vulnerabilities and issues in responding to climatic hazards. (high confidence) {16.6.1}
49
50 The scale and nature of climate risks is partly determined by the responses to climate change, not only in
51 how they reduce risk, but also how they may create other risks (sometimes inadvertently, and sometimes to
52 others than those who implement the response, in other places, or later in time).
53
54 Solar Radiation Modification (SRM) approaches have potential to offset warming and ameliorate
55 other climate hazards, but their potential to reduce risk or introduce novel risks to people and
56 ecosystems is not well understood (high confidence). SRM effects on climate hazards are highly dependent
57 on deployment scenarios and substantial residual climate change or overcompensating change would occur
Do Not Cite, Quote or Distribute 16-6 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 at regional scales and seasonal timescales (high confidence). Due in part to limited research, there is low
2 confidence in projected benefits or risks to crop yields, economies, human health, or ecosystems. Large
3 negative impacts are projected from rapid warming for a sudden and sustained termination of SRM in a high-
4 CO2 scenario. SRM would not stop CO2 from increasing in the atmosphere or reduce resulting ocean
5 acidification under continued anthropogenic emissions (high confidence). There is high agreement in the
6 literature that for addressing climate change risks SRM is, at best, a supplement to achieving sustained net
7 zero or net negative CO2 emission levels globally. Co-evolution of SRM governance and research provides a
8 chance for responsibly developing SRM technologies with broader public participation and political
9 legitimacy, guarding against potential risks and harms relevant across a full range of scenarios. [Cross-
10 Working Group Box SRM}
11
12 Recent global estimates of the economic cost of climate impacts exhibit significant spread and
13 generally increase with global average temperature, as well as vary by other drivers, such as income,
14 population and composition of the economy (high confidence). The wide variation across disparate
15 methodologies does not allow a robust range of damage estimates to be identified with confidence, though
16 the spread of estimates increases with warming in all methodologies, indicating higher risk (in terms of
17 economic costs) at higher temperatures (high confidence). Reconciling methodological variance is a priority
18 for facilitating use of different lines of evidence; however, that some new estimates are higher than the AR5
19 range indicates that global aggregate economic impacts could be higher than previously assessed (low
20 confidence due to the lack of robustness and comparability across methodologies). {Cross-Working Group
21 Box ECONOMIC in Chapter 16}
22
23 Reasons for Concern across scales
24
25 The five major Reasons for Concern (RFCs), describing risks associated with (1) unique and threatened
26 systems, (2) extreme weather events, (3) distribution of impacts, (4) global aggregate impacts, and (5) large-
27 scale singular events, were updated using expert elicitation. RFC risk levels were assessed with no or low
28 adaptation, but limits to adaptation are a factor in the identification of very high risk levels.
29
30 Compared to AR5 and SR15, risks increase to high and very high levels at lower global warming levels
31 for all five RFCs (high confidence), and transition ranges are assigned with greater confidence.
32 Transitions from high to very high risk emerge in all five RFCs, compared to just two RFCs in AR5
33 (high confidence). {16.6.3, Figure 16.15}
34
35 For unique and threatened systems (RFC1), as before, levels of risk at a given level of warming are
36 higher than for the other RFCs. Risks are already (at current warming of 1.1ºC) in the transition from
37 moderate to high (very high confidence), compared to moderate in AR5 and SR15, based on observed and
38 modelled impacts. The transition to very high risk occurs between 1.2ºC and 2.0ºC warming (high
39 confidence). {16.6.3.1}
40
41 For risks from extremes (RFC2), the transition to high risk is between 1.0ºC and 1.5ºC (high
42 confidence) and to very high risk (new in AR6) between 1.8 and 2.5ºC (medium confidence). {16.6.3.2}
43
44 For risks disproportionately affecting particularly vulnerable societies and socio-ecological systems,
45 including disadvantaged people and communities in countries at all levels of development (RFC3), current
46 risk is moderate (high confidence) and the transition to high risk is between 1.52.0ºC warming (medium
47 confidence). The transition to very high risk occurs at between 2.03.5ºC warming (medium confidence).
48 {16.6.3.3}
49
50 The risk of global aggregate impacts, including monetary damages, lives affected, species lost or
51 ecosystem degradation at a global scale (RFC4), has begun to transition to moderate risk (medium
52 confidence), with a transition to high risk between 1.52.5ºC (medium confidence) and to very high risk (new
53 in AR6) at between 2.5 and 4.5ºC (low confidence). {16.6.3.4}
54
55 Present-day risks associated with large-scale singular events (sometimes called tipping points or
56 critical thresholds) (RFC5) are already moderate (high confidence), with a transition to high risk between
Do Not Cite, Quote or Distribute 16-7 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 1.52.5ºC (medium confidence) and to very high risk (new in AR6) between 2.54ºC (low confidence).
2 {16.6.3.5}
3
4 Limiting global warming to 1.5ºC would ensure risk levels remain moderate for RFC3, RFC4 and
5 RFC5 (medium confidence) but risk for RFC2 would have transitioned to a high risk at 1.5ºC and
6 RFC1 would be well into the transition to very high risk (high confidence). Remaining below 2ºC
7 warming (but above 1.5ºC) would imply that risk for RFC3 through 5 would be transitioning to high,
8 and risk for RFC1 and RFC2 would be transitioning to very high (high confidence). By 2.5ºC warming,
9 RFC1 will be in very high risk (high confidence) and all other RFCs will have begun their transitions to very
10 high risk (medium confidence for RFC2 and RFC3, low confidence for RFC4 and RFC5).
11
12 RFC1, RFC2 and RFC5 include risks that are irreversible, such as species extinction, coral reef degradation,
13 loss of cultural heritage, or loss of a small island due to sea level rise. Once such risks materialise, as is
14 expected at very high risk levels, the impacts would persist even if global temperatures would subsequently
15 decline to levels associated with lower levels of risk in an `overshooting' scenario (high confidence).
16 {16.6.3}
17
18
Do Not Cite, Quote or Distribute 16-8 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 16.1 Introduction and Framing
2
3 16.1.1 Objective of the Chapter
4
5 Anthropogenic climate change poses risks to many human and ecological systems. These risks are
6 increasingly visible in our day-to-day lives, including a growing number of disasters that already bear a
7 fingerprint of climate change. There is increasing concern about how these risks will shape the future of our
8 planet our ecosystems, our well-being and development opportunities. Policy makers are asking what is
9 known about the risks, and what can be done about them. Many people and especially youth around the
10 world are calling for urgency, ambition and action. Companies are wondering how to manage new threats to
11 their bottom line, or how to grasp new opportunities. On top of this growing concern about climate change,
12 the COVID-19 pandemic has exposed vulnerabilities to shocks, significantly aggravated climate-related
13 risks, and posed new questions about how to achieve a green, resilient and inclusive recovery (see Cross-
14 Chapter Box COVID in Chapter 7).
15
16 The three synthesis chapters of this report (16, 17 and 18) aim to address these concerns. They synthesize
17 information from across all thematic and regional Chapters of the Working Group (WGII) Sixth Assessment
18 Report (AR6) and the recent IPCC Special Reports on Global Warming of 1.5ºC, on Climate Change and
19 Land, and on Ocean and Cryosphere in a Changing Climate (SR15, SRCCL and SROCC), but also include
20 an independent assessment of the literature, especially literature that cuts across sectors and regions.
21
22 Chapter 16 lays the groundwork by synthesizing the state of knowledge on the observed impacts of climate
23 change (Section 16.2) and ongoing adaptation responses (Section 16.3), the limits to adaptation (Section
24 16.4), and the key risks we should be concerned about, how these risks evolve with global temperature
25 change, and also how they depend on future development and adaptation efforts (Sections 16.5 and 16.6). It
26 thus brings together elements that were assessed in different chapters in previous assessments, especially the
27 Third, Fourth and Fifth Assessment Reports (TAR, AR4, and AR5, respectively). Background on specific
28 methodological aspects of this chapter is provided in Supplementary Material..
29
30 The strong link between risks, adaptation and development connects this chapter closely to Chapters 17 and
31 18. Chapter 17 assesses decision-making: what do we know about the ways to manage risks in a warming
32 climate (including in the context of the key risks and limits to adaptation identified in this chapter)? Chapter
33 18 puts all of this information into the perspective of climate-resilient development pathways: how can we
34 achieve sustainable development given the additional challenges posed by climate change?
35
36 16.1.2 Risk Framing
37
38 In the IPCC AR6, `risk' is defined as the potential for adverse consequences for human or ecological
39 systems, recognizing the diversity of values and objectives associated with such systems. Relevant adverse
40 consequences include those on lives, livelihoods, health and well-being, economic, social and cultural assets
41 and investments, infrastructure, services (including ecosystem services), ecosystems and species (Chapter 1
42 this volume, SR15). The AR6 definition explicitly notes that `risks can arise from potential impacts of
43 climate change as well as human responses to climate change.'
44
45 The main risks assessed here relate to the potential impacts of climate change. In recent years, the growing
46 visibility of current climate impacts has resulted in a stronger focus on understanding and managing such
47 risk across timescales, rather than just for the longer-term future. Examples include the rapid growth in
48 attribution of specific extreme weather events, the use of scientific evidence of climate change impacts in
49 legal cases, the context of the Paris Agreement's Article 8 on `averting, minimizing and addressing loss and
50 damage' associated with climate change, but also the stronger links between adaptation and disaster risk
51 reduction, including early warning systems, wider discussions on how to build resilience in the face of a
52 more volatile climate, and attention for limits to adaptation that are already being reached.
53
54 Of course the scale of these risks is also determined by the responses to climate change, mainly in how they
55 reduce risk, but also how they may create risks (sometimes inadvertently, and sometimes to others than those
56 who implement the response, in other places, or later in time). Our focus is on adaptation responses, given
57 that mitigation is covered in WGIII AR6, but we acknowledge certain important interactions, such as
Do Not Cite, Quote or Distribute 16-9 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 biomass-production as an alternative to fossil fuels which can compete with food production and thus
2 aggravate adaptation challenges. Given that solar radiation modification (SRM) could also be considered a
3 response with significant implications for climate risks across scales, this chapter also includes Cross-
4 Working Group Box SRM.
5
6 This assessment focuses primarily on adverse consequences of climate change. However, climate change
7 also has positive implications (benefits and opportunities) for certain people and systems, although there are
8 gaps in the literature on these positive effects. Some risks assessed in this chapter are actually about a
9 balance between positive and negative effects of climate change (and of response options, especially
10 adaptation). In those contexts, we assess the combined effect of both, aiming to identify not only the
11 aggregate impacts (the balance between positive and negative effects) but also the distributional aspects
12 (winners and losers). A more comprehensive discussion of the decision-making related to such trade-offs in
13 relation to adaptation is provided in Chapter 17.
14
15 This chapter's assessment takes a global perspective, although many risks and responses materialise at the
16 local or national scale. We use case studies to illustrate the ways these risks aggregate across scales, again
17 with particular concern for distributional aspects.
18
19 16.1.3 Storyline of the Chapter, and What's New Compared to Previous Assessments
20
21 Figure 16.1 illustrates the elements covered by the chapter, which can be summarised as four key questions.
22
23
24
25 Figure 16.1. Illustrative storyline of the chapter highlighting the central questions addressed in the various sections,
26 going from realized risks (observed impacts) to future risks (key risks and reasons for concern), informed by adaptation-
27 related responses and the limits to adaptation. The pink arrows illustrate actions to reduce hazard, exposure and
28 vulnerability, which shape risks over time. Accordingly, the green areas at the centre of the propeller diagrams indicate
29 the ability for such solutions to reduce risk, up to certain adaptation limits, leaving the white residual risk (or observed
30 impacts) in the centre. The shading of the right-hand side propeller diagram compared to the non-shaded one on the left
31 reflects some degree of uncertainty about future risks. The figure builds on the conceptual framework of risk-
32 adaptation-relationships used in SROCC (Garschagen et al., 2019).
Do Not Cite, Quote or Distribute 16-10 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1
2
3 16.1.3.1 What Impacts are Being Experienced?
4
5 This assessment of climate related impacts that are already taking place is covered in Section 16.2, which
6 aims to differentiate between observed changes in climate hazards (also called `climate impact drivers' in
7 IPCC Working Group I) and the exposure and vulnerability of human and ecological systems.
8
9 Observed impacts of climate change were synthesized in the TAR, AR4 and AR5. The TAR found that
10 recent regional climate changes had already affected many physical and biological systems, with preliminary
11 indications that some human systems had been affected, primarily through floods and droughts. AR4 found
12 likely2 discernible impacts on many physical and biological systems, and more limited evidence for impacts
13 on human environments. AR5 devoted a separate chapter to observed impacts, which found growing
14 evidence of impacts on human and ecological systems on all continents and across oceans (Cramer et al.,
15 2014).
16
17 Section 16.2 reports on the expanded literature since then, generally reflecting a growing and more certain
18 impact of climate change on humans and ecological systems.
19
20 16.1.3.2 What Responses are Being Undertaken?
21
22 Section 16.3 provides, for the first time, a comprehensive synthesis of observed adaptation-related responses
23 to the rising risks.
24
25 Such adaptation responses were first covered in the TAR, and further developed in the AR4 and AR5. For
26 instance, AR5 Chapter 15 notes that adaptation to climate change was transitioning from a phase of
27 awareness to the construction of actual strategies and plans in societies (Mimura et al., 2014) but did not
28 include a comprehensive mapping of responses.
29
30 Based on such a comprehensive mapping, Section 16.3 finds growing evidence of adaptation-related
31 responses, although these are dominated by minor modifications to usual practices or measures for dealing
32 with extreme weather events, and there is limited evidence for the extent to which they reduce climate risk.
33
34 16.1.3.3 What are the Limits to Adaptation?
35
36 The literature on limits to adaptation, which is covered in Section 16.4, has strongly evolved since AR5,
37 including links to discussions on Loss and Damage in the UNFCCC. While the SPM of AR4 noted that there
38 was no clear picture of the limits to adaptation, or the cost, AR5 Chapter 16 (Klein et al., 2014) reported
39 increasing insights emerging from the interactions between climate change and biophysical and
40 socioeconomic constraints, and highlighted the fact that limits could be both hard and soft. It also noted that
41 residual losses and damages will occur from climate change despite adaptation and mitigation action.
42 However, AR5 Chapter 16 still found that the empirical evidence needed to identify limits to adaptation of
43 specific sectors, regions, ecosystems, or species that can be avoided with different GHG mitigation pathways
44 was lacking.
45
46 Section 16.4 provides a more comprehensive assessment of limits to adaptation, highlighting again that
47 limits to adaptation are not fixed, but are properties of dynamic socio-ecological systems. They are shaped
48 not only by the magnitude of the climate hazards (e.g., the amount of sea level rise in low lying coasts and
49 islands), and the exposure and vulnerability to those hazards (e.g., people and assets in those areas), but also
50 by physical, infrastructural and social tolerance thresholds and adaptation choices of actors in societies (e.g.,
51 the decision to migrate from locations strongly impacted by climate change). The evolution of such socio-
2 In this Report, the following terms have been used to indicate the assessed likelihood of an outcome or a result:
Virtually certain 99100% probability, Very likely 90100%, Likely 66100%, About as likely as not 3366%,
Unlikely 033%, Very unlikely 010%, and Exceptionally unlikely 01%. Additional terms (Extremely likely: 95
100%, More likely than not >50100%, and Extremely unlikely 05%) may also be used when appropriate. Assessed
likelihood is typeset in italics, e.g., very likely). This Report also uses the term `likely range' to indicate that the assessed
likelihood of an outcome lies within the 17-83% probability range.
Do Not Cite, Quote or Distribute 16-11 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 economic systems over time, including their interaction with the changing physical climate, determines the
2 evolution of limits to adaptation.
3
4 16.1.3.4 What Future Risks are of Greatest Concern?
5
6 The fourth and final element of the chapter is the question about the risks we face, and which ones we should
7 be most concerned about. This is addressed in Section 16.5 and 16.6.
8
9 Section 16.5.1 presents a full discussion of `key risks', synthesized from across all chapters, defined as those
10 risks that are potentially severe and therefore especially relevant to the interpretation of `dangerous
11 anthropogenic interference with the climate system' in the terminology of UNFCCC Article 2.
12
13 In 2015 the Paris Agreement established the goal of `holding the increase in the global average temperature
14 to well below 2°C above pre-industrial levels and pursuing efforts to limit the temperature increase to 1.5°C
15 above pre-industrial levels'. However, assessment of key risks across a range of future warming levels
16 remains a high priority for several reasons: (1) understanding risks at higher levels of warming can help
17 prepare for them, should efforts to limit warming be unsuccessful (UNEP, 2017); (2) understanding risks at
18 higher levels can inform the benefits of limiting warming to lower levels; (3) in addition, there is continued
19 debate about whether warming limits should be at or rather somewhere below 2ºC (in particular at 1.5ºC);
20 and (4) there is a more explicit recognition that key risks can result not only from increased warming, but
21 also from changes in the exposure and vulnerability of society, and from a lack of ambitious adaptation
22 efforts. So relatively limited warming does not automatically imply that key risks will not occur. In assessing
23 key risks, we have applied four criteria: magnitude of adverse consequences, likelihood of adverse
24 consequences, temporal characteristics of the risk, and ability to respond. Of course, this is an aggregated
25 approach to what is dangerous; it should be noted that in practice, `dangerous' will occur at a myriad of
26 temperature levels depending on who or what is at risk (and their circumstances), geographic scale and time
27 scale.
28
29 A new element is that we particularly look at a set of eight `representative key risks' that exemplify the
30 underlying set of key risks identified in the earlier chapters: risk to the integrity of low-lying coastal socio-
31 ecological systems, risk to terrestrial and ocean ecosystems, risk to critical physical infrastructure and
32 networks, risk to living standards (including economic impacts, poverty and inequality), risk to human
33 health, risk to food security, risk to water security, and risk to peace and mobility (Section 16.5.2.3).
34 Another increased focus relates to the issue of compound risks. This includes risks associated with
35 compound hazards (WGI AR6 Chapter 11, Seneviratne et al., 2021), but also implications for future risk
36 when repeated impacts erode vulnerability, as well as through transboundary effects (including effects both
37 from one system to a neighbouring one, as well as from one system to a distant one), also discussed in the
38 cross-chapter box on interregional risks and adaptation (Cross-Chapter Box INTEREG in this Chapter).
39
40 Section 16.6 maps the representative key risks in Section 16.5 to the Sustainable Development Goals, noting
41 both direct and indirect implications for Climate Resilient Development as assessed in Chapter 18.
42
43 Finally, section 16.6 presents an updated assessment of the so-called `Reasons for Concern' (RFC): risks
44 related to unique and threatened systems, extreme events, distribution of impacts, aggregate impacts
45 (including the cross-chapter box on the global economic impacts of climate change and the social cost of
46 carbon, Cross-Working Group Box ECONOMIC) and the risk of irreversible and abrupt transitions.
47
48 The AR4 and AR5 each also evaluated the most important climate risks, framed firstly in terms of the state
49 of knowledge relevant to Article 2 of the UNFCCC. The TAR first synthesized this knowledge in five RFCs.
50 AR4 identified a set of `key vulnerabilities', and provided an update of the RFCs. AR5 further refined a new
51 risk framework developed in SREX, and used it to assess `key risks' and provide another update of the
52 overarching Reasons for Concern, drawing as well on Cramer et al. (2014) assessment of observed changes.
53
54 Our risk assessment also further builds on risk assessments from the Special Reports that are part of the AR6
55 cycle, i.e. SR15; SRCCL, and SROCC. While since AR4 the RFC assessment framework has remained
56 largely consistent, refinements in methodology have included the consideration of different risks, the role of
57 adaptation, use of confidence statements, more formalized protocols and standardized metrics (Zommers et
Do Not Cite, Quote or Distribute 16-12 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 al., 2020). In subsequent assessment cycles, the risk level at a given temperature has generally increased,
2 reflecting accumulating scientific evidence (Zommers et al., 2020).
3
4 16.1.4 Drivers of Exposure and Vulnerability
5
6 While this chapter focuses on climate-related impacts, risks and responses, these all take place against a
7 backdrop of trends in exposure and vulnerability driven by demographics, socio-economic development
8 (including inequalities) and ecosystem degradation. Other global trends that are shaping climate risks include
9 technological innovation, shifts in global power relations, and resource scarcity (Retief et al., 2016). Note
10 that these global trends may increase but also reduce exposure and/or vulnerability, for instance when
11 growing incomes, savings and social protection systems increase resilience in the face of shocks and stresses.
12 Drivers and future trends in vulnerability and exposure next to climate-induced changes in natural hazards
13 therefore need to be considered in comprehensive risk assessments and eventually adaptation solutions, but
14 empirical research suggests that they remain to be underemphasized in current national adaptation planning
15 (Garschagen et al., 2021a).
16
17 While these risk drivers are often listed separately, they are often closely interconnected, including between
18 human and ecological systems, and increasingly also through climate risks and responses (e.g., Simpson et
19 al., 2021). Climate impacts increasingly affect these drivers, and may compete with financial resources that
20 could otherwise be applied for development, mitigation, adaptation and resilience building, also affecting
21 inequalities (e.g., Taconet et al., 2020).
22
23 16.1.4.1 Demographics
24
25 Population growth (or decline) can result in increasing (or decreasing) pressure on natural resources (e.g.,
26 soils, water and fish stocks) (IPBES, 2019), and can result in the expansion of densely populated areas
27 (Cardona et al., 2012; Day et al., 2016). A majority of the population in the coming decades will be in urban
28 areas. While urbanization can have many benefits that reduce vulnerability, such as employment
29 opportunities and increased income, better access to healthcare and education, and improved infrastructure,
30 unsustainable urbanisation patterns can create challenges for resource availability, exacerbate pollution
31 levels (Rode et al., 2015), and increase exposure to some risks. For example, ~10% of the global population
32 live in Low Elevation Coastal Zones (in 2000; areas <10 m of elevation) (McGranahan et al., 2007;
33 Neumann et al., 2015), which is expected to increase by 5% to 13.6% by 2100 depending on the population
34 scenario (Neumann et al., 2015; Jones and O'Neill, 2016). Building assets and infrastructure in naturally
35 risk-prone areas are also projected to increase (Magnan et al., 2019), which may also lead to environmental
36 degradation that can further aggravate risk, e.g., destruction of wetlands that buffer against floods (Schuerch
37 et al., 2018; Oppenheimer et al., 2019). Demographic trends, coupled with changes in income, can also result
38 in increasing demands for land, food, water and energy, and therefore to major changes in land use and cover
39 change (Arneth, 2019). The observed and projected population decline in some rural areas also has
40 implications for vulnerability and exposure. In addition, demographic changes such as aging may increase
41 vulnerability to some climate hazards, including heat stress (Byers et al., 2018; Rohat et al., 2019a; Rohat et
42 al., 2019b).
43
44 16.1.4.2 Biodiversity and Ecosystems
45
46 Rapidly accelerating trends in human impacts on global ecosystems and biodiversity, especially in the past 5
47 decades, have resulted precipitous declines in the numbers of many wild species on land and in the ocean,
48 transformation of the terrestrial land surface for agricultural production, and the pervasive spread of alien
49 and invasive species (IPBES, 2019). As a result, the capacity of ecosystems to support human society is
50 thought to be coming under threat. For instance, the fraction of all primary production being appropriated for
51 human use has doubled over the course of the 20th Century (to about 25% in 2005), although it has grown at
52 a slower rate than human population (Krausmann et al., 2013). Future projections significantly depend on
53 bioenergy production, signalling one of the feedbacks between responses to climate change and climate
54 risks.
55
56 16.1.4.3 Poverty Trends and Socioeconomic Inequalities Within and Across Societies
57
Do Not Cite, Quote or Distribute 16-13 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 Poverty contributes to exposure and vulnerability by limiting access of individuals, households and
2 communities to economic resources and restraining adaptive capacities (e.g., for food and energy supply, or
3 for financing adaptation responses) (Hallegatte and Rozenberg, 2017). Over the past decades, until the
4 COVID-19 pandemic, global poverty rates have declined rapidly. Between 1981 and 2015, the share of
5 global population living in extreme poverty (under the international poverty line of US$1.90 per day)
6 declined from 42% to 10%, leaving 736 million people in extreme poverty, concentrated in South Asia and
7 Sub-Saharan Africa (World Bank, 2018). This general reduction in poverty across the world is accompanied
8 by a decrease in vulnerability to many types of climate change impacts (medium confidence). However, the
9 COVID-19 pandemic has significantly increased extreme poverty by about 100 million people in 2020, with
10 disproportionate economic impacts on the poorest, most fragile and smaller countries (World Bank, 2021)
11 and significant implications for vulnerability to climate change (see also Cross-Chapter Box COVID in
12 Chapter 7).
13
14 The majority of the population in poverty are smallholder farmers and pastoralists, whose livelihoods
15 critically depend on climate-sensitive natural ecosystems, e.g., through semi-subsistence agriculture where
16 food consumption is primarily dependent on households' own food production (Mbow et al., 2019). A
17 significant share of this population is affected by armed conflict, which deters economic development and
18 growth and increases local dependence on subsistence agriculture (Serneels and Verpoorten, 2015;
19 Braithwaite et al., 2016; Tollefsen, 2017), and aggravating humanitarian challenges (e.g., ICRC, 2020).
20 Extreme weather events, particularly droughts, can result in poverty traps keeping people poor or making
21 them poorer, resulting in widening inequalities within and across countries.
22
23 Climate risks are also strongly related to other inequalities, often but not always intersecting with poverty.
24 AR5 found with very high confidence that differences in vulnerability and exposure arise from
25 multidimensional inequalities, often produced by uneven development processes. These inequalities relate to
26 geographic location, as well as economic, political and socio-cultural aspects, such as wealth, education,
27 race/ethnicity, religion, gender, age, class/caste, disability, and health status (Oppenheimer et al., 2014).
28 Since AR5, a number of studies have confirmed and refined this assessment, especially also regarding socio-
29 economic inequality and poverty (Hallegatte et al., 2016; Hallegatte and Rozenberg, 2017; Pelling and
30 Garschagen, 2019; Hallegatte et al., 2020). Poor people more often live in exposed areas such as wastelands
31 or riverbanks (Garschagen and Romero-Lankao, 2015; Winsemius et al., 2018). Also, poor people lose more
32 of their total wealth to climatic hazards, receive less post-shock support from their often-times equally poor
33 social networks, and are often not covered by social protection schemes (Leichenko and Silva, 2014;
34 Hallegatte et al., 2016). Countries with high inequality tend to have above-average levels of exposure and
35 vulnerability to climate hazards (BEH UNU-EHS, 2016). Many socio-economic models used in climate
36 research have been found to have a limited ability to capture and represent the poor at a larger scale (Rao et
37 al., 2019; Rufat et al., 2019). However, an analysis of 92 countries found that relative income losses and
38 other climate change impacts were disproportionately high among the poorest (Hallegatte and Rozenberg,
39 2017, see Section 16.2.6). There have also been advances in detecting and attributing the impacts of climate
40 change and vulnerability at household scale and specifically on women's agency and adaptive capacity (Rao
41 et al., 2019). The distribution of impacts and responses (adaptation and mitigation) affects inequality, not just
42 between countries, but also within countries (e.g., Tol, 2020) and between different people within societies.
43 Distribution has so far largely been thought of in a geographical sense, but identifying those most at risk
44 requires an additional focus on the social distribution of impacts, responses, as well as of resilience, as
45 influenced for instance by differential social protection coverage (Tenzing, 2020).
46
47 Many climate responses interact with all of these global risk drivers. Some raise additional equity concerns
48 about marginalising those most vulnerable and exacerbating social conflicts (Oppenheimer et al., 2019),
49 leading to wider questions about the governance of climate risks (and impacts) across scales. Hence, our
50 assessment of impacts, responses, and risks is complemented by the assessment of governance and the
51 enabling environment for risk management in Chapter 17, and of climate-resilient development in Chapter
52 18.
53
54
55 16.2 Synthesis of Observed Impacts
56
Do Not Cite, Quote or Distribute 16-14 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 This section synthesizes the observed impacts of changes in climate-related systems (see Section 16.2.1) on
2 different natural, human, and managed systems (outlined in Chapters 2-8) and regions (outlined in Chapters
3 9-15). To stay as specific as possible given the required level of aggregation, we decided in favour of a
4 summary along specific prominent indicators such as `crop yields' or `areas burned by wildfires' instead of
5 an assessment across broad categories such as `food production' which could include a broad range of
6 measures ranging from climate induced changes in growing seasons to growing seasons to impacts on
7 livestock and fisheries etc. or `wildfires' which could also cover impacts on the frequency, intensity, timing,
8 or emissions and health impacts of wildfires. However, this decision for specificity certainly implies a
9 decision against comprehensiveness. In addition, the level of specificity has to be adjusted given the
10 literature basis which is quite broad regarding crop yields but still limited and less harmonized regarding
11 indicators when it comes to e.g., conflicts. A broader discussion can be found in the sectoral or regional
12 chapters that all cover `observed impacts' individually. Section 16.2.1 provides key definitions, followed by
13 recent advances in available methods and data for climate impact attribution (Section 16.2.2), and the
14 assessment of observed impacts (Section 16.2.3). It is important to note that the assessment is primarily
15 based on peer-reviewed literature, i.e. it is limited to the regions and phenomena for which such studies are
16 available. So `no assessment' in a certain region does not apply the considered type of impact did not occur
17 in this region.
18
19 16.2.1 Definitions
20
21 The section adopts the general definition of detection as `demonstration that a considered system has
22 changed without providing reasons for the change' and attribution as identifying the causes of the observed
23 change or a specific event (see Glossary).
24
25 Based on these general definitions and following the approach applied in WGII AR5 Chapter 18 (Cramer et
26 al., 2014), we define an observed impact as the difference between the observed state of a natural, human,
27 or managed system and a counterfactual baseline that characterizes the system's state in the absence of
28 changes in the climate-related systems defined here as climate system including the ocean and the
29 cryosphere as physical or chemical systems.
30
31 The difference between the observed and the counterfactual baseline state is considered the change in the
32 natural, human, or managed system that is attributed to the changes in the climate-related systems (impact
33 attribution).The counterfactual baseline may be stationary or may change over time, for example due to
34 direct human influences such as changes in land use patterns, agricultural or water management affecting
35 exposure and vulnerability to climate related hazards (see Section 16.2.3 for methods on how to construct the
36 counterfactual).
37
38 In line with the AR5 definition, `changes in climate-related systems' here refer to any long-term trend,
39 irrespective of the underlying causes; thus, an observed impact is not necessarily an observed impact of
40 anthropogenic climate forcing. For example, in this section sea level rise is defined as relative sea level rise
41 measured against a land-based reference frame (tide gauge measurements), meaning that it is driven not only
42 by thermal expansion and loss of land ice influenced by anthropogenic climate forcing, but also by vertical
43 land movements. As attribution of coastal damages to sea level rise does not distinguish between these
44 components it does not imply attribution to anthropogenic forcing. Where the literature does allow
45 attribution of changes in natural, human or managed systems to anthropogenic climate forcing (`joint
46 attribution', Rosenzweig et al., 2007), this is highlighted in the assessment. Often the attribution of changes
47 in the natural, human or managed systems to anthropogenic forcing can be done in a two-step approach
48 where i) an observed change in a climate-related system is attributed to anthropogenic climate forcing
49 (`climate attribution') and ii) changes in natural, human, or managed systems are attributed to this change in
50 the climate-related system (`impact attribution').
51
52 For climate attribution the main challenge is the separation of externally human forced changes in the
53 climate-related systems from their internal variability while for impact attribution it often is the separation of
54 the effects of other external forcings (i.e., direct human influences or natural disturbances) from the impacts
55 of the changes in the climate-related systems. Direct influences not related to changes in the climate-related
56 systems could e.g., be pollution and land use changes amplifying biodiversity losses, intensification of
57 fishing reducing fish stocks, and increasing protection reducing losses due to river floods. The direct human
Do Not Cite, Quote or Distribute 16-15 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 or natural influences may counter the impacts of climate change (e.g., climate change may have reduced
2 flood hazards but exposure may have increased as people have moved to flood-prone areas, resulting in no
3 change in observed damages). Given the definition of impact attribution, that means that there may be an
4 observed impact of climate change without the detection of a change in the natural, human or managed
5 system. This is different from `climate attribution' where detection and attribution are consecutive steps.
6
7 Changes in climate related systems can certainly also affect natural, human and managed systems through
8 indirect effects on land use, pollution or exposure. However, these indirect effects are barely addressed in
9 existing studies.
10
11 In addition to impact attribution, there is research on the identification of natural, human, or managed
12 systems' response to short-term (typically daily, monthly or annual) weather fluctuations or individual
13 extreme weather events. As different from impact attribution we separately define:
14
15 `Identification of weather sensitivity' refers to the attribution of the response of a system to fluctuations in
16 weather and short-term changes in the climate-related systems including individual extreme weather events
17 (e.g., a heatwave or storm surge).
18
19 Typical questions addressed include: `How much of the observed variability of crop yields is due to
20 variations in weather conditions compared to contributions from management changes?' (e.g., Ray et al.,
21 2015; Müller et al., 2017) and `Can weather fluctuations explain part of the observed variability in annual
22 national economic growth rates?' (e.g., Burke et al., 2015). Identification of weather sensitivity may also
23 address the effects of individual climate extremes, for example asking, `Was the observed outbreak of
24 cholera triggered by an associated flood event?' (e.g., Rinaldo et al., 2012; Moore et al., 2017b). It is
25 important to note that sensitivity could be described in diverse ways and that for example the fraction of the
26 observed variability in a system explained by weather variability differs from the strength of the systems'
27 response to a specific change in a weather variable. Nevertheless, all these different measures are integrated
28 in the `identification of weather sensitivity' assessment where `sensitivity' should not be considered a
29 quantitative one dimensional mathematical measure.
30
31 In this chapter we explicitly distinguish between assessment statements related to `climate attribution' (listed
32 in Table SM16.21), `impact attribution' (listed in Table SM16.22), and `identification of weather sensitivity'
33 (listed in Table SM16.23). The identification of `weather sensitivity' does not necessarily imply that there
34 also is an impact of long-term climate change on the considered system. However, if the probability or
35 intensity of an extreme weather event has increased due to anthropogenic forcing (`climate attribution')
36 (NASEM, 2016; WGI AR6 Chapter 11 Seneviratne et al., 2021) and the event is also identified as an
37 important driver of an observed fluctuation in a natural, human or managed system (`identification of
38 weather sensitivity'), then the observed fluctuation is considered (partly) attributed to long-term climate
39 change (`impact attribution') and even to anthropogenic forcing.
40
41 16.2.2 Methods and Data for Impact Attribution Including Recent Advances
42
43 By definition the counterfactual baseline required for impact attribution cannot be observed. However, it
44 may be approximated by impact model simulations forced by a stationary climate e.g. derived by de-trending
45 the observed climate (Diffenbaugh et al., 2017; Mengel et al., 2021) while other relevant drivers (e.g., land
46 use changes or application of pesticides) of changes in the system of interest (e.g., a bird population) evolve
47 according to historical conditions. To attribute to anthropogenic climate forcing, the anthropogenic trends in
48 climate are estimated from a range of different climate models and subtracted from the observed climate e.g.,
49 Abatzoglou and Williams (2016) for changes in the extent of forest fires or Diffenbaugh and Burke (2019)
50 for effects on economic inequality) or the `no anthropogenic climate forcing' baseline is directly derived
51 from a large ensemble of climate model simulations not accounting for anthropogenic forcings e.g.,
52 Kirchmeier-Young et al. (2019b) for the extent of forest fires). In any case it has to be demonstrated that the
53 applied impact models are able to explain the observed changes in natural, human or managed systems by
54 e.g., reproducing the observations when forced by observed changes in climate-related systems and other
55 relevant drivers.
56
Do Not Cite, Quote or Distribute 16-16 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 In a situation where an influence of other direct human drivers can be excluded (e.g., by restriction to remote
2 areas not affected by direct human interventions), the `no climate-change' baseline can also be approximated
3 by data from early observational periods with no or minor levels of climate change. In particular, the
4 contribution of climate change to the observed changes in ecosystems is often also determined by a `multiple
5 lines of evidence' approach where the baseline is not formally quantified but the observed changes are
6 identified as a signal of climate change compared to a no-climate change situation based on process
7 understanding from e.g. paleo data and laboratory or field experiments in combination with individual long
8 term observational records and the large scale spatial or temporal pattern of observed changes that can hardly
9 be explained by alternative drivers (Parmesan et al., 2013).
10
11 To date, explicit accounting for direct human or natural influences is often hampered by an incomplete
12 understanding of the processes and limited observational data. There are, however, first studies
13 demonstrating the potential of detailed process-based or empirical modelling that explicitly account for
14 known variations in direct human or natural drivers and separate their effects from the ones induced by
15 changes in the climate-related systems. Examples are Butler et al. (2018) for the separation of growing
16 season adjustments from within growing season climate effects on US crop yields; Wang and Hijmans
17 (2019), separating effects of shifts in land use from climate effects; Jongman et al. (2015); Formetta and
18 Feyen (2019), and Tanoue et al. (2016) for the separation of changes in exposure and vulnerability from
19 climate effects on river floods; Kirchmeier-Young et al. (2019b) for wildfire attribution; Venter et al. (2018)
20 for the attribution of ecosystem structural changes to climate change versus other disturbances.
21
22 There also has been significant progress in the compilation of fragmented and distributed observational data
23 (e.g., Cohen et al. (2018) for phenological ecosystem changes, Poloczanska et al. (2013) for distributional
24 shifts in marine ecosystems, the new global fire atlas (Andela et al., 2019) including information about
25 individual fire size, duration, speed and direction), as well as regional downscaling (e.g., Ray et al. (2015))
26 allowing for the identification of an overall picture of the impacts of progressing climate change. Given the
27 ever increasing body of literature on observed changes in natural, human, and managed systems there also is
28 a first machine learning approach for an automated identification for relevant literature that could
29 complement or support expert assessments as the one provided here (Callaghan et al., 2021).
30
31 16.2.3 Observed Impacts
32
33 In this section we synthesize observed impacts across a range of ecosystems, sectors, and regions. Figure
34 16.2 summarizes the attribution of observed (regional) changes in natural, human or managed systems
35 (orange symbols and confidence ratings), the quantification of weather sensitivity of those systems (blue
36 symbols and confidence ratings), and the attribution of underlying changes in the climate-related systems to
37 anthropogenic forcing (grey symbols and confidence ratings). The Figure can be read as a summary and
38 Table of content for the underlying Tables 16.B.1 on climate attribution, 16.B.2 on impact attribution, and
39 16.B.3 on identification of weather sensitivity that provide the more detailed explanations behind each
40 regional or global assessment, including all references. The synthesis was generated in collaboration with
41 `detection and attribution contact persons' from the individual chapters that each includes its own assessment
42 of observed impacts, and contributing authors on individual topics. The synthesis of `climate attribution'
43 studies in Table SM16.21 was particularly informed by the WGI assessment.
44
45 If Figure 16.2 only provides an assessment of attributed impacts on a given system (e.g., Phenology shifts in
46 terrestrial ecosystems) but does not include an associated `identification of weather sensitivity' that does not
47 mean that the system is not sensitive to weather fluctuations. The focus of our assessment was on `impacts
48 attribution' and we only provide an assessment of `weather sensitivities' if the literature has turned out to
49 provide only limited evidence on impacts of long-term climate change but rather addressed the system's
50 responses to short term weather fluctuations.
51
52 16.2.3.1 Ecosystems
53
54 The collapse or transformation of ecosystems is one of the most abrupt potential tipping points associated
55 with climate change. Climate change has started to induce such tipping points with the first examples
56 including mass mortality in coral reef ecosystems (e.g., Donner et al., 2017; Hughes et al., 2018; Hughes et
57 al., 2019) (high confidence), and changes in vegetation cover triggered by wildfires with climate change
Do Not Cite, Quote or Distribute 16-17 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 suppressing the recovery of the former cover (Tepley et al., 2017; Davis et al., 2019) (low confidence
2 because of the still limited number of studies). Another example of an abrupt change in an ecosystem
3 triggered by a climate extreme is the shift from kelp- to urchin-dominated communities along parts of the
4 Western North America coast due a marine heatwave (Rogers-Bennett and Catton, 2019; McPherson et al.,
5 2021, see `Marine ecosystems - Kelp forest', Table SM16.22) where anthropogenic climate forcing has been
6 shown to have increased the probability for an event of that duration by at least a factor of 33 (Laufkötter et
7 al., 2020). Many terrestrial ecosystems on all continents show evidence of significant structural
8 transformation, including woody thickening and `greening' in more water-limited ecosystems, with a
9 significant role played by rising atmospheric CO2 fertilization in these trends (high confidence) (Fang et al.,
10 2017; Stevens et al., 2017; Burrell et al., 2020). Climate change is identified as a major driver of increases in
11 burned areas in the Western US (high confidence, see `Terrestrial ecosystems - Burned areas', Table
12 SM16.22).
13
14 There is also a clear footprint of climate change on species distribution, with appreciable proportions of
15 tropical species expanding into the ranges of temperate species, and boreal species moving into Arctic
16 regions (high confidence, see `Marine ecosystems - Range reduction and shift' and `Terrestrial ecosystems -
17 Range reduction and shift', Table SM16.22). Climate change has also shifted the phenology of animals and
18 plants on land and in the ocean (high confidence, see `Marine ecosystems - Phenology shift' and `Terrestrial
19 ecosystems - Phenology shifts', Table SM16.22). Both processes have led to emerging hybridisation,
20 competition, temporal or spatial mismatches in predator-prey, guest-host relationships, and invasion of alien
21 plant pests or pathogens (Edwards and Richardson, 2004; Bebber et al., 2013; Parmesan et al., 2013; Millon
22 et al., 2014; Thackeray et al., 2016).
23
24 16.2.3.2 Water Distribution - River Flooding and Reduction in Water Availability
25
26 Observed trends in high river flows strongly vary across regions but also with the considered time period
27 (Gudmundsson et al., 2019; Gudmundsson et al., 2021) as influenced by climate oscillations such as the El
28 NiñoSouthern Oscillation (Ward et al., 2014). On global scale the spatial pattern of observed trends is
29 largely explained by observed changes in climate conditions as demonstrated by multi-model hydrological
30 simulations forced by observed weather while the considered direct human influences only play a minor role
31 on global scale (Gudmundsson et al., 2021, see `Water distribution - Flood hazards', Table SM16.22). The
32 annual total number of reported fatalities from flooding shows a positive trend (1.5% per year from 1960-
33 2013, Tanoue et al., 2016) which appears to be primarily driven by changes in exposure dampened by a
34 reduction in vulnerability while climate induced increases in affected areas only show a weak positive trend
35 on global scale (see `Water distribution - Flood induced fatalities', Table SM16.22). However, the signal of
36 climate change in flood induced fatalities may be lost in the regional aggregation where effects of increasing
37 and decreasing hazards may cancel out. Thus, a climate driven increase in flood induced damages becomes
38 detectable in continental subregions with increasing discharge while the signal of climate change may not be
39 detectable without disaggregation (Sauer et al., 2021, see `Water distribution: Flood-induced economic
40 damages', Table 16.2), see `Water distribution: Flood-induced economic damages', Table 16.2). Compared
41 to river floods the analysis of impacts of long-term changes in the climate related systems on the reduction in
42 water availability is much more fragmented and reduced to individual case studies regarding associated
43 societal impacts (see `Water distribution - Reductions in water availability + induced damages and fatalities',
44 Table SM16.22). At the same time weather fluctuations have led to reductions in water availability with
45 severe societal consequences and high numbers of drought-induced fatalities and damages in particular in
46 Africa and Asia (see `Water distribution - Reductions in water availability + induced damages and fatalities,
47 Table SM16.23) and impacts on malnutrition (see `Food system - Malnutrition, Table SM16.23). Although
48 anthropogenic climate forcing has increased droughts' intensity or probability in many regions of the world
49 (medium confidence), (`Atmosphere - Droughts, Table SM16.21), the existing knowledge has not yet been
50 systematically linked to attribute long-term trends in malnutrition, fatalities, and damages induced by
51 reduced water availability to anthropogenic climate forcing or long-term climate change. For impacts of
52 individual attributable drought events see Table 4.5 of Chapter 4 and `Water distribution - Reductions in
53 water availability + induced damages and fatalities, Table SM16.23.
54
55 16.2.3.3 Coastal Systems
56
Do Not Cite, Quote or Distribute 16-18 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 With their enormous destructive power tropical cyclones represent a major risk for coastal systems (see
2 `Coastal systems - Damages', Table SM16.23). Despite its relevance, confidence in the influence of
3 anthropogenic climate forcing on the strength and occurrence probability of tropical storms themselves is
4 still low (see `Coastal systems: Tropical cyclones', Table SM16.21). However, anthropogenic climate
5 forcing has become the dominant driver of sea level rise (high confidence) (see `Coastal systems - Mean and
6 extreme sea levels', Table SM16.21) and has increased the risk of coastal flooding, including inundation
7 induced by tropical cyclones. In addition, anthropogenic climate forcing has increased the amount of rainfall
8 associated with tropical cyclones (high confidence) (Risser and Wehner, 2017; Van Oldenborgh et al., 2017;
9 Wang et al., 2018) for hurricane Harvey in 2017 (Patricola and Wehner, 2018) and for hurricanes Katrina in
10 2005, Irma in 2017, and Maria in 2017 (see `Atmosphere - Heavy precipitation', Table SM16.21). Assuming
11 that the extreme rainfall is a major driver of the total damages induced by the tropical cyclone, the
12 contribution of anthropogenic climate forcing to the occurrence probability of the observed rainfall (fraction
13 of attributable risk) can also be considered the fraction of attributable risk of the hurricane-induced damages
14 or fatalities (Frame et al., 2020; Clarke et al., 2021, see `Coastal systems - Damages', Table SM16.22).
15 However, first studies do not only quantify the change in occurrence probabilities but translate the actual
16 change in climate-related systems into the additional area affected by flooding in a process-based way
17 (Strauss et al. (2021), contribution of anthropogenic SLR to damages induced by hurricane Sandy; Wehner
18 and Sampson (2021), contribution increased precipitation to damages induced by hurricane Harvey) and
19 attribute a considerable part of the observed damage to anthropogenic climate forcing. In addition, disruption
20 of local economic activity in Annapolis, Maryland and loss of areas and settlements in Micronesia and
21 Solomon Islands have been attributed to relative sea level rise (Nunn et al., 2017; Albert et al., 2018; Hino et
22 al., 2019) while permafrost thawing and sea ice retreat are additional drivers of observed coastal damages in
23 Alaska (Albert et al., 2016; Smith and Sattineni, 2016; Fang et al., 2017).
24
25 16.2.3.4 Food System
26
27 Crop yields respond to weather variations but also to increasing atmospheric CO2, changes in management
28 (e.g., fertilizer input, changes in varieties), diseases, and pests. However, the weather signal is clearly
29 detectable in national and subnational annual yield statistics in main production regions (see `Food system -
30 Crop yields', Table SM16.23). Over the last decades crop yields have increased nearly everywhere mainly
31 due to technological progress (e.g., Lobell and Field, 2007 (global); Butler et al., 2018 (US); Hoffman et al.,
32 2018 (Sub-Saharan Africa); Agnolucci and De Lipsis, 2019 (Europe)) with only minor areas not
33 experiencing improvements in maize, wheat, rice, and soy yields. However meanwhile, stagnation or decline
34 in yields is also observed on parts of the harvested areas (high confidence) (~20% to 40% of harvested areas
35 of maize, wheat, rice and soy with wheat being most affected) (Ray et al., 2012; Iizumi et al., 2018).
36 Evidence on the contribution of climate change to recent trends is still limited (see `Food system - Crop
37 yields', Table SM16.22). Current global-scale process-based simulations forced by simulated historical and
38 pre-industrial climate miss an evaluation to what degree simulations reproduce observed yields (Iizumi et al.,
39 2018). Global scale empirical approaches do not explicitly account for extreme weather events but growing
40 season average temperatures and precipitation (e.g., Lobell et al., 2011; Ray et al., 2019). In addition, studies
41 are constrained by only fragmented information about changes in agricultural management such as growing
42 season adjustments. Some of these limitations have be overcome in regional studies indicating a climate
43 induced increase (28% of observed trend since 1981) in maize yields in the US (Butler et al., 2018 based on
44 a detailed accounting of impacts of extreme temperatures and growing season adjustments) and a climate
45 induced decrease in millet and sorghum yields (1020% for millet and 515% for sorghum in 2000-2009
46 compared to pre-industrial conditions) in Africa and a negative effect of historical climate change on
47 potential wheat yields (27% reduction from 1990 to 2015) in Australia (Hochman et al., 2017; Sultan et al.,
48 2019) based on detailed process-based modelling including a dedicated evaluation against observed yield
49 fluctuations). However, these findings need additional support by independent studies while results are
50 relatively convergent that climate change has been an important driver of the recent declines in wheat yields
51 in Europe (medium confidence) (Moore and Lobell, 2015; Agnolucci and De Lipsis, 2019; Ray et al., 2019).
52
53 Due to complex interactions with socio-economic conditions, climate-induced trends in crop yields and
54 production do not directly transmit to crop prices, availability of food, or nutrition status. This complexity, in
55 addition to the limited availability of long-term data, has so far impeded the detection and attribution of a
56 long-term impact of climate change on associated food security indicators. However, in a few cases,
57 observed crop prices (e.g., domestic grain price in Russia and Africa, Götz et al., 2016; Mawejje, 2016;
Do Not Cite, Quote or Distribute 16-19 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 Baffes et al., 2019) are shown to be sensitive to fluctuations in local weather through its impact on
2 production (see `Food system - Food prices', Table SM16.23). In addition, there is growing evidence that
3 climate extremes (in particular droughts) have led to malnutrition (in particular stunting of children) in the
4 historical period (medium confidence, see `Food system - Malnutrition', Table SM16.23) but without an
5 attribution of changes to long-term climate change.
6
7 16.2.3.5 Temperature-related Mortality
8
9 There is nearly universal evidence that non-optimal ambient temperatures increase mortality (high
10 confidence), with notable heterogeneity only in the shape of the temperature-mortality relationship across
11 geographical regions but often sharply growing relative risks at the outer 5% of the local historical
12 temperature distributions (Gasparrini et al., 2015; Guo et al., 2018; Carleton et al., 2020; Zhao et al., 2021,
13 see `Other societal impacts - Heat-related mortality', Table SM16.23). Significant advances have been made
14 since AR5 regarding the study of temperature-related excess mortality in previously under-researched
15 regions, such as developing countries and (sub-)tropical climates e.g. Africa (South-East Asia: Dang et al.,
16 2016; Ingole et al., 2017; Mazdiyasni et al., 2017; Wichmann, 2017; e.g., Scovronick et al., 2018; Alahmad
17 et al., 2019; the Middle East: Gholampour et al., 2019; and Latin America: Péres et al., 2020). Progress has
18 also been made with regard to temporal changes in temperature-related excess mortality and underlying
19 population vulnerability over time. Heat-attributable mortality fractions have declined over time in most
20 countries due to general improvements in health care systems, increasing prevalence of residential air
21 conditioning, and behavioural changes. These factors, which determine the susceptibility of the population to
22 heat, have predominated over the influence of temperature change (see `Other societal impacts - Heat-related
23 mortality', Table SM16.22, De'Donato et al., 2015; Arbuthnott et al., 2016; Vicedo-Cabrera et al., 2018a).
24 Important exceptions exist, e.g., where unprecedented heat waves have occurred recently. No conclusive
25 evidence emerges regarding recent temporal trends in excess mortality attributable to cold exposure (Vicedo-
26 Cabrera et al., 2018b). Quantitative detection and attribution studies of temperature-related mortality are still
27 rare. One study (Vicedo-Cabrera et al.), using data from 43 countries, found that 37% (range 20.576.3%) of
28 average warm-season heat-related mortality during recent decades can be attributed to anthropogenic climate
29 change (medium confidence, see `Other societal impacts - Heat-related mortality', Table SM16.22). Studying
30 excess mortality associated with past heat waves, such as the 2003 or 2018 events in Europe, even higher
31 proportions of deaths attributable to anthropogenic climate change have been reported for France and the UK
32 (Mitchell et al., 2016; Clarke et al., 2021). Formal attribution studies encompassing cold-related mortality
33 are quasi non-existent. The very few studies from Europe and Australia (Christidis et al., 2010; Åström et al.,
34 2013; Bennett et al., 2014) find weak impacts of climate change on cold-associated excess mortality, with
35 contradictory outcomes both towards higher and lower risks (low confidence, see `Other societal impacts -
36 Heat-related mortality', Table SM16.22).
37
38 16.2.3.6 Water-borne Diseases
39
40 Infectious diseases with water-associated transmission pathways constitute a large burden of disease
41 globally. Since AR5 the evidence has strengthened that waterborne diseases, and especially gastrointestinal
42 infections, are highly to moderately sensitive to weather variability (medium confidence, see `Water
43 distribution - Water-borne diseases', Table SM16.23). Increased temperature and high precipitation, with
44 associated flooding events, have been shown to generally increase the risk of diarrhoeal diseases. There are
45 however a number of studies that describe important exceptions and modifications to this general
46 observation. While high temperatures favour bacterial diarrhoeal diseases, virally transmitted diarrhoea is on
47 the contrary mostly associated with low temperatures (Carlton et al., 2016; Chua et al., 2021). Socio-
48 economic determinants, such as the existence of single household water supplies (Herrador et al., 2015) or
49 combined sewer overflows (Jagai et al., 2017), have been shown to critically increase the risk of
50 gastrointestinal infections linked to heavy rainfall in high-income countries. Also, for both low- and high-
51 income countries it has been found that gastrointestinal diseases increase following a heavy rainfall event
52 only if preceded by a dry period (Carlton et al., 2014; Setty et al., 2018). Yet, so far there is no consistent
53 evidence on the role of droughts in favouring waterborne disease transmission (Levy et al., 2016). As
54 exemplified by the large cholera outbreak following the 2010 earthquake in Haiti, the existence of
55 functioning sanitation systems is critical for preventing waterborne disease outbreaks, while climatic factors
56 (especially rainfall) are important in driving the transmission dynamics once the outbreak has started
57 (Rinaldo et al., 2012). Other socio-economic factors, such as human mobility and water management project
Do Not Cite, Quote or Distribute 16-20 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 (e.g., dam constructions) also modify the strength of the association between climatic factors and waterborne
2 diseases, as shown by recent studies in Africa (Perez-Saez et al., 2015; Finger et al., 2016).
3
4 Whereas the weather sensitivity of waterborne diseases is well-established for all world regions (see `Water
5 distribution - Water-borne diseases', Table SM16.23), studies attempting to attribute recent trends in
6 waterborne disease to climate change are non-existent, except for investigations on the distribution of marine
7 Vibrio bacteria and associated disease outbreaks in the coastal North Atlantic and the Baltic Sea regions
8 (Baker-Austin et al., 2013; Baker-Austin et al., 2016; Vezzulli et al., 2016; Ebi et al., 2017). These
9 investigations provide evidence that increases in sea surface temperatures over recent decades as well as
10 during recent summer heat waves are linked to increased concentrations of Vibrio bacteria in coastal waters
11 and an associated rise in environmentally acquired Vibrio infections in humans.
12
13 16.2.3.7 Vector-borne Diseases
14
15 Vector-borne diseases constitute a large burden of infectious diseases worldwide and are highly sensitive to
16 fluctuations of weather conditions including extreme events. Thus, both extreme rainfall and droughts have
17 increased infections (high confidence, see documentation of cases in `Other societal impacts - Vector-borne
18 diseases', Table SM16.23). For example, in Sudan, anomalous high rainfall increased Anopheles mosquito
19 breeding sites, leading to malaria outbreaks (Elsanousi et al., 2018) while in Barbados and Brazil, drought
20 conditions in urban areas have enhanced dengue incidence due to changes in water storage behaviour
21 creating breeding sites for Aedes mosquitoes around human dwellings (Lowe et al., 2018; Lowe et al.,
22 2021)). In the Caribbean and Pacific island nations, weather extremes, such as storms and flooding have led
23 to outbreaks of dengue due to disruption to water and sanitation services, leading to increased exposure to
24 Aedes mosquito breeding sites (Descloux et al., 2012; Sharp et al., 2014; Uwishema et al., 2021). In South
25 and Central America, and Asia, dengue incidence has been shown to sensitive to variations in temperature
26 and the monsoon season in addition to variations induced by urbanization and population mobility (high
27 confidence (South and Central America); medium confidence (Asia); see `Other societal impacts - Vector-
28 borne diseases', Table SM16.23).
29
30 The attribution of changes in disease incidence to long-term climate change is often limited by relatively
31 short reporting periods often only covering 10-15 years. Most studies then attribute trends in the occurrence
32 of vector-borne diseases to the trends in climate across the same observational period and do not refer to an
33 early `no climate change' baseline climate. This means that they also capture trends induced by longer term
34 climate oscillations. Nevertheless, we list them in Table SM16.22 on `impact attribution' to clearly
35 distinguish them from the analysis of interannual fluctuations. The overall consistency of their findings
36 across regions and time windows indicates that climate change is an important driver of the observed
37 latitudinal or altitudinal range expansions of vector-borne diseases into previously colder areas (medium to
38 high confidence, see `Other societal impacts - Vector-borne diseases', Table SM16.22). In highland areas of
39 Africa and South America, epidemic outbreaks of malaria have become more frequent due to warming trends
40 that allow Anopheles mosquitoes to persist at higher elevations (Pascual et al., 2006; Siraj et al., 2014). In the
41 US, ticks that transmit Lyme disease have expanded their range northwards due to warmer temperatures
42 (high confidence, (Kugeler et al., 2015; McPherson et al., 2017; Lin et al., 2019; Couper et al., 2020, see
43 `Other societal impacts - Vector-borne diseases', Table SM16.22). In Southern Europe, climate suitability
44 for Aedes mosquitoes, which transmit dengue and chikungunya, and Culex mosquitoes, which transmit West
45 Nile virus, has also increased and contributed to unprecedented outbreaks including the 2018 West Nile fever
46 outbreak (medium confidence) (Medlock et al., 2013; Paz et al., 2013; Roiz et al., 2015; ECDC, 2018, see
47 `Other societal impacts - Vector-borne diseases', Table SM16.22).
48
49 16.2.3.8 Economic Impacts
50
51 Since the AR5, there has been significant progress regarding the identification of economic responses to
52 weather fluctuations: Evidence has increased that extreme weather events such as tropical cyclones,
53 droughts, and severe fluvial floods have not only caused substantial immediate direct economic damage
54 (high confidence, see `Coastal Systems - Damages, Table SM16.23, `Water distribution - Reductions in
55 water availability + induced damages and fatalities, Table SM16.23, and `Water distribution - Flood-induced
56 economic damages, Table SM16.22), but have also reduced economic growth in the short-term (year of, and
57 year after event) (Strobl, 2011; Strobl, 2012; Fomby et al., 2013; Felbermayr and Gröschl, 2014, Loyaza et
Do Not Cite, Quote or Distribute 16-21 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 al. 2012) (high confidence) as well as long-term (up to 10-15 years after event) (medium confidence) (Hsiang
2 and Jina, 2014; Berlemann and Wenzel, 2016; Berlemann and Wenzel, 2018; Krichene et al., 2020; Tanoue
3 et al., 2020, see `Other societal impacts - Macroeconomic output', Table SM16.23). Short- and long-term
4 reductions of economic growth by extreme weather events affect both, developing and industrialized
5 countries, but have been shown to be more severe in developing than in industrialized economies thereby
6 increasing inequality between countries (high confidence, see `Other societal impacts - Between country
7 inequality', Table SM16.23). Further, extreme weather events have increased within-country inequality since
8 poorer people are more exposed and suffer relatively higher well-being losses than richer parts of the
9 population (medium confidence, see `Other societal impacts - Between country inequality', Table SM16.23).
10 Going beyond extreme weather events, economic production depends non-linearly on temperature
11 fluctuations: below a certain threshold temperature, economic production increases with temperature
12 whereas it decreases above a certain threshold temperature (high confidence) (Burke et al., 2015; Pretis et al.,
13 2018; Kalkuhl and Wenz, 2020; Kotz et al., 2021).
14
15 So far, there are few individual studies attributing observed economic damages to long term climate change
16 except for damages induced by river flooding, droughts, and tropical cyclones (see `Coastal systems -
17 Damages', `Water distribution - Flood induced damages', and `Water distribution - Reduction in water
18 availability + induced damages and fatalities', Table SM16.22) extremes to anthropogenic forcing. In
19 addition, the empirical findings on the sensitivity of macroeconomic development to weather fluctuations
20 and extreme weather events have been used to estimate the cumulative effect of historical warming on long
21 term economic development (see `Other societal impacts - Macroeconomic output', Table SM16.22):
22 anthropogenic climate change is estimated to have reduced GDP growth over the last 50 years with
23 substantially larger negative effects on developing countries and in some cases positive effects on colder
24 industrialized countries (low confidence) (Diffenbaugh and Burke, 2019). Globally, between-country
25 inequality has decreased over the last 50 years. Climate change is estimated to have substantially slowed
26 down this trend, i.e., increased inequality compared to a counterfactual no climate change baseline (low
27 confidence) (Diffenbaugh and Burke, 2019). On a regional level, decreasing rainfall trends in Sub-Saharan
28 Africa (SSA) may have increased the GDP per capita gap between SSA and other developing countries (low
29 confidence) (Barrios et al., 2010). Overall, more research is needed on the impact channels through which
30 extreme weather events and weather variability can hinder economic development, especially in the long-
31 term.
32
33 16.2.3.9 Social Conflict
34
35 There are few studies directly attributing changes in conflict risk to climate change in the modern era (van
36 Weezel, 2020), preventing a confident assessment of the effect of long-term changes in the climate-related
37 systems on armed conflict (see `Other societal impacts - Social conflict', Table SM16.22). However, a
38 sizeable literature links the prevalence of armed conflict within countries to within- and between-year
39 variations in rainfall, temperature or drought exposure; often via reduced-form econometric analysis or
40 statistical models that control for important non-climatic factors, such as agricultural dependence, level of
41 economic development, state capacity, and ethnopolitical marginalization (see `Other societal impacts -
42 Social conflict' in Table SM16.23). Overall, there is more consistent evidence that climate variability has
43 influenced low-intensity organized violence than major civil wars (Detges, 2017; Nordkvelle et al., 2017;
44 Linke et al., 2018). Likewise, there is more consistent evidence that climate variability has affected dynamics
45 of conflict, such as continuation, severity, and frequency of violent conflict events, than the likelihood of
46 initial conflict outbreak (Yeeles, 2015; Eastin, 2016; Von Uexkull et al., 2016, Section 7.2.7). Moreover,
47 research suggests with medium confidence (medium evidence, medium agreement) that weather effects on
48 armed conflict have been most prominent in contexts marked by a large population, low socioeconomic
49 development, high political marginalization, and high agricultural dependence (Theisen, 2017; Koubi, 2019;
50 Buhaug et al., 2020; Ide et al., 2020).
51
52 Some studies also seek to evaluate potential indirect links between climate and weather anomalies and
53 prevalence of armed conflict via food price shocks or forced migration. While there is robust evidence that
54 the likelihood of social unrest in the developing world generally increases in response to rapid growth in
55 food prices (Bellemare, 2015; Rudolfsen, 2018), the magnitude of the climate effect on unrest via food prices
56 is less well established (Martin-Shields and Stojetz, 2019). Similarly, research shows with high confidence
57 that climate variability and extremes have affected human mobility (see `Other societal impacts -
Do Not Cite, Quote or Distribute 16-22 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 Displacement and migration', Table SM16.23), but there is low agreement and limited evidence that
2 weather-induced migration has increased the likelihood of armed conflict (Section 7.2, Brzoska and Fröhlich,
3 2016; Kelley et al., 2017; Selby et al., 2017; Abel, 2019). Research on weather-related effects on interstate
4 security generally conclude that periods of transboundary water scarcity are more likely to facilitate
5 increased international cooperation than conflict (Bernauer and Böhmelt, 2020).
6
7 In general, the historical influence of climate on conflict is judged to be small when compared to dominant
8 conflict drivers (Mach et al., 2019). Much of this research is limited to (parts of) Sub-Saharan Africa, which
9 raises some concerns about selection bias and generalizability of results (Adams et al., 2018).
10
11 16.2.3.10 Displacement and Migration
12
13 Given the complexity of human migration processes and decisions (e.g., Boas et al., 2019, Cattaneo et al.,
14 2019) and the paucity of long-term, reliable and internally consistent observational data on displacement
15 (IDMC, 2019; IDMC, 2020) and migration (Laczko, 2016) the contribution of long-term changes in climate
16 related systems to observed human displacement or migration patterns has not been quantified so far, except
17 for individual examples for displacement induced by inland flooding where the heavy precipitation has been
18 attributed to anthropogenic climate forcing and coastal flooding (see `Other societal impacts - Displacement
19 and migration', Table SM16.22; CCP2).
20
21 However, new evidence has emerged since the AR5 that further documents widespread effects of weather
22 fluctuations and extreme events on migration (see `Other societal impacts- Displacement and migration' in
23 Table SM16.23). Numerous studies find significant links between temperature or precipitation anomalies, or
24 extreme weather events such as storms or floods, and internal as well as international migration (Coniglio
25 and Pesce, 2015; Cattaneo and Peri, 2016; Nawrotzki and DeWaard, 2016; Beine and Parsons, 2017for
26 international migration; and IDMC, 2019 for internal displacement). Internal displacement of millions of
27 people every year is triggered by natural hazards, mainly floods and storms (IDMC, 2019). The effects of
28 weather fluctuations and extremes on migration are considered more important for temporary mobility and
29 displacement than permanent migration, and more influential on short-distance movement, including
30 urbanization, than international migration (McLeman, 2014; Hauer et al., 2020; Hoffmann et al., 2020,
31 Section 7.2.6). Importantly, these links are conditional on the socio-economic situation in the origin; e.g.,
32 poor populations may be `trapped' and not be able to migrate in the face of adverse climate or weather
33 conditions (Black et al., 2013; Adams, 2016). Many studies have also explored the channels through which
34 climate or weather influence migration, and have identified incomes in the agricultural sector as one of the
35 main channels (Nawrotzki et al., 2015; Viswanathan and Kavi Kumar, 2015; Cai et al., 2016a). In particular,
36 declines in agricultural incomes and employment due to changed weather variability may foster increased
37 rural-urban movement; and the resulting pressures on urban wages in turn fosters international migration
38 (Marchiori et al., 2012; Maurel and Tuccio, 2016). Another possible but controversial channel is violent
39 conflict, which may be fostered (though not exclusively caused) by adverse climate conditions such as
40 drought, and in turn lead to people seeking refugee status, although evidence of such an indirect effect is
41 weak (Brzoska and Fröhlich, 2016; Abel et al., 2019; Schutte et al., 2021).
42
43 16.2.3.11 Case study on climate change and the outbreak of the Syrian civil war
44
45 Separating between climatic and non-climatic factors in impact attribution is often challenging, as
46 highlighted by the debate surrounding the causes of the Syrian civil war. During the years 20062010, the
47 Fertile Crescent region in Eastern Mediterranean and Western Asia was hit by the worst drought on
48 meteorological record, compounding a consistent drying of the region over the past half century (Trigo et al.,
49 2010; Hoerling et al., 2012; Mathbout et al., 2018, SR15 BOX 3.2). The magnitude of the multiyear drought
50 is estimated to have become two to three times more likely as a result of increased CO2 forcing (Kelley et al.,
51 2015). The drought had a devastating impact on agricultural production in the northeast of Syria. In 2007
52 2008 alone, average crop yields dropped by 32% in irrigated areas and as much as 79% in rain-fed areas (De
53 Châtel, 2014), and herders in the northeast lost around 85% of their livestock (Werrell et al., 2015).
54 Successive years with little or no income eventually forced people to leave their farms in great numbers and
55 seek employment in less affected parts of the country, adding to existing pressures on housing, labour
56 market, and public goods provision (Gleick, 2014; Kelley et al., 2015). In March 2011, by which time the
Do Not Cite, Quote or Distribute 16-23 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 `Arab Spring' uprisings had gained momentum and spread across much of the region, anti-regime protests
2 broke out in Syria, first in the southern city of Dara'a and then in Damascus and throughout the country.
3
4 Yet, the attribution of the Syrian civil war to climate change has triggered a heated debate. A number of
5 studies argue that the principal drivers of the drought-induced economic collapse were political rather than
6 environmental in nature, shaped by adverse economic reforms and unsustainable agricultural policies,
7 promoting water-intensive irrigation schemes for cotton cultivation and implementing abrupt subsidy cuts at
8 the peak of the drought, implying that many poor farmers no longer could afford fertilizers or fuel to power
9 irrigation pumps (Barnes, 2009; De Châtel, 2014; Eklund and Thompson, 2017; Selby et al., 2017). Thus, the
10 200610 drought did not precipitate similar devastating socioeconomic impacts on agrarian communities
11 across the borders in Turkey, Iraq or Jordan, although environmental conditions were comparable (Trigo et
12 al., 2010; Eklund and Thompson, 2017; Feitelson and Tubi, 2017).
13
14 However, the relevant attribution question is not whether the same drought would produce the same
15 consequences under different political and socio-economic conditions but rather, given the same political and
16 socio-economic context, how would the outcomes have differed in the absence of the climate event?
17 Research still provides very limited insights into whether and how the escalation process would have
18 evolved differently in a counterfactual no-climate change world.
19
20 Thus, the role of the drought in augmenting pre-existing internal migration, and the role of the distress
21 migration in accentuating demographic, economic, and social pressures in receiving areas, remain contested.
22 Estimates of the number of people who abandoned their farms in response to the drought range from less
23 than 4060,000 families (Selby et al., 2017) to more than 1.5 million displaced (Gleick, 2014). However, the
24 numbers have to be seen in the context of prevailing population growth, significant rural-urban migration,
25 and the preceding inflow of around 1.5 million refugees from neighbouring Iraq (De Châtel, 2014;
26 Hoffmann, 2016). In addition, research suggests that the migrants played a peripheral role in the initial social
27 mobilization in March 2011 (Fröhlich, 2016).
28
29 While it is undisputed that the drought caused direct economic losses, its overall additional impact on the
30 Syrian economy, relative to other prevalent drivers of economic misery, including rampant unemployment,
31 increasing inequalities, declining rural productivity, and loss of oil revenues (Aïta, 2009; Landis, 2012; De
32 Châtel, 2014; Selby, 2019) has not been quantified.
33
34 In addition, the protesters' demands centred around contentious political rather than economic issues,
35 including release of political prisoners, ending of torture and indiscriminate violence by security forces, and
36 abolishment of the near 50-year old state of emergency (Selby et al., 2017; Ash and Obradovich, 2020). The
37 mobilization in Syria in the spring of 2011 also made explicit references to events across the Middle East and
38 North African region. Analyses of regional and social media and networks show high level of interaction
39 across the Arab world, and the initial Syrian uprising adopted a mobilization model and rhetorical frames
40 similar to those developed in Tunisia and Egypt (Leenders, 2013; 2014). However, the Syrian uprising stands
41 out in how it was met with overwhelming violent force by the police and security forces, which changed the
42 character of the resistance and opened up for militarization of non-state actors that further escalated the
43 conflict (Heydemann, 2013; Leenders, 2013; Bramsen, 2020).
44
45 In summary, the drought itself is shown to be attributable to greenhouse gas emissions. The agricultural
46 losses and internal migration from rural to urban areas can be directly linked to the drought and in this way
47 are partly attributable to greenhouse gas emissions, although there are no studies comparing the observed
48 losses and number of people displaced to a counterfactual situation of a weaker drought in a `no climate
49 change' situation. Current research does not provide enough evidence to attribute the civil war to climate
50 change. In contrast, it is likely that social uprisings would have occurred even without the drought.
51
52
Do Not Cite, Quote or Distribute 16-24 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1
2 Figure 16.2: Impact of Climate Change or Weather Fluctuations.
3
4
5 16.3 Synthesis of Observed Adaptation-related Responses
6
7 A new development since AR5, is there is now growing evidence assessing progress on adaptation across
8 sectors, geographies and spatial scales. Uncertainty persists around what defines adaptation and how to
9 measure it (Cross-Chapter Box FEASIB in Chapter 18, UNEP, 2021). As a result, most literature
10 synthesizing responses are based on documented or reported adaptations only, and are thus subject to
11 substantial reporting bias.
Do Not Cite, Quote or Distribute 16-25 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1
2 We document implemented adaptation-related responses that could directly reduce risk. Adaptation as a
3 process is more broadly covered in Chapter 17 (Section 17.4.2), including risk management, decision-
4 making, planning, feasibility (see Cross-Chapter Box FEASIB in Chapter 18), legislation and learning. Here,
5 we focus on a subset of adaptation activities: adaptation-related responses of species, ecosystems, and human
6 societies that have been implemented, observed, and could directly reduce risk. We consider all adaptation-
7 related responses to assumed, perceived, or expected climate risk, regardless of whether or not impacts or
8 risks have been formally attributed to climate change.
9
10 We use the term `adaptation-related responses', recognising that not all responses reduce risk. While
11 `adaptation' implies risk reduction, we use the broader term `responses' to reflect that responses may
12 decrease risk, but in some cases may increase risk.
13
14 It is not currently possible to conduct a comprehensive global assessment of effectiveness, adequacy, or the
15 contribution of adaptation-related responses to changing risk due to an absence of robust empirical literature.
16 This constrains assessment of adaptation progress and gaps in the context of over-shoot scenarios. Given
17 limited evidence to inform comprehensive global assessment of effectiveness and adequacy, we assess
18 evidence that adaptation responses in human systems indicate transformational change. Chapter 17 considers
19 adaptation planning and governance, including adaptation solutions, success, and feasibility assessment
20 (Cross-Chapter Box FEASIB in Chapter 18), discussed further in Box 16.1 (also see Cross-Chapter Box
21 PROGRESS in Chapter 17).
22
23 In natural ecosystems or species, detectable changes can be considered as `impact' or `response'. The
24 distinction between `observed impacts' (16.2) and `observed responses' (16.3) is not always clear. For
25 example, autonomous distributional shifts in wild species induced by increasing temperatures (an observed
26 impact) may reduce risk to the species (an autonomous adaptation response), but this process can be
27 enhanced or supported by human intervention such as intentional changes in land use. Observed autonomous
28 changes in natural ecosystems or species unsupported by human intervention are treated as impacts (see
29 Section 16.2).
30
31 Adaptation-related responses are frequently motivated by a combination of climatic and non-climatic drivers,
32 and interact with other transitions to affect risk. For societal responses, it is difficult to say whether they are
33 triggered by observed or anticipated changes in climate, by non-climatic drivers, or a combination of all
34 three. In the case of observed impacts, assessment typically focuses on detection and attribution vis à vis a
35 counterfactual of no climate change. While there has been some effort to attribute reduced climate risk to
36 adaptation-related responses (Toloo et al., 2013a; Toloo et al., 2013b; Hess et al., 2018; Weinberger et al.,
37 2018), in many cases this has not been feasible given difficulties in defining adaptation and empirically
38 disentangling the contribution of intersecting social transitions and changing risks. Literature on adaptation-
39 related response frequently draws on theories-of-change to assess the likely contribution of adaptations to
40 changes in risk, including maladaptation and co-benefits.
41
42 16.3.1 Adaptation-related Responses by Natural Systems
43
44 There is growing evidence of shifts in species distributions and ecosystem structure and functioning in
45 response to climate change (Chapter 2). While many species are increasingly responding to climate change,
46 there is limited evidence that these responses will be fully adaptive, and for many species the rate of
47 response appears insufficient to keep pace with the rate of climate change under mid- and high-range
48 emissions scenarios (medium confidence). There is relatively limited, but growing, empirical data to
49 document adaptation of natural systems in the absence of human interventions. For example, Scheffers et al.
50 (2016) reviewed climate responses across diverse species, reporting widespread and extensive observed
51 changes in organisms (genetics, physiology, morphology), populations (phenology, abundance and
52 dynamics), species (distributions), and ecosystems. A systematic review by Franks et al. (2014) synthesized
53 evidence from 38 empirical studies of changes in terrestrial plant populations, finding evidence to support a
54 mix of plastic and evolutionary responses. Boutin and Lane (2014) similarly reviewed adaptive responses in
55 mammals, finding most species' responses due to phenotypic plasticity. Charmantier and Gienapp (2014)
56 reviewed responses to climate change among birds, finding emerging evidence that birds from a range of
57 taxa show advancement in their timing of migration and breeding in response to warming. Aragão et al.
Do Not Cite, Quote or Distribute 16-26 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 (2018) reviewed adaptation responses in marine systems, including 12 studies of live marine mammals. They
2 observed widespread evidence of shifting distributions and timing of biological events (Chapter 2, Chapter 3,
3 and Cross-Chapter Paper 1 on Biodiversity Hotspots).
4
5 Some ecosystems and species' responses may be insufficient to keep pace with rates of climate change. It
6 is difficult to distinguish whether adaptations are due to genotypic change or to phenotypic plasticity. Long-
7 term natural adaptations will require the former, but the latter may provide short-term coping mechanisms to
8 `buy time' to respond to climate changes or lay foundations for evolutionary adaptation. There is mixed
9 evidence regarding evolutionary versus plastic responses, with relatively limited evidence of longer-term
10 evolutionary responses of species that can be associated with climate change. Similarly, it is difficult to
11 assess whether responses are indeed potentially adaptive (e.g., coping, shifting, migrating) or simply
12 reflective of impacts (e.g., stress, damage). Among mammal responses reviewed by Boutin and Lane (2014),
13 for example, only 4 of 12 studies found some evidence that responses were adaptive. Even where adaptive
14 responses are occurring, they may not be sufficient to keep pace with the rate of climate change. found, for
15 example, that among the twelve studies in their review that directly assessed the sufficiency of responses to
16 keep pace with the rate of climate change, eight concluded that responses would be insufficient to avert
17 extinction.
18
19 16.3.2 Adaptation-related Responses by Human Systems
20
21 The literature that seeks to assess adaptation progress is growing at the global (Berrang-Ford et al., 2021a),
22 regional (Bowen and Ebi, 2015; England et al., 2018; Robinson, 2018a; Wirehn, 2018; Olazabal et al., 2019;
23 Thomas et al., 2019a; Biesbroek et al., 2020; Canosa et al., 2020; Robinson, 2020b), national (Hegger et al.,
24 2017; Lesnikowski et al., 2019a; Lesnikowski et al., 2019b), and municipal (Araos et al., 2016; Reckien et
25 al., 2018; Reckien et al., 2019; Lesnikowski et al., 2020; Singh et al., 2021) levels, using National
26 Communications (Gagnon-Lebrun and Agrawala, 2007; Lesnikowski et al., 2015; Muchuru and Nhamo,
27 2017), local climate change action plans (Regmi et al., 2016b; Regmi et al., 2016a; Reckien et al., 2018;
28 Reckien et al., 2019), adaptation project proposals, and reported adaptations in the peer reviewed literature.
29 There remains persistent publication bias in the evidence base on adaptation given the difficulty of
30 integrating diverse knowledge sources (see Section 16.3.3). To better assess how adaptation is occurring in
31 human systems, we draw on this literature base and characterize evidence of adaptation across regions and
32 sectors in terms of five key questions (Table 16.4, Ford et al., 2013; Biagini et al., 2014; Ford et al., 2015a;
33 Bednar and Henstra, 2018; Reckien et al., 2018; Tompkins et al., 2018): What types of hazards are
34 motivating adaptation-related responses? Who is responding? What types of responses are being
35 documented? What evidence is available on adaptation effectiveness, adequacy, and risk reduction? To
36 characterize evidence that adaptation responses indicate transformation, we use a typology based on four
37 dimensions of climate adaptation: scope, depth, and speed, and consideration of limits to adaptation (Section
38 16.4, Termeer et al., 2017; Berrang-Ford et al., 2021a).
39
40 16.3.2.1 What Hazards are Motivating Adaptation-related Responses?
41
42 Drought and precipitation variability are the most prevalent hazards in the adaptation literature, particularly
43 in the context of food and livelihood security. Adaptation frequently occurs in response to specific rapid or
44 slow-onset physical events that can have adverse impacts on people. In some cases, people adapt in
45 anticipation of climate change in general or to take advantage of new opportunities created by hazards (e.g.,
46 increased navigability due to melting sea ice). There is evidence that prior experience with hazards increases
47 adaptation response (Barreca et al., 2015). Following drought and precipitation variability, the next specific
48 hazards that are most frequently documented in the global adaptation literature are heat and flooding. Heat,
49 while less salient, appears to be a driver of adaptation across all regions and sectors (Stone Jr et al., 2014;
50 Hintz et al., 2018; Nunfam et al., 2018). Drought, extreme precipitation, and inland flooding are commonly
51 reported in the context of water and sanitation (Bauer and Steurer, 2015; Lindsay, 2018; Kirchhoff and
52 Watson, 2019; Hunter et al., 2020; Simpson et al., 2020). Flooding is frequently reported as a key hazard for
53 adaptation in cities, followed by drought, precipitation variability, heat, and sea level rise (Broto and
54 Bulkeley, 2013; Araos et al., 2016; Georgeson et al., 2016; Mees, 2017; Reckien et al., 2018; Hunter et al.,
55 2020).
56
57
Do Not Cite, Quote or Distribute 16-27 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1
2 Figure 16.3: Salience of different types of hazards in the scientific literature on adaptation-related responses (i.e.,
3 responses that people undertake to reduce risk from climate change and associated hazards). Updated from a systematic
4 review of 1,682 scientific publications (2013-2019) reporting on adaptation-related responses in human systems
5 (Berrang-Ford et al., 2021a). Numbers in table reflect the number of publications reporting. Darker colours denote more
6 extensive reporting on a hazard as a motivating factor for the response. Publications are counted in all relevant regions
7 or sectors.
8
9
10 16.3.2.2 Who is Responding?
11
12 Individuals and households play a central role in adaptation globally. The most frequently reported actors
13 engaged in adaptation-related responses in the scientific literature are individuals and households,
14 particularly in the global south (Fig. 16.4). Regionally, household- and individual-level adaptation is
15 documented most extensively in Africa and Asia, and to a lesser but still substantial extent in North America
16 (Fig. 16.4).
17
18 National and local governments are also frequently engaged in reported adaptation across most regions.
19 In Africa and Asia, reported adaptations have been primarily associated with individuals, households,
20 national governments, NGOs, and international institutions, with more limited reporting of involvement from
21 sub-national governments or the private sector (Ford et al., 2015a; Ford and King, 2015; Hunter et al., 2020).
22 Engagement by sub-national governments in adaptation is more frequently documented in Europe and North
23 America (Craft and Howlett, 2013; Craft et al., 2013; Bauer and Steurer, 2014; Lesnikowski et al., 2015; Shi
24 et al., 2015; Austin et al., 2016). Reporting of private sector engagement is generally low. Civil society
25 participation in adaptations is reported across all regions. Consistent with this, local governments are also
26 widely reported in documented adaptation responses, particularly where municipal jurisdiction is high,
27 including cities, infrastructure, water, and sanitation.
28
29
Do Not Cite, Quote or Distribute 16-28 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1
2 Figure 16.4: Who is responding, by geographic region and sector? Cell contents indicate the number of publications
3 reporting engagement of each actor in adaptation-related responses. Darker colours denote a high number of
4 publications. Based on a systematic review of 1,682 scientific publications (2013-2019) reporting on adaptation-related
5 responses in human systems (Berrang-Ford et al., 2021a). SIS: Small Island States; Terr: Terrestrial and freshwater
6 ecosystems.
7
8
9 16.3.2.3 What Types of Responses are Documented?
10
11 Behavioural change is the most common form of adaptation. The scientific literature presents extensive
12 evidence of behavioural adaptation -- change in the strategies, practices, and actions that people, particularly
13 individuals and households, undertake to reduce risk (Figure 16.5). This includes, for example, household
14 measures to protect homes from flooding, protect crops from drought, relocation out of hazard zones, and
15 shifting livelihood strategies (Porter et al., 2014). This is followed by adaptation via technological innovation
16 and infrastructural development, nature-based adaptation (enhancing, protecting, or promoting ecosystem
17 services), and institutional adaptation (enhancing multilevel governance or institutional capabilities).
18 Behavioural adaptation is most frequently documented in Asia, Africa, and Small Island States, and in the
19 agriculture, health, and development sectors. In the agricultural sector, households are adopting or changing
20 to crops and livestock that are more adapted to drought, heat, moisture, pests, and salinity (Arku, 2013;
21 Kattumuri et al., 2017; Wheeler and Marning, 2019). Studies in Africa and Asia have documented shifts in
22 farming and animal husbandry practice (Arku, 2013; Garcia de Jalon et al., 2016; Gautier et al., 2016;
23 Chengappa et al., 2017; Epule et al., 2017; Kattumuri et al., 2017; Abu and Reed, 2018; Asadu et al., 2018;
24 Haeffner et al., 2018; Shaffril et al., 2018; Wiederkehr et al., 2018; Zinia and McShane, 2018; Currenti et al.,
25 2019; Fischer, 2019a; Fischer, 2019b; Schofield and Gubbels, 2019; Sereenonchai and Arunrat, 2019;
26 Wheeler and Marning, 2019; Mayanja et al., 2020). In Small Island Nations, studies have documented
27 household flood protections measures such as raising elevation of homes and yards, creating flood barriers,
28 improving drainage, moving belongings, and in some cases, relocating (Middelbeek et al., 2014; Currenti et
29 al., 2019; Klock and Nunn, 2019).
30
31 The mix of adaptation response types differs across regions and sectors. Technological and infrastructural
32 responses are widely reported in Europe, and globally in the context of cities and water and sanitation (Mees,
33 2017; Hintz et al., 2018). Responses to flood risk in Europe include the use of flood and climate resistant
34 building materials, large scale flood management, and water storage and irrigation systems (van Hooff et al.,
35 2015; Mees, 2017). Technological and infrastructural responses are also documented to some extent in
36 agriculture, including for example breeding more climate resilient crops, precision farming and other high-
37 tech solutions such as genetic modification (Makhado et al., 2014; Fisher et al., 2015; Costantini et al., 2020;
38 Fraga et al., 2021; Grusson et al., 2021; Naulleau et al., 2021). While less common, institutional responses
39 are more prominent in North America and Australasia as compared to other regions, and include zoning
40 regulations, new building codes, new insurance schemes, and coordination mechanisms (Craft and Howlett,
41 2013; Craft et al., 2013; Parry, 2014; Ford et al., 2015b; Beiler et al., 2016; Lesnikowski et al., 2016; Labbe
42 et al., 2017; Sterle and Singletary, 2017; Hu et al., 2018; Conevska et al., 2019). Institutional adaptations are
43 more frequently reported in cites than other sectors. Institutional adaptation may be particularly subject to
44 reporting bias, however, with many institutional responses likely to be reported in the grey literature (see
45 Chapter 17). Nature-based solutions are less frequently reported, except in Africa, where they are relatively
46 well-documented, and in the content of terrestrial systems where reports included species regeneration
Do Not Cite, Quote or Distribute 16-29 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 projects, wind breaks, erosion control, reforestation, and riparian zone management (Munji et al., 2014;
2 Partey et al., 2017; Muthee et al., 2018).
3
4
5
6
7 Figure 16.5: Type of adaptation responses by global region. Percentages reflect the number of articles mentioning each
8 type of adaptation over the total number of articles for that region. Radar values do not total 100% per region since
9 publications frequently report multiple types of adaptation; for example, construction of drainage systems
10 (infrastructural), changing food storage practices by households (behavioural), and planting of tree cover in flood prone
11 areas (nature-based) in response to flood risk to agricultural crops. Data updated and adapted from Berrang-Ford et al.
12 (2021a), based on 1682 scientific publications reporting on adaptation-related responses in human systems.
13
14
15 Some but not all adaptation-related responses are engaging vulnerable populations in planning or
16 implementation (high confidence) (Araos et al., 2021). Consideration of vulnerable populations is most
17 frequently focused on low-income populations and women through the inclusion of informal or formal
18 institutions or representatives in adaptation planning, or through targeted adaptations to reduce risk in these
19 populations (high confidence). Consideration of vulnerable groups in adaptation responses are more
20 frequently reported in the global south (medium confidence). Engagement in adaptation planning of
21 vulnerable elderly, migrants, and ethnic minorities remains low across all global regions (medium
22 confidence). There is negligible literature on consideration of disabled peoples in planning and
23 implementation of adaptation-related responses (medium confidence).
24
25 16.3.2.4 Adaptation Effectiveness, Adequacy, and Risk Reduction
26
27 Despite a lack of systematic methods for assessing general adaptation effectiveness, there is some evidence
28 of risk reduction for particular places and hazards, especially flood and heat vulnerability. There is some
29 evidence of a reduction in global vulnerability, particularly for flood risk (Jongman et al., 2015; Tanoue et
30 al., 2016; Miao, 2019) and extreme heat (Bobb et al., 2014; Boeckmann and Rohn, 2014; Gasparrini et al.,
31 2015; Arbuthnott et al., 2016; Chung et al., 2017; Sheridan and Allen, 2018; Folkerts et al., 2020).
32 Investment in flood protection, including building design and monitoring and forecasting, have reduced
33 flood-related mortality over time and are cost-effective (Bouwer & Jonkman 2018; Ward et al. 2017).
Do Not Cite, Quote or Distribute 16-30 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 Declining heat sensitivity, primarily reported in developed nations, has also been observed, and has been
2 linked to air conditioning, reduced social vulnerability, and improved population health (Boeckmann and
3 Rohn, 2014; Chung et al., 2017; Kinney, 2018; Sheridan and Allen, 2018). Formetta and Feyen (2019)
4 demonstrate declining global all-cause mortality and economic loss due to extreme weather events over the
5 past four decades, with the greatest reductions in low income countries, and with reductions correlated with
6 wealth. Studies that correlate changes in mortality or economic losses with wealth indicators, to infer
7 changes in vulnerability or exposure, lack direct empirical measures of vulnerability or exposure and are
8 limited in their ability to assess how indirect effects of extreme events (e.g., morbidity, relocation, social
9 disruption) may have changed or how changes may redistribute risk across populations.
10
11 There remain persistent difficulties in defining and measuring adaptation effectiveness and adequacy for
12 many climate risks. No studies have systematically assessed the adequacy and effectiveness of adaptation at
13 a global scale, across nations or sectors, or for different levels of warming. There has, however, been
14 progress in operationalizing assessment of adaptation feasibility (Cross-Chapter Box FEASIB in Chapter
15 18). Effectiveness of adaptation-related responses reflects whether a particular response actually reduces
16 climate risk, typically through reductions in vulnerability and exposure (Fig 1.7 in Section 1.4). Some
17 adaptation-related responses may increase risk or create new risks (maladaptation) or have no or negligible
18 impact on risk. Adequacy of adaptation-related responses refers to the extent to which responses are
19 collectively sufficient to reduce the risks or impacts of climate change (Fig 1.7 in Section 1.4). A set of
20 adaptation-related responses may, for example, result in reduced climate risk (effectiveness), but these
21 reductions may be insufficient to offset the level of risk and avoid loss and damages. Feasibility reflects the
22 degree to which climate responses are possible or desirable, and integrates consideration of potential
23 effectiveness. A feasibility assessment drawing on these methods is presented in (Cross-Chapter Box
24 FEASIB in Chapter 18).
25
26 Global adaptation is predominantly slow, siloed, and incremental with little evidence of transformative
27 adaptation (high confidence). In the absence of a general method to assess the adequacy of adaptation
28 actions, we assessed evidence for transformational adaptation documented in peer-reviewed publications
29 identified by a global stock-taking initiative (Berrang-Ford et al., 2021b) and in other AR6 chapters (2-15)
30 (see Supplemental Material, SM16.1 for details). `Transformational adaptation' refers to the degree to which
31 adaptations have been implemented widely (scope), reflect major shifts (depth), occur rapidly (speed), and
32 challenge limits to adaptation (limits, Pelling et al., 2015; Few et al., 2017; Termeer et al., 2017, Table 16.1).
33
34
35 Table 16.1: Evidence of transformational adaptation assessed across four components (depth, scope, speed,
36 and limits). Transformational adaptation does not imply adequacy or effectiveness of adaptation (low
37 transformation may be sufficient for some climate risks, and high transformation may be insufficient to
38 offset others). Nevertheless, these components provide a systematic framework for tracking adaptation
39 progress and assessing the state of adaptation-related responses. The `high' categories across each
40 component reflect more transformative scenarios. Methods are described in Supplementary Material
41 (SM16.1).
Transformative potential of adaptation
Dimensions Low Medium High
Overall Adaptation is largely sporadic Adaptation is expanding and Adaptation is widespread and
and consists of small increasingly coordinated, implemented at or very near its
adjustments to business-as- including wider implementation full potential across multiple
usual. Coordination and and multi-level coordination. dimensions.
mainstreaming are limited and
fragmented.
Do Not Cite, Quote or Distribute 16-31 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
Depth Adaptations are largely Adaptations reflect a shift away Adaptations reflect entirely new
Scope expansions of existing from existing practices, norms, or practices involving deep
practices, with minimal change structures to some extent. structural reform, complete
in underlying values, change in mindset, major shifts in
assumptions, or norms. perceptions or values, and
changing institutional or
behavioral norms.
Adaptations are largely Adaptations affect wider Adaptations are widespread and
localized and fragmented, with geographic areas, multiple areas substantial, including most
limited evidence of and sectors, or are mainstreamed possible sectors, levels of
coordination or mainstreaming and coordinated across multiple governance, and actors.
across sectors, jurisdictions, or dimensions.
levels of governance.
Speed Adaptations are implemented Adaptations are implemented Change is considered rapid for a
given context
slowly. moderately quickly.
Limits Adaptations may approach but Adaptations may overcome some Adaptations exceed many soft
do not exceed or substantively soft limits but do not challenge or limits and approach or challenge
challenge soft limits. approach hard limits. hard limits.
1
2
3 Based on the literature, the overall transformative nature of adaptation across most global regions and sectors
4 is low (high confidence) (Figure 16.6). Documented adaptations tend to involve minor modifications to usual
5 practices taken to address extreme weather conditions (high confidence). For example, changing crop variety
6 or timing of crop planting to address floods or droughts, new types of irrigation, pursuing supplementary
7 livelihoods, and home elevations are widely reported but typically do not reflect radical or novel shifts in
8 practice or values and are therefore considered low-depth (high confidence) (see Supplementary Material,
9 SM16.1 for more examples). Adaptations documented in the literature are also frequently focused on a single
10 sector or small geographic area (high confidence). Actions taken by individuals or households are generally
11 small in scope (Hintz et al., 2018; Hlahla and Hill, 2018) unless they are widely adopted (e.g., by farmers
12 across a region) or address numerous aspects of life. National policies are more likely to be broad in scope
13 (Puthucherril et al., 2014), although they frequently focus on a single sector and are therefore still limited.
14 The speed of adaptation is rarely noted explicitly, but the average speed documented in the literature is slow
15 (medium confidence) (Cross-Chapter Box FEASIB in Chapter 18). Adaptation efforts frequently encounter
16 either soft or hard limits (see Section 16.4), but there is limited evidence to suggest these limits are being
17 challenged or overcome (medium confidence).
18
19 Few documented responses are simultaneously widespread, rapid, and novel (high confidence). Some
20 examples exist, such as village relocations or creation of new multi-stakeholder resource governance systems
21 (Schwan and Yu, 2018; McMichael and Katonivualiku, 2020), but these are rare. In general, adaptations that
22 are broad in scope tend to be slow (medium confidence), suggesting that achieving high transformation in all
23 four categories (depth, scope, speed, and limits) may be particularly challenging or even involve trade-offs.
24
25
Do Not Cite, Quote or Distribute 16-32 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1
2 Figure 16.6. Evidence of transformative adaptation by sector and region. Evidence of transformational adaptation does
3 not imply effectiveness, equity, or adequacy. Evidence of transformative adaptation is assessed based on the scope,
4 speed, depth, and ability to challenge limits of responses reported in the scientific literature (see Supplementary
5 Material for methods). Studies relevant to multiple regions or sectors are included in assessment for each relevant
6 sector/region.
7
8
9 16.3.2.5 Observed Maladaptation and Co-benefits
10
11 There is increasing reporting of maladaptation globally (Table 16.2, Section 17.5.1) (high confidence).
12 Maladaptation has been particularly reported in the context of agricultural, forestry, and fisheries practices,
13 migration in the global south, and some infrastructure based-interventions. Urban heat adaptations have been
14 linked to maladaptation that increase health risks and/or energy consumption. Heat poses significant risks to
15 the evolutionary tolerance levels of humans, animals, and crops (Asseng et al., 2021), and current adaptation
16 interventions for reducing urban heat like cool or evaporation roofs and street trees may be insufficient to
17 reduce heat-related vulnerabilities in some urban areas at higher levels of warming (Krayenhoff et al., 2018)
18 (see also Section 16.4 on adaptation limits). There is evidence that autonomous adaptation by individuals and
19 households can shift risk to others, with net increases in vulnerability. Intensification of pasture use as a
20 coping response to climate-induced drought has been observed to increase risks to livestock reproduction and
21 human life expectancy due to overgrazing, suggesting responses to pastoral vulnerability can cross tolerance
22 limits for animals, humans, and food available for foraging (Suvdantsetseg et al., 2017).
23
24 Evidence on realized co-benefits of implemented adaptation responses with other priorities in the sustainable
25 development goals is emerging among the areas of poverty reduction, food security, health and well-being,
26 terrestrial and freshwater ecosystem services, sustainable cities and communities, energy security, work and
27 economic growth, and mitigation (Table 16.2) (high confidence). Evidence on co-benefits of adaptation for
28 mitigation is particularly strong, and is observed in various agricultural, forestry and land use management
29 practices like agroforestry, climate smart agriculture and afforestation (Kremen and Miles, 2012; Christen
30 and Dalgaard, 2013; Mbow et al., 2014; Locatelli et al., 2015; Suckall et al., 2015; Wichelns, 2016;
Do Not Cite, Quote or Distribute 16-33 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 Kongsager, 2018; Debray et al., 2019; Loboguerrero et al., 2019; Morecroft et al., 2019; Chausson et al.,
2 2020) as well as in the urban built environment (Perrotti and Stremke, 2020; Sharifi, 2020). Evidence on co-
3 benefits of implemented responses for other SDG priority areas is less developed, however, in the areas of
4 education, gender inequality and reduced inequalities, clean water and sanitation, industry, innovation and
5 infrastructure, consumption and production, marine and coastal ecosystem protection, and peace, justice, and
6 strong institutions. This indicates a gap between some assumed likely co-benefits of adaptation and empirical
7 evidence on the realization of these co-benefits within the context of implemented adaptation responses
8 (Berga, 2016; Froehlich et al., 2018; Gattuso et al., 2018; Morris et al., 2018; Chausson et al., 2020; Karlsson
9 et al., 2020; Krauss and Osland, 2020).
10
11
12 Table 16.2: Observed examples of maladaptation and co-benefits from adaptation-related responses in
13 human systems
Implemented Observed maladaptation References
adaptations
Agricultural & forestry practices
Intensified Increased competition for resources such as water Bele et al. (2014); D'haen et al. (2014);
cultivation of and nutrients; reduced soil fertility; invasive Chapman et al. (2016); Ifeanyi-obi et al.
marginal lands: species; degraded environment; increased (2017); Suvdantsetseg et al. (2017);
clearing of virgin greenhouse gas emissions; reduced crops Villamayor-Tomas and Garcia-Lopez (2017);
forests for farmland; diversity and reduced harvest, thus increasing Afriyie et al. (2018); Ticehurst and Curtis
frequent weeding; food insecurity in rural areas; accelerated illegal (2018); Tran et al. (2018); Neset et al. (2019);
poorly-managed logging practices; increased vulnerability of Work et al. (2019); Yamba et al. (2019);
irrigation schemes; herders and translated into poor health and Singh and Basu (2020)
dependence on working conditions (Mongolia)
rainfed agriculture
Agroforestry systems Higher water demand where trees were combined Nordhagen and Pascual (2013); D'haen et al.
with crops and livestock; replaced native trees (2014); Hoang et al. (2014); Ruiz-Mallen et
with non-indigenous trees; Reduced resilience of al. (2015); Kibet et al. (2016); Chengappa et
certain plants (e.g., cocoa); degraded soil and al. (2017); Haji and Legesse (2017); Abdulai
water quality and accelerated environmental et al. (2018); Antwi-Agyei et al. (2018);
degradation in Africa and Asia (Pakistan, Nepal, Mersha and van Laerhoven (2018); Ullah et
India, China, Philippines) al. (2018); Krishnamurthy et al. (2019)
Agricultural Soil degradation and high dependency on external Nordhagen and Pascual (2013); D'haen et al.
transitions: inputs in South and Central America (El- (2014); Warner et al. (2015); Kibet et al.
Commercialization Salvador, Guatemala, Honduras, Nicaragua, and (2016); (Warner and Kuzdas, 2016); Haji and
of common property; Peru); dependency on foreign corporation seed Legesse (2017); Antwi-Agyei et al. (2018);
market-integration systems; land enclosures. Adaptation that forced Mersha and van Laerhoven (2018);
and sedentarisation local farmers in Costa Rica to switch crops to Krishnamurthy et al. (2019); Neset et al.
of pastoralists; commercially viable products (e.g., from rice to (2019)
adoption and sugar cane) impoverished the land by removing
expansion of nutrients and affecting food security for
commercial crops smallholder farmers.
Proper, improper, Fertilizer and agrochemicals negatively affected Postigo (2014); Rodriguez-Solorzano (2014);
and increased use of soil quality and accelerated environmental Fezzi et al. (2015); Sujakhu et al. (2016);
agrochemicals, degradation in several parts of Africa (Ghana, Begum and Mahanta (2017); de Sousa et al.
pesticides, and Nigeria) and Asia (Pakistan, Nepal, India, China, (2018); Tang et al. (2018); Yamba et al.
fertilizers Philippines). In Europe (Sweden and Finland) (2019)
there are concerns about the risk of pests and
weeds developing immunity to pesticides, and
drainage systems and rain transferred chemicals
to other fields, thereby affecting arable land. In
South and Central America (El-Salvador,
Guatemala, Honduras, Nicaragua, and Peru)
agrochemicals led to soil degradation, and high
dependency on external input was reported. Loss
of soil nutrients, increased GHG emissions
(Sweden, Finland); high nitrate and phosphate
concentration (Great Britain)
Do Not Cite, Quote or Distribute 16-34 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
Tree planting The lack of shaded trees increased vulnerability to Benito-Garzon et al. (2013); Hoang et al.
landslides in areas where Robusta coffee was (2014); Ruiz Meza (2015); Chengappa et al.
grown (Mexico); new tree species to cope with (2017); Abdulai et al. (2018); Ullah et al.
climate change increased sensitivity and displaced (2018)
non-indigenous trees (India; Tanzania and
Kenya); Cocoa planted under shade trees had
higher mortality rate and more stress (Ghana);
Eucalyptus trees planted to reduce soil erosion
had high water demand (Pakistan); In certain
urban areas, trees planted to provide shade
damaged buildings during heavy storms.
Fisheries & water management
Increased fishing Fishery depletion and exacerbated negative trends Goulden et al. (2013); Mazur et al. (2013);
activity
in the ecosystem that threatened fishermen's Rodriguez-Solorzano (2014); Pershing et al.
subsistence (2016); Kanda et al. (2017); Kihila (2018);
Pinsky et al. (2018)
Shrimp farming A driver of deforestation of mangroves in Johnson et al. (2016); Jamero et al. (2017);
Bangladesh; imposes external cost on paddy Paprocki and Huq (2018); Sovacool (2018);
farmers; salinity levels are relatively higher in Morshed et al. (2020)
paddy plots closer to shrimp ponds. Coral mining
increased vulnerability to flooding (in small
islands in the Philippines)
Water irrigation Increased land loss; redistributed risk among Barnett and O'Neill (2013); Olmstead (2014);
infrastructure for agrarian stakeholders; affected the rural poor Warner and Kuzdas (2016); Work et al.
agriculture; water (Cambodia; Costa Rica); uneven distribution of (2019)
desalination in cost and benefits (US-Mexico border);
response to water Desalination plants to led disproportionately high
shortages cost for low income water users.
Storage of large Water rendered unsafe for drinking due Boelee et al. (2013); Trewin et al. (2013);
quantities of water in contamination by fecal coliforms in Zimbabwe; Kanda et al. (2017)
the home drought-induced changes in water harvesting and
storage increased breeding sites for mosquitoes
(Australia); Water storage facilities and tanks
provided ideal breeding conditions for
mosquitoes and flies bringing both vectors and
diseases closer to people (Ethiopia).
Increased number of Reduced river and ground water flow Mazur et al. (2013); Christian-Smith et al.
farm dams for water downstream; water grabs from shared surface or (2015); (Hurlbert and Mussetta, 2016); Work
storage; groundwater groundwater resources with poorly defined et al.)
extraction and property rights shifted vulnerability to other
interbasin water groups and ecosystems (Cambodia; California):
transfers water extractions increased risks for the
environment and food security, while transfers
reduced hydropower generation and resulted in
higher costs paid by electricity consumers and
health impacts from air pollution caused by more
electricity generation from natural gas
(California); increase the concentration in hands
of the more powerful large farmers (Argentina)
Built environment
Do Not Cite, Quote or Distribute 16-35 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
Seawalls and Coastal erosion, beach losses, changes in water Macintosh (2013); Maldonado et al. (2014);
infrastructural current, and destruction of natural ecosystems in Porio (2014); Betzold (2015); Renaud et al.
development along Asia, Australasia, Europe, and North America; (2015); Gundersen et al. (2016); Sayers et al.
coastlines increased or shifted erosion from protected to (2018); Craig (2019); Javeline and Kijewski-
unprotected areas in Fiji, Marshall Islands, Nuie, Correa (2019); Loughran and Elliott (2019);
Smart or green Kiribati, Norway; failed or sped up flood waters Rahman and Hickey (2019); Piggott-
luxury real estate and worsened conditions for riparian habitat and McKellar et al. (2020); Simon et al. (2020)
development downstream residents; harmed nearby reefs and Dahl et al. (2017)
designed to reduce impeded autonomous adaptation practise that
impacts from storm could be effective (Bangladesh).
surges and erosion
along coastal area; Redistributed risk and vulnerability; displaced Caprotti et al. (2015); Magnan et al. (2016);
artificial islands. and diminished adaptive capacity of vulnerable Atteridge and Remling (2018); Ajibade
groups, created new population of landless (2019); Salim et al. (2019); Thomas and
peasants; negatively affected neighbouring Warner (2019)
coastal areas and local ecology (Lagos, Miami,
Hanoi, Jakarta, Manila; Maldives)
Subsidized insurance Rebuilding in risky areas Shearer et al. (2014); O'Hare et al. (2016);
premiums for Craig (2019); Loughran and Elliott (2019)
properties located in
flood-prone areas,
levees, dykes
Autonomous flood Sand bags used to reduce coastal erosion released Schaer (2015); Wamsler and Brink (2015);
strategies such as plastics into the sea and led to loss of recreational (Chapman et al., 2016); Magnan et al. (2016);
sand bags, digging value of beaches; sand walls shifted the flood Mycoo (2018); Rahman and Hickey (2019)
channels and sand impacts across space and time and were more
walls around homes. detrimental to poor informal urban settlers
(Dakar); caused erosion and degraded coastal
lands (South Africa).
Top-down Ignored the complexities of the landscapes and Cartwright et al. (2013); Goulden et al.
technocratic socio-ecological systems; constrained (2013); Nordhagen and Pascual (2013); Carr
adaptation with no autonomous adaptation due to time and labour and Thompson (2014); Nyamadzawo et al.
consideration for demands of public work; increased gender (2015); Ruiz-Mallen et al. (2015); Djoudi et
ecosystem vulnerability; hamper women's water rights al. (2016); Gautier et al. (2016); Gundersen et
biodiversity, local (South Africa); altered local gender norms al. (2016); Barnett and McMichael (2018);
adaptive capacity, (Ethiopia); led to a mismatch that undermine Kihila (2018); Mersha and van Laerhoven
and gender issues local-level processes that are vital to local (2018); Clay and King (2019); Currenti et al.
adaptive capacity (Rwanda) (2019); Yang et al. (2019)
Migration & relocation
Out-migration or Migration mostly undertaken by poorer household Su et al. (2017);Aziz and Sadok
rural to urban weakened local subsistence production capacity; (2015);Bhatta and Aggarwal (2016);Clay and
migration in response disrupted family structures; reduced labour King (2019); Elagib et al. (2017);Gao and
to food insecurity and available for agricultural work; increased burden Mills (2018); Kattumuri et al. (2017);
agricultural of responsibilities on women; fostered loss of Magnan et al. (2016); Ofoegbu et al. (2016);
livelihood solidarity within communities; increased divorce Rademacher-Schulz et al.
depreciation rates; exacerbated conflicts among different (2014);Rademacher-Schulz et al.
groups; increased pressure on urban housing and (2014);Wiederkehr et al. (2018); Yegbemey
social services; expanded slum settlements around et al. (2017); Yila and Resurreccion (2013);
riparian and coastal areas including flood plains Nizami et al. (2019); Mersha and Van
and swamplands (Ethiopia, Namibia, Benin, Laerhoven (2016); Ojha et al. (2014); Radel
Botswana, Nigeria, Ghana, Kenya, Niger, Mail, et al. (2018); Gioli et al. (2014); Hooli
Tanzania, Zimbabwe, South Africa, Morocco, (2016); Koubi et al. (2016)
Nepal, Pakistan, Bangladesh China, India,
Australia, Nicaragua). Out-migration from small
communities had devastating consequences on
their fragile economies, thereby reducing
community resilience in the long term (Australia).
Do Not Cite, Quote or Distribute 16-36 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
Certain autonomous, Expansion of informal settlements in cities Monnereau and Abraham (2013); Maldonado
forced, and planned (Solomon Islands); relocation to areas prone to et al. (2014); Pritchard and Thielemans
relocation landslide and soil erosion or insufficient housing (2014); Averchenkova et al. (2016); Lei et al.
(Fiji); disproportionate burden on vulnerable (2017); Barnett and McMichael (2018);
Temporary communities (China); temporary relocation Currenti et al. (2019)
resettlement (India) created gender inequality associated with minimal
privacy; poor access to private toilets; sexual
harassment; reduced sleep; insufficient or food
rationing; exploitation and abuse of children
(India); inadequate funding and governance
mechanism for community-based relocation
caused loss of culture, economic decline and
health concerns (Alaska); relocation of supply
chain to reduce exposure to climate change
resulted in adverse outcomes for communities
along the supply chain.
Implemented Observed co-benefits References
adaptations
Agricultural practices
Integrated Mitigation, especially carbon sequestration (but Furman et al. (2014); Lwasa et al. (2014);
Kibue et al. (2015); Nyasimi et al. (2017);
agricultural practices see (Sommer et al., 2018)); improved household Aryal et al. (2018); Han et al. (2018);
Kakumanu et al. (2018); Sikka et al. (2018);
(e.g., climate smart equity regarding farming decisions, particularly Debray et al. (2019); Kerr et al. (2019);
(Teklewold et al., 2019a); Teklewold et al.
agriculture, urban inclusion of women; food security (2019b); Wang et al. (2020) Sommer et al.
(2018)
and peri-urban
Islam et al. (2020)
agriculture and
forestry;
agroecology;
silvopasture; soil
desalinization;
drainage
improvement;
integrated soil-crop
system management;
no tillage farming;
rainwater harvesting;
check dams)
Improved irrigation Mitigation, especially avoided emissions;
systems improved crop yields
Conservation Mitigation, especially carbon sequestration; Helling et al. (2015); Sapkota et al. (2015);
agriculture (e.g.crop
diversification; soil increased crop yields; food security; reduced heat Kimaro et al. (2016); Mainardi (2018);
conservation; cover
cropping) and water stress; increased food security Asmare et al. (2019); Gonzalez-Sanchez et al.
(2019)
Return to traditional Mitigation, especially carbon sequestration Pienkowski and Zbaraszewski (2019)
farming practices
Place-specific Mitigation, especially carbon sequestration; Sushant (2013); Balaji et al. (2015); Helling
et al. (2015); Jorgensen and Termansen
practices & improved crop yields; food security (2016); Sen and Bond (2017); Wilkes et al.
(2017); Kakumanu et al. (2018); Mainardi
innovations: animal (2018); Sikka et al. (2018) Yadav et al.
(2020)
cross-breeding; direct
crop seeding; site-
specific nutrient
management;
irrigation
innovations; use of
riparian buffer strips;
Do Not Cite, Quote or Distribute 16-37 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
use of green winter
land; rice-rice system
Land and water management
Agroforestry Mitigation, especially carbon sequestration; Holler (2014); Suckall et al. (2015); Sharma
biodiversity and ecosystem conservation; et al. (2016); Nyasimi et al. (2017); Pandey et
improved food security; plant species al. (2017); Schembergue et al. (2017); Ticktin
diversification; diversification of household et al. (2018); Debray et al. (2019); Jezeer et
livelihoods; improved household incomes; al. (2019); Krishnamurthy et al. (2019);
improved access to forage material; energy access Nyantakyi-Frimpong et al. (2019); Tschora
and reduced fuel wood gathering time and and Cherubini (2020)
distance for women; soil and water conservation;
aesthetic improvements in landscapes
Afforestation and Mitigation, especially carbon sequestration; Holler (2014); Etongo et al. (2015);
reforestation biodiversity and ecosystem conservation; new Diederichs and Roberts (2016); Acevedo-
programs; employment opportunities; diversification of Osorio et al. (2017); Nyasimi et al. (2017);
Forest management household livelihoods; increased household Krishnamurthy et al. (2019); Rahman et al.
practices (e.g., tree incomes; improved access to fuel wood; (2019) Wolde et al. (2016)
thinning) harvesting opportunities from enclosures
Ecosystem-based Mitigation, especially carbon sequestration; Fedele et al. (2018)
Roberts et al. (2012); Morris et al. (2019);
adaptations like habitat enhancement and protection for marine (Jones et al., 2020)
mangrove restoration species; prevention of floor-related deaths,
and natural coastal injuries, and damage; improved nutrition and
defences income generation for local communities,
improved water quality
Sustainable water Mitigation, especially avoided emissions; reduced Spencer et al. (2017); Siraw et al. (2018);
management
water demand; increased awareness about Stanczuk-Galwiaczek et al. (2018)
impacts of water consumption; decreased
incidence of fecaloral disease transmission;
decreased use of drinking water for irrigation;
reduced soil loss; increased groundwater
retention; increased vegetation cover; increased
food security and health and well-being;
increased forage for livestock and amount of
cultivated area; enhanced recreational areas
Return to traditional Mitigation, especially carbon sequestration; Duguma et al. (2014)
land management increased water availability for household and
practices (e.g., the livestock use; increase in presence of edible and
Ngitili system) medicinal plants; regional economic growth;
reduced land management conflicts; increased
household income and access to education for
children; improved access to wood fuel and
reduced collection time for women; improved
wildlife habitat.
REDD+ participation Mitigation, especially carbon sequestration; McElwee et al. (2017); Spencer et al. (2017)
to maintain intact improved air quality; water and soil conservation;
forest ecosystems slowed rate of vector-borne disease; improved
mental well-being associated with cultural
continuity; clean water; nutritional and spiritual
value of forest-derived foods; protection from
violence related to natural resource extraction
Urban planning and design
Do Not Cite, Quote or Distribute 16-38 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
Spatial planning Mitigation, particularly avoided emissions; public Beiler et al. (2016); Belanger et al. (2016)
walkable health increases in physical activity, reductions
neighbourhood in air pollution and urban heat island effect
design; strategic
densification.
Urban greening (e.g Mitigation, particularly avoided emissions; public Samora-Arvela et al. (2017); Vahmani and
tree planting; health improvements increases in physical Jones (2017); Newell et al. (2018); Alves et
construction of activity, reductions in air and noise pollution, al. (2019); De la Sota et al. (2019)
stormwater retention reduced urban heat island effect, improved mental
areas; construction of health; urban flood risk management; water
green roofs and cool savings; energy savings
roofs; provision of
rainwater barrels;
pervious pavement
materials)
Improved building Mitigation, particularly avoided emissions; Barbosa et al. (2015); Koski and Siulagi
efficiency standards improved air quality; reduced urban heat island; (2016); Balaban and Puppim de Oliveira
improved natural indoor lighting (2017); Landauer et al. (2019)
Use of local building Mitigation, particularly avoided emissions Lundgren-Kownacki et al. (2018)
materials
1
2
3 16.3.3 Knowledge Gaps in Observed Responses
4
5 Many adaptation responses are not documented, and reporting bias is a key challenge for assessment of
6 observed responses. Evidence of absence (i.e., where no adaptations are occurring) is different from absence
7 of evidence (where responses are occurring but are not documented), with implications for understanding
8 trends in global responses.
9
10 Adaptation is being reported differently across different sources of knowledge. The peer-reviewed
11 literature, for example, has been primarily reporting reactive adaptation at the individual, household, and
12 community levels, while the grey literature has been more mixed, reporting adaptation across governmental
13 levels and civil society, with less focus on individuals and households (Ford et al., 2015a; Ford and King,
14 2015). Synthesis of impacts and responses within the private sector is particularly limited (Averchenkova et
15 al., 2016; Minx et al., 2017), further suggesting that knowledge accumulation on climate responses has been
16 particularly slow, and that more robust evidence synthesis is required to fill key knowledge gaps.
17
18 The potential for under-reporting is most acute in the context of minorities, remote and marginalized
19 groups, who are often also be the most affected by the impacts of climate change and least able to respond
20 to, or benefit from, the responses to, climate change (Araos et al., 2021). Deficits in reporting on impacts and
21 responses are well-recognized in the global south, among vulnerable populations (e.g., women, socio-
22 economically disadvantaged, indigenous, people living with disabilities), and within civil society (ibid.).
23
24 There is growing support for more comprehensive and systematic approaches to assess adaptation
25 progress (Berrang-Ford et al., 2015; Ford et al., 2015a; Ford and King, 2015; Ford and Berrang-Ford, 2016;
26 Biesbroek et al., 2018). Since AR5, there is increased recognition of the value of integrating diverse
27 knowledge sources to fill knowledge gaps in observation of impacts and responses (Chapter 17; Cross-
28 Chapter Box PROGRESS in Chapter 17). Van Bavel, for example, found that the involvement of local and
29 diverse knowledge can improve the detection (medium confidence) and attribution (medium confidence) of
30 health impacts, and improve the action (high confidence) (Van Bavel et al., 2020).
31
32
33 [START CROSS-CHAPTER BOX INTEREG HERE]
34
35 Cross-Chapter Box INTEREG: Inter-regional Flows of Risks and Responses to Risk
36
Do Not Cite, Quote or Distribute 16-39 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 Authors: Birgit Bednar-Friedl (Austria, Chapter 13), Christopher Trisos (South Africa, Chapter 9), Laura
2 Astigarraga (Uruguay, Chapter 12), Magnus Benzie (Sweden/United Kingdom), Aditi Mukherji (India,
3 Chapter 4), Maarten Van Aalst (The Netherlands, Chapter 16)
4
5 Introduction
6
7 Our world today is characterized by a high degree of interconnectedness and globalization which establish
8 pathways for the transmission of climate-related risks across sectors and borders (high confidence)
9 (Challinor et al., 2018; Hedlund et al., 2018). While AR5 has pointed to this connection of risks across
10 regions as `cross-regional phenomena' (Hewitson et al., 2014), only a few countries so far have integrated
11 interregional aspects into their climate change risks assessments (Liverman, 2016; Surminski et al., 2016;
12 Adams et al., 2020) and adaptation is still framed as a predominantly national or local issue (Dzebo and
13 Stripple, 2015; Benzie and Persson, 2019).
14
15 Interregional risks from climate change - also called cross-border, transboundary, transnational or indirect
16 risks - are risks that are transmitted across borders (e.g., transboundary water use) and/or via teleconnections
17 (e.g., supply chains, global food markets) (Moser and Hart, 2015). The risks can result from impacts,
18 including compound or concurrent impacts, that cascade across several tiers, in ways that either diminish or
19 escalate risk within international systems (Carter et al., 2021). Risk transmission may occur through trade
20 and finance networks, flows of people (Cross-Chapter Box MIGRATE in Chapter 7), biophysical flows
21 (natural resources such as water) and ecosystem connections. But not only risks are transmitted across
22 borders and systems, but also the adaptation response may reduce risks at the origin of the risk, along the
23 transmission channel or at the recipient of the risk (Carter et al., 2021). This Cross-Chapter Box discusses
24 four interregional risk channels (trade, finance, food, and ecosystems) and how adaptation can govern these
25 risks.
26
27
Do Not Cite, Quote or Distribute 16-40 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1
2 Figure Cross-Chapter Box INTEREG.1: Interregional climate risks: the example of the trade transmission channel,
3 illustrated for the Thailand flood 2011 (Abe and Ye, 2013; Haraguchi and Lall, 2015; Carter et al., 2021)
4
5
6 Trade
7
8 Most commodities are traded on global markets and supply chains have become increasingly globalized. For
9 instance, specialized industrial commodities like semi-conductors are geographically concentrated in a few
10 countries (Challinor et al., 2017) (Liverman, 2016). When climatic events like flooding or heat affect the
11 location of these extraction and production activities, economies are not only disrupted locally but also
12 across borders and in distant countries (high confidence), as exemplified by the Thailand flood 2011 that led
13 to a shortage of key inputs to the automotive and electronics industry not only in Thailand but also in Japan,
14 Europe, and the USA (Figure Cross-Chapter Box INTEREG.1). For many industrialized countries like the
15 United Kingdom, Japan, the USA and the European Union, there is increasing evidence that the trade
16 impacts of climate change are significant and can have substantial domestic impacts (medium confidence)
17 (Nakano, 2017; Willner et al., 2018, Section 13.9.1; Benzie and Persson, 2019; Knittel et al., 2020).
18 Enhanced trade can transmit risks across borders and thereby amplify damages (Wenz and Levermann,
19 2016), but it can also increase resilience (Lim-Camacho et al., 2017; Willner et al., 2018).
Do Not Cite, Quote or Distribute 16-41 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1
2 Finance
3
4 Climate risks can also spread through global financial markets (Mandel et al., 2021). For the case of coastal
5 and riverine flooding with low adaptation 2080 (RCP 8.5-SSP5), the financial system is projected to amplify
6 direct losses by a factor of 2 (global average), but reach up to a factor of 10 for countries that are central
7 financial hubs (Mandel et al., 2021, Figure 13.28). Indirect impacts may also arise through indirect effects on
8 foreign direct investment, remittance flows, and official development assistance (Hedlund et al., 2018).
9
10 Food
11
12 The global supply of agricultural products is concentrated to a few main breadbaskets (Bren d'Amour et al.,
13 2016; Gaupp et al., 2020, Chapter 5). For instance, Central and South America is one of the regions with the
14 highest potential to increase food supplies to more densely populated regions in Asia, Middle East and
15 Europe (Chapter 12). The exports of agricultural commodities (coffee, bananas, sugar, soybean, corn,
16 sugarcane, beef livestock) have gained importance in the past two decades as international trade and
17 globalization of markets have shaped the global agri-food system (Chapter 5).
18
19 The export of major food crops like wheat, maize and soybeans from many of the world's water scarce areas
20 Middle East, North Africa, parts of South Asia, North China Plains, south-west USA, Australia to
21 relatively water abundant parts of the world carries a high virtual water content (the net volume of water
22 embedded in trade) (high confidence) (Hoekstra and Mekonnen, 2012; Dalin et al., 2017; Zhao et al., 2019,
23 Chapter 4). Both importing and exporting countries are exposed to transboundary risk transmission through
24 climate change impacts on distant water resources (Sartori et al., 2017; Zhao et al., 2019; Ercin et al., 2021).
25 Climate change is projected to exacerbate risk and add new vulnerabilities for risk transmission (medium
26 confidence). Rising atmospheric CO2 concentration is projected to decrease water efficiency of growing
27 maize and temperate cereal crops in parts of USA, East and Mediterranean Europe, South Africa, Argentina,
28 Australia and South East Asia with important implications for future trade in food grains (Fader et al., 2010).
29 By 2050 (SRES B2 scenario) virtual water importing countries in Africa and the Middle East may be
30 exposed to imported water stress as they rely on imports of food grains from countries which have
31 unsustainable water use (Sartori et al., 2017). Until 2100, virtual trade in irrigation water is projected to
32 almost triple (for SSP2-RCP6.5 scenarios) and the direction of virtual water flows is projected to reverse
33 with the currently exporting regions like South Asia becoming importers of virtual water (Graham et al.,
34 2020). An additional 10-120% trade flow from water abundant regions to water scarce regions will be
35 needed to sustain environmental flow requirements on a global scale by end of the century (Pastor et al.,
36 2019). Exports of agricultural commodities contribute to deforestation, over-exploitation of natural resources
37 and pollution, affecting the natural capital base and ecosystem services (Agarwala and Coyle, 2020; Rabin et
38 al., 2020, Section 12.5.4).
39
40 Species and ecosystems
41
42 The spatial distributions of species on land and in the oceans are shifting due to climate change, with these
43 changes projected to accelerate at higher levels of global warming (Pecl et al., 2017). These `species on the
44 move' have large effects on ecosystems and human well-being, and present challenges for governance (Pecl
45 et al., 2017). For example, the number of transboundary fish stocks are projected to increase as key fisheries
46 species are displaced by ocean warming (Pinsky et al., 2018). Conflict over shifting mackerel fisheries has
47 already occurred between European countries (Spijkers and Boonstra, 2017), because few regulatory bodies
48 have clear policies on shifting stocks; this leaves species open to unsustainable exploitation in new waters in
49 the absence of regularly updated catch allocations to reflect changing stock distributions (Caddell, 2018).
50
51 Human health will also be affected as vector-borne diseases such as malaria and dengue shift geographic
52 distributions (Caminade et al., 2014). There is also evidence that many warm-adapted invasive species, such
53 as invasive freshwater cyanobacterium, have spread to higher latitudes due to climate change (Chapter 2).
54
55 Adaptation to interregional climate risks
56
Do Not Cite, Quote or Distribute 16-42 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 Adaptation responses to reduce interregional risks can be implemented at a range of scales: at the point of the
2 initial climate change impact (e.g. assistance for recovery after an extreme event, development of resilient
3 infrastructure, climate-smart technologies for agriculture); at or along the pathway via which impacts are
4 transmitted to the eventual recipient (e.g. trade diversification, re-routing of transport); in the recipient
5 country (e.g. increasing storage to buffer supply disruptions), or by third parties (e.g. adaptation finance,
6 technology transfer) (Bren d'Amour et al., 2016; Carter et al., 2021; Talebian et al., 2021). A knowledge gap
7 exits on the need for, effectiveness of, and limits to adaptation under different socio-economic and land-use
8 futures.
9
10 Due to regional and global interdependencies, climate resilience has a global, multi-level public good
11 character (Banda, 2018). The benefits of adaptation are therefore shared beyond the places where adaptation
12 is initially implemented. Conversely, adaptation may be successful at a local level whilst redistributing
13 vulnerability elsewhere or even driving or exacerbating risks in other places (Atteridge and Remling, 2018).
14 International cooperation is therefore needed to ensure that inter-regional effects are considered in adaptation
15 and that adaptation efforts are coordinated to avoid maladaptation. However, regional and global scale
16 governance of adaptation is only just beginning to emerge (Persson, 2019).
17
18 The UNFCCC Paris Agreement frames adaptation as a `global challenge' (Article 7.2) and establishes the
19 global goal on adaptation (Article 7.1), which provides space for dialogue between Parties on the global
20 scale challenge of adaptation and the need for renewed political and financial investment in adaptation,
21 including to address interregional effects (Benzie et al., 2018).
22
23 National Adaptation Plans (NAPs) can evolve to consider inter-regional effects as well as domestic ones
24 (Liverman, 2016; Surminski et al., 2016; European Environment, 2020). Regional and international
25 coordination of NAPs, coupled with building capacities and addressing existing knowledge gaps at the
26 country level, can help to ensure that resources are oriented towards reducing interregional risks and building
27 systemic resilience to climate change globally (Booth et al., 2020; Wijenayake et al., 2020).
28
29 Given the important role of private actors in managing interregional climate risks (Goldstein et al., 2019;
30 Tenggren et al., 2019), efforts will be needed to align public and private strategies for managing
31 interregional climate risks to avoid maladaptation and ensure just and equitable adaptation at different scales
32 (Talebian et al., 2021).
33
34 [END CROSS-CHAPTER BOX INTEREG HERE]
35
36
37 A new development since AR5, there is now growing evidence assessing progress on adaptation across
38 sectors, geographies and spatial scales. Uncertainty persists around what defines adaptation and how to
39 measure it (Cross-Chapter Box FEASIB in Chapter 18, UNEP, 2021). As a result, most literature
40 synthesizing responses are based on documented or reported adaptations only, and are thus subject to
41 substantial reporting bias.
42
43 We document implemented adaptation-related responses that could directly reduce risk. Adaptation as a
44 process is more broadly covered in Chapter 17 (Section 17.4.2), including risk management, decision-
45 making, planning, feasibility (see Cross-Chapter Box FEASIB in Chapter 18), legislation and learning. Here,
46 we focus on a subset of adaptation activities: adaptation-related responses of species, ecosystems, and human
47 societies that have been implemented, observed, and could directly reduce risk. We consider all adaptation-
48 related responses to assumed, perceived, or expected climate risk, regardless of whether or not impacts or
49 risks have been formally attributed to climate change.
50
51 We use the term `adaptation-related responses', recognising that not all responses reduce risk. While
52 `adaptation' implies risk reduction, we use the broader term `responses' to reflect that responses may
53 decrease risk, but in some cases may increase risk.
54
55 Given limited evidence to inform comprehensive global assessment of effectiveness and adequacy, we assess
56 evidence that adaptation responses in human systems indicate transformational change. Chapter 17 considers
57 adaptation planning and governance, including adaptation solutions, success, and feasibility assessment
Do Not Cite, Quote or Distribute 16-43 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 (Cross-Chapter Box FEASIB in Chapter 18). It is not currently possible to conduct a comprehensive global
2 assessment of effectiveness, adequacy, or the contribution of adaptation-related responses to changing risk
3 due to an absence of robust empirical literature (discussed further in Cross-Chapter Box PROGRESS in
4 Chapter 17).
5
6 In natural ecosystems or species, detectable changes can be considered as `impact' or `response'. The
7 distinction between `observed impacts' (16.2) and `observed responses' (16.3) is not always clear. For
8 example, autonomous distributional shifts in wild species induced by increasing temperatures (an observed
9 impact) may reduce risk to the species (an autonomous adaptation response), but this process can be
10 enhanced or supported by human intervention such as intentional changes in land use. Observed autonomous
11 changes in natural ecosystems or species unsupported by human intervention are treated as impacts (see
12 Section 16.2).
13
14 Adaptation-related responses are frequently motivated by a combination of climatic and non-climatic drivers,
15 and interact with other transitions to affect risk. For societal responses, it is difficult to say whether they are
16 triggered by observed or anticipated changes in climate, by non-climatic drivers, or as is the case in many
17 societal responses, a combination of all three. In the case of impacts, assessment typically focuses on
18 detection and attribution vis a vis a counterfactual of no climate change. While there has been some effort to
19 attribute reduced climate risk to adaptation-related responses (Toloo et al., 2013a; Toloo et al., 2013b; Hess
20 et al., 2018; Weinberger et al., 2018), in many cases this has not been feasible given difficulties in defining
21 adaptation and empirically disentangling the contribution of intersecting social transitions and changing
22 risks. Literature on adaptation-related response frequently draws on theories-of-change to assess the likely
23 contribution of adaptations to changes in risk, including maladaptation and co-benefits.
24
25
26 16.4 Synthesis of Limits to Adaptation Across Natural and Human Systems
27
28 This section builds on previous IPCC Reports (i.e., AR5, SR15, SROCC, SRCCL) to advance concepts and
29 emphasize remaining gaps in understanding about limits to adaptation. We provide case studies to illustrate
30 these concepts and synthesize regional and sectoral limits to adaptation across natural and human systems
31 that informs key risks (Section 16.5) and Reasons for Concern (Section 16.6). We also identify residual risks
32 - risks that remain after efforts to reduce hazards, vulnerability, and/or exposure - associated with limits to
33 adaptation.
34
35 16.4.1 Definitions and Conceptual Advances Since AR5
36
37 16.4.1.1 Limits to Adaptation since AR5
38
39 AR5 introduced the concept of limits to adaptation and provided a functional definition that has been used in
40 subsequent Special Reports (SR15, SROCC, SRCCL) and is also used for AR6 (see also Chapter 1).
41
42 A limit is defined as the point at which an actor's objectives or system's needs cannot be secured from
43 intolerable risks through adaptive actions (Klein et al., 2014). Tolerable risks are those where adaptation
44 needed to keep risk within reasonable levels is possible, while intolerable risks are those where practicable or
45 affordable adaptation options to avoid unreasonable risks are unavailable. This highlights that limits to
46 adaptation are socially constructed and based on values that determine levels of reasonable or unreasonable
47 risk as well as on available adaptation options, which vary greatly across and within societies.
48
49 Limits are categorized as being either `soft' or `hard'. Soft limits may change over time as additional
50 adaptation options that are practicable or affordable become available. Hard limits will not change over time
51 as no additional adaptive actions are possible. When a limit is exceeded, then intolerable risk may
52 materialize and the actor's objectives or system's needs may be either abandoned or transformed (Figure
53 Box16.1.1).
54
55 For human systems, soft and hard limits are largely distinguished by whether or not constraints to adaptation
56 are able to be overcome. Constraints to adaptation (also called barriers) are factors that make it harder to plan
57 and implement adaptation actions such as limited financial resources, ineffective institutional arrangements
Do Not Cite, Quote or Distribute 16-44 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 or insufficient human capacity. Soft limits are mostly associated with human systems, due in part to the role
2 of human agency in addressing constraints. For natural systems, the magnitude and rate of climate change
3 and capacity of adaptation to such change largely determine the type of limit. Hard limits are largely
4 associated with natural systems and are mostly due to inability to adapt to biophysical changes.
5
6 Using this understanding of limits, subsequent Special Reports have assessed relevant literature (Mechler et
7 al., 2020). SR15 identifies several regions, sectors and ecosystems including coral reefs, biodiversity,
8 human health, coastal livelihoods, small island developing states, and the Arctic that are projected to
9 experience limits at either 1.5ºC or 2ºC. SRCCL states that land degradation due to climate change may
10 result in limits to adaptation being reached in coastal regions and areas affected by thawing permafrost.
11 SROCC details that risks of climate-related changes in the ocean and cryosphere may result in limits for
12 ecosystems and vulnerable communities in coral reef environments, urban atoll islands and low-lying Arctic
13 locations before the end of this century in case of high emissions scenarios.
14
15 A key area of advancement since AR5 is how incremental and transformational adaptation relate to limits to
16 adaptation. Incremental adaptation maintains `the essence and integrity of a system or process at a given
17 scale' while transformational adaptation `changes the fundamental attributes of a social-ecological system'
18 (Matthews, 2018). Both incremental and transformational adaptation may expand the adaptive possibilities
19 for a system, providing additional adaptation options after a system reaches a soft limit (Felgenhauer, 2015;
20 Pelling et al., 2015; Termeer et al., 2017, see also Chapter 1 and 17; Alston et al., 2018; Panda, 2018;
21 Mechler and Deubelli, 2021). However, it is critical to note that adaptation, whether incremental or
22 transformational, must support securing an actor's objectives or system's needs from intolerable risks. Once
23 objectives or needs have been abandoned or transformed, a limit to adaptation has occurred. However,
24 objectives or needs may change over time as values of a society change (Taebi et al., 2020), thus adding
25 further complexity to assessing limits to adaptation.
26
27 16.4.1.2 Residual risk since AR5
28
29 The term `residual risk' was not assessed in detail in AR5 and was used interchangeably with other terms
30 including `residual impacts', `residual loss and damage' and `residual damage'. SR15 includes discussion of
31 residual risks without an explicit definition and relates these to loss and damage and limits to adaptation,
32 concluding that residual risks rise as global temperatures increase from 1.5°C to 2°C. SRCCL refers to
33 residual risks arising from limits to adaptation related to land management. Such residual risk can emerge
34 from irreversible forms of land degradation, such as coastal erosion when land completely disappears,
35 collapse of infrastructure due to thawing of permafrost, and extreme forms of soil erosion. SROCC advanced
36 the conceptualization of residual risk and integrated it within the risk framework, defining residual risk as the
37 risk that remains after actions have been taken to reduce hazards, exposure and/or vulnerability. Residual
38 risk is therefore generally higher where adaptation failure, insufficient adaptation or limits to adaptation
39 occur. We use the SROCC definition of residual risk for our assessment in the following sections and
40 identify residual risks that are associated with limits to adaptation.
41
42
43 [START BOX 16.1 HERE]
44
45 Box 16.1: Linking Adaptation Constraints, Soft and Hard Limits
46
47 McNamara et al. (2017) provides an example of community-scaled adaptation that highlights how
48 constraints affect limits, the relationship between soft and hard limits, and the potential need to abandon or
49 transform objectives. In Boigu Island, Australia, community members are already adapting to perceived
50 climate change hazards - including sea level rise and coastal erosion - to secure their objective of sustaining
51 livelihoods and way of life in their current location. Existing seawall and drainage systems provide
52 inadequate protection from flooding during high tides, leading residents to elevate their houses to prevent
53 damages. However, these adaptation measures have proved to be insufficient. Standing saltwater for
54 extended periods of time after floods has resulted in losses and damages including erosion of
55 infrastructure, increased soil salinity, and heightened public health concerns. Additional adaptation efforts
56 are constrained by scarcity of elevated land which inhibits movement of infrastructure within the community
57 and lack of financial, technical and human assets to improve coastal protection measures.
Do Not Cite, Quote or Distribute 16-45 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1
2 These constraints are leading to a soft limit to adaptation where risks would become unreasonable as sea
3 levels continue to rise and practicable and affordable adaptation options are limited to currently available
4 approaches. This soft limit could be overcome through addressing constraints and allowing further
5 adaptation to take place, such as providing financial, technical and human resources for more effective
6 coastal protection and drainage systems that would reduce flooding. However, if the effectiveness of these
7 new adaptation measures decreases as sea levels rise further and if constraints are not able to be overcome,
8 another soft limit may be reached. Eventually, if constraints are not addressed, no further adaptation
9 measures are implemented and climate hazards intensify, the area could become uninhabitable. This would
10 then be a hard limit for adaptation there would be no adaptation options available that would allow the
11 community to sustain livelihoods and way of life in its present location. This hard limit to adaptation may
12 necessitate abandoning the objective of remaining in the community. The objective of the community may
13 then transform to sustaining their livelihoods in a less vulnerable location which would necessitate
14 relocation. However, such transformation of the community's objectives may be hindered by the expressed
15 resistance of residents to migrate, due to their strong sense of place.
16
17 [END BOX 16.1 HERE]
18
19
20 16.4.2 Insights from Regions and Sectors about Limits to Adaptation
21
22 Here we provide example case studies to highlight constraints that may lead to soft limits, potential
23 incremental and transformational adaptation options that may overcome soft limits, evidence of hard limits
24 and residual risks.
25
26 16.4.2.1 Small Island Developing States (SIDS)
27
28 An expanding volume of empirical research highlights existing adaptation constraints that may lead to soft
29 limits in SIDS. Investigation of national communications among 19 SIDS found that financial constraints,
30 institutional challenges and poor resource endowments were the most-frequently reported as inhibiting
31 adaptation for a range of climate impacts (Robinson, 2018b). Governance, financial and information
32 constraints such as unclear property rights and lack of donor flexibility have led to hasty implementation of
33 adaptation projects in Kiribati, whereas in Vanuatu and the Solomon Islands, limited awareness of rural
34 adaptation needs and weak linkages between central governance and local communities have resulted in an
35 urban bias in resource allocation (Kuruppu and Willie, 2015). Limited availability and use of information
36 and technology also present constraints to adaptation many SIDS suffer from lack of data and established
37 routines to identify loss and damage, and the combination of poor monitoring of slow-onset changes and
38 influence of non-climatic determinants of observed impacts challenges attribution (Thomas and Benjamin,
39 2018). The fact that climate information is often available only in the English language represents another
40 common constraint for island communities (Betzold, 2015). Although indigenous and local knowledge
41 systems can provide important experience-based input to adaptation policies (Miyan et al., 2017), socio-
42 cultural values and traditions such as attachment to place, religious beliefs and traditions can also constrain
43 adaptation in island communities, particularly for more transformational forms of adaptation (Ha'apio et al.,
44 2018; Oakes, 2019).
45
46 Soft limits to adaptation for coastal flooding and erosion are already being experienced in Samoa due largely
47 to financial, physical and technological constraints (Crichton and Esteban, 2018). While sea walls have been
48 erected to minimize coastal erosion, these defences need regular upgrading and replacement as high swells,
49 tropical cyclones and constant wave action erode their effectiveness. The high costs of installing, upgrading
50 and enlarging such infrastructure has led to sea walls only being used in specific locations, leaving
51 communities that are beyond the extent of these measures exposed to inundation and erosion. Native tree
52 replanting has also been implemented but coastal flooding and erosion persist as large swells lead to high
53 failure rates of replanting efforts. Across SIDS, adaptation to coastal flooding and erosion in particular is
54 increasingly facing soft limits due to high costs, unavailability of technological options and limited physical
55 space or environmental suitability for hard engineering or ecosystem-based approaches (Mackey and Ware,
56 2018; Nalau et al., 2018).
57
Do Not Cite, Quote or Distribute 16-46 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 Retreat and relocation constitute transformative adaptation options, although evidence of permanent
2 community-scale relocation in response to climate change remains limited at present (Kelman, 2015;
3 McNamara and Des Combes, 2015). Material and emotional cost of emigration as well as loss of homeland,
4 nationhood, and other intangible assets and values imply that relocation is generally considered a last resort
5 (Jamero et al., 2017) and may mean abandoning objectives of remaining in existing locations, hence
6 exceeding adaptation limits.
7
8 Hard limits in SIDS are mostly due to adaptation being unable to prevent intolerable risks from escalating
9 climate hazards such as sea-level rise and related risks of flooding and surges, severe tropical cyclones, and
10 contamination of groundwater. Emerging evidence suggests that shortage of water and land degradation have
11 already contributed to migration of multiple island communities in the Pacific (Handmer and Nalau, 2019).
12
13 Residual risks for SIDS include loss of marine and terrestrial biodiversity and ecosystem services, increased
14 food and water insecurity, destruction of settlements and infrastructure, loss of cultural resources and
15 heritage, collapse of economies and livelihoods and reduced habitability of islands (Section 3.5.1, Section
16 15.3).
17
18 16.4.2.2 Agriculture in Asia
19
20 Lack of financial resources is found to be a significant constraint that contributes to soft limits to adaptation
21 in agriculture across Asia. Although smallholder farmers are currently adapting to climate impacts, lack of
22 finance and access to credit prevents upscaling of adaptive responses and has led to losses (Bauer, 2013;
23 Patnaik and Narayanan, 2015; Bhatta and Aggarwal, 2016; Loria, 2016). Other constraints further contribute
24 to soft limits including governance and associated institutional factors such as ineffective agricultural
25 policies and organizational capacities (Tun Oo et al., 2017), information and technology challenges such as
26 limited availability and access to technologies on the ground (Singh et al., 2018), socio-cultural factors such
27 as the social acceptability of adaptation measures that are affected by gender (Huyer, 2016; Ravera et al.,
28 2016), and limited human capacity (Masud et al., 2017). A wide range of pests and pathogens are predicted
29 to become problematic to regional food crop production as average global temperatures rise (Deutsch et al.,
30 2018) increasing crop loss across Asia for which farmers are already experiencing a variety of adaptation
31 constraints including financial, economic and technological challenges (Sada et al., 2014; Tun Oo et al.,
32 2017; Fahad and Wang, 2018). Extreme heat waves are projected in the densely populated agricultural
33 regions of South Asia leading to increased risk of heat stress for farmers and resultant constraints on their
34 ability to implement adaptive actions (Im et al., 2017). However, socio-economic constraints appear to have
35 a higher influence on soft limits to adaptation in agriculture than biophysical constraints (Thomas et al.,
36 2021). For example, an examination of farmers' adaptation to climate change in Turkey found that
37 constraints related to access to climate information and access to credit will likely limit the yield benefits of
38 incremental adaptation (Karapinar and Özertan, 2020). In Nepal, conservation policies restrict traditional
39 grazing inside national parks, which promotes intensive agriculture and limits other cropping systems that
40 have been implemented as climate change adaptation (Aryal et al., 2014).
41
42 In Bangladesh, small and landless farm households are already approaching soft limits in adapting to
43 riverbank erosion (Alam et al., 2018). While wealthier farming households can implement a range of
44 adaptation responses including changing planting times and cultivating different crops, poorer households
45 have limited access to financial institutions and credit to implement such measures. Their adaptation
46 responses of shifting to homestead gardening and animal rearing are insufficient to maintain their livelihoods
47 and these households are more likely to engage in off-farm work or migrate.
48
49 (Palao et al., 2019) identify the possible need for transformational adaptation in Asian-Pacific agricultural
50 practices due to changes in biophysical parameters as global average temperatures rise. In this context,
51 transformational adaptation would consist of changing farming locations to different provinces or different
52 elevations for the production of specific crops or introducing new farming systems. Nearly 50% of maize in
53 the region along with 18% of potato and 8% of rice crops would need to either be shifted in location or use
54 new cropping systems, with the most significant transformation being needed in China, India, Myanmar and
55 the Philippines. For maize suitability by 2030, seven provinces in the east and northeast of China are
56 projected to experience over 50% reduction in suitability and two northern states in India may experience
Do Not Cite, Quote or Distribute 16-47 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 70% reduction in suitability. Cassava and sweet potato may play a critical role in food resilience in these
2 areas, as these crops are more resilient to climate change (Prain and Naziri, 2019).
3
4 In terms of hard limits, the rate and extent of climate change is critical as agriculture is climate-dependent
5 and sensitive to changes in climate parameters. Poudel and Duex (2017) document that over 70% of the
6 springs used as water sources in Nepalese mountain agricultural communities had a decreased flow and
7 approximately 12% had dried up over the past decade. While there are some adaptation measures to address
8 reduced water availability e.g., the introduction of water-saving irrigation technology among Beijing
9 farmers to alleviate water scarcity in metropolitan suburbs (Zhang et al., 2019) these actions still depend on
10 some level of water availability. If climate hazards intensify to the point where water supply cannot meet
11 agricultural demands, hard limits to adaptation will occur.
12
13 Residual risks associated with agriculture in Asia include declines in fisheries, aquaculture and crop
14 production, particularly in South and Southeast Asia (Section 10.3.5), increased food insecurity (Section
15 10.4.5), reductions of farmers' incomes by up to 25% (Section 10.4.5), loss of production areas (Section
16 10.4.5) and reduced physical work capacity for farmers - between 5-15% decline in south-southwest Asia
17 and China under RCP8.5 (Section 5.12.4).
18
19 16.4.2.3 Livelihoods in Africa
20
21 For livelihoods dependent on small-scale rain-fed agriculture in Africa, climate hazards include floods and
22 droughts. However, governance, financial and information/awareness/technology challenges are identified as
23 the most significant constraints leading to soft limits, followed by social and human capacity constraints
24 (Thomas et al., 2021). Finance and land tenure constraints restrict Ghanaian farmers when considering
25 adaptation responses due to climate variability (Guodaar et al., 2017). Similarly, in East Africa, farmers with
26 small pieces of land have limited economic profitability, making it difficult to invest in drought and/or flood
27 management measures (Gbegbelegbe et al., 2018).
28
29 Increasing droughts and floods require costlier adaptation responses to reduce risks, such as using drought-
30 tolerant species (Berhanu and Beyene, 2015) and coping strategies for flood-prone households (Schaer,
31 2015; Musyoki et al., 2016), resulting in soft limits for poorer households who cannot afford these responses.
32 In Namibia weak governance and poor integration of information, such as disregarding knowledge of urban
33 and rural residents in flood management strategies, has resulted in soft limits to adaptation, leading to
34 temporary or permanent relocation of communities (Hooli, 2016). Shortage of land namely high population
35 pressure and small per capita land holding leads to continuous cultivation and results in poor soil fertility.
36 This low productivity is further aggravated by erratic rainfall causing soft limits as farmers cannot produce
37 enough and must depend on food aid (Asfaw et al., 2019).
38
39 Relocation due to flooding is discussed as a transformation adaptation action taken in Botswana where the
40 government decided to permanently relocate hundreds of residents to a nearby dryland area (Shinn et al.,
41 2014). Some residents permanently relocated whereas others only temporarily relocated against the
42 government's instructions. Such relocation processes must attend to micro-politics and risks of existing
43 systemic issues of inequality and vulnerability.
44
45 In terms of hard limits, land scarcity poses a hard limit when implementing organic cotton production, an
46 adaptation response supporting sustainable livelihoods (Kloos and Renaud, 2014).
47
48 Residual risks associated with livelihoods in Africa include poorer households becoming trapped in cycles of
49 poverty (Section 9.9.3), increased rates of rural-urban migration (Section 9.8.4), decline of traditional
50 livelihoods such as in agriculture (Section 9.9.3, Section 9.11.3.1) and fisheries (Section 9.11.1.2) and loss of
51 traditional practices and cultural heritage (Section 9.9.2).
52
53 16.4.3 Regional and Sectoral Synthesis of Limits to Adaptation
54
55 16.4.3.1 Evidence on Limits to Adaptation
56
Do Not Cite, Quote or Distribute 16-48 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 There is high agreement and medium evidence that there are limits to adaptation across regions and sectors.
2 However, much of the available evidence focuses on constraints that may lead to limits at some point with
3 little detailed information on how limits may be related to different levels of socio-economic or
4 environmental change (high confidence). Figure 16.7 assesses evidence on constraints and limits for broad
5 categories of region and sector. Small Islands and Central and South America show most evidence of
6 constraints being linked to adaptation limits across sectors while ocean and coastal ecosystems and health,
7 wellbeing and communities show most evidence of constraints being linked to limits across regions (medium
8 confidence).
9
10
11 Figure 16.7 Evidence on constraints and limits to adaptation by region and sector. Data from (Thomas et al. 2021),
12 based on 1682 scientific publications reporting on adaptation-related responses in human systems. See SM16.1 for
13 methods. Low evidence: <20% of assessed literature has information on limits, literature mostly focuses on constraints
14 to adaptation Medium evidence: between 20-40% of assessed literature has information on limits, literature provides
15 some evidence of constraints being linked to limits High evidence: > 40% of assessed literature has information on
16 limits, literature provides broad evidence of constraints being linked to limits
17
18
19 There are clusters of evidence with additional details on limits to adaptation, as detailed in Table 16.3.
20 Evidence on limits to adaptation is largely focused on terrestrial and aquatic species and ecosystems, coastal
21 communities, water security, agricultural production, and human health and heat (high confidence).
22
23 Beginning at 1.5°C, autonomous and evolutionary adaptation responses by terrestrial and aquatic species and
24 ecosystems face hard limits, resulting in biodiversity decline, species extinction and loss of related
25 livelihoods (high confidence). Interventionist adaptation strategies to reduce risks for species and ecosystems
26 face soft limits due to governance, financial and knowledge constraints (medium confidence).
27
28 As sea levels rise and extreme events intensify, coastal communities face soft limits due to financial,
29 institutional and socio-economic constraints reducing the efficacy of coastal protection and accommodation
30 approaches and resulting in loss of life and economic damages (medium confidence). Hard limits for coastal
31 communities reliant on nature based coastal protection will be experienced beginning at 1.5°C (medium
32 confidence).
33
34 Beginning at 3°C, hard limits are projected for water management measures, leading to decreased water
35 quality and availability, negative impacts on health and wellbeing, economic losses in water and energy
36 dependent sectors and potential migration of communities (medium confidence).
37
Do Not Cite, Quote or Distribute 16-49 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 Soft and hard limits for agricultural production are related to water availability and the uptake and
2 effectiveness of climate-resilient crops which is constrained by socio-economic and political challenges
3 (medium confidence).
4
5 Adaptation measures to address risks of heat stress, heat mortality and reduced capacities for outdoor work
6 for humans face soft and hard limits across regions beginning at 1.5°C and are particularly relevant for
7 regions with warm climates (high confidence).
8
9
10 Table 16.3: Adaptation limits and residual risks for select actors and systems. Asterisks indicate confidence
11 level *=low confidence, **=medium confidence, ***=high confidence, ****=very high confidence.
Actor/system at risk Adaptation limits Residual risks
Biodiversity decline, local
extinctions, half of all species
Terrestrial species in Hard: autonomous adaptation unable to overcome currently considered to be at risk
islands at risk to loss of loss of habitat and lack of physical space (***) (Box of extinction occur on islands
habitat CCP1.1) (Box CCP 1.1)
9% of species face complete range
loss (*) mountain-top endemics
and species at poleward
boundaries of African continent at
Terrestrial species across Hard: beyond 2°C many species will lack suitable risk of range loss due to
Africa at risk to habitat climate conditions by 2100 despite migration and disappearing cold climates (***)
changes dispersal (***) (9.6.4.1) (9.6.4.1)
Greater risks of loss of endemic
African aquatic organisms Hard: thermal changes above optimal physiological fish species than generalist fish
at risk to habitat changes limits will reduce available habitats (9.6.2.4) species (9.6.2.4)
African coastal and Over 90% of east African coral
marine ecosystems at risk Hard: at 2°C bleaching of east African coral reefs reefs destroyed at 2C (***)
to habitat changes (***) (9.6.2.3) (9.6.2.3)
Hard: coral restoration and management no longer Loss of more than 80% of healthy
effective after 2°C (***), enhanced coal and reef coral cover, loss of livelihoods
Coral reefs at risk to shading no longer effective after 3°C (**) (Figure dependent on coral reefs (***)
oceanic changes 3.23) (Figure 3.23, Table 8.7)
Cold-adapted species
whose habitats are Species extinctions in the case of
restricted to polar and high Hard: evolutionary responses unable to keep pace species losing its climate space
mountaintop areas at risk with the rate of climate change and degraded state of entirely on a regional or global
to loss of climate space ecosystems (2.6.1, CCP 1.2.4.2) scale (2.6.1, CCP 1.2.4.2)
Ecosystems in North Soft: governance constraints hinder implementation of
America at risk to multiple adaptation strategies Hard: some species unable to
climate hazards adapt (Table 14.8) -
Soft: financial and knowledge constraints lead to
limits for interventionist approaches such as
Ecosystems and species at translocation of species or ecosystem restoration Species extinctions and changes,
risk to multiple climate Hard: some habitats unable to be effectively restored irreversible major biome shifts
hazards (2.6.6) (2.6.6)
With 1-1.1m of sea level rise,
value of coastal urban
infrastructure at risk in Australia is
A$164 to >226 billion while in NZ
it is NZ$43 billion. Sea level rise
will also result in significant
cultural and archaeological sites
disturbed and increasing flood risk
Coastal settlements in and water insecurity with health
Australia and New Soft and hard: limits in the efficacy of coastal and well-being impacts on
Zealand at risk to sea level protection and accommodation approaches as sea Australia's small northern islands
rise levels rise and extreme events intensify (Box 11.5) (Box 11.5)
Do Not Cite, Quote or Distribute 16-50 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
Soft: socio-economic, institutional and financial
constraints may lead to soft limits well in advance of
technical limits of hard engineering measures (CCP
2.3.2, 2.3.4) Hard: Nature based measures (e.g.
restoration of coral reefs, mangroves, marshes) reach at 3°C, globally up to 510 million
Human settlements in hard limits beginning at 1.5°C of global warming. people and up to US$12,739
coastal areas in the 1 in Retreat strategies reach hard limits as availability and billion in assets at risk by 2100
100 year floodplain at risk affordability of land decreases (CCP 2.3.2.3, CCP (CCP 2.2.1)
to coastal flooding 2.3.5)
Hard: domestic freshwater resources unable to Migration of communities due to
recover from increased drought, sea level rise and water shortages with impacts on
Communities in small decreased precipitation by 2030 (RCP8.5+ ice-sheet well-being, community cohesion,
islands at risk to collapse), 2040 (RCP8.5) or 2060 (RCP4.5) (Box 4.2, livelihoods and people-land
freshwater shortages 4.7.2) relationships (Box 4.2)
Communities in North Soft: financial and technological constraints lead to
America at risk to poor limits in ability to treat water for harmful algal
water quality blooms. (Table 14.8)
At 3°C, two thirds of the
population of Southern Europe at
risk to water security with
significant economic losses in
Communities in Western water and energy dependent
and Central Europe at risk Hard: at 3°C, geophysical and technological limits sectors (**) (13.2.2, 13.6,
to water shortages reached in Southern Europe (13.10.3.3) 13.10.2.3)
Increasing competition and
conflict associated with high
Soft: improved water management as an adaptation economic losses (**); glacier
Communities in Central strategy unable to overcome lack of trust and shrinkage leading to loss of related
and South America at risk stakeholder flexibility, unequal power relations and livelihoods and cultural values
to water shortages reduced social learning. (12.5.3.4) (12.5.3.1, Table 8.7)
Agricultural production in Soft: above 3°C, unavailability of water will limit At 3-4°C, yield losses for maize
Europe at risk to heat and irrigation as an adaptation response (***) (13.5.1, may reach up to 50% (**) (13.5.1,
drought 13.10.2.2) 13.10.2.2)
Costs of adaptation and residual
damages are US$63 billion at
Soft: socio-economic and political constraints limit 1.5°C. US$80 billion at 2°C and
uptake of climate-resilient crops (5.4.4.3) Hard: after US$128 billion at 3°C, with
Crops at risk to 2°C, cultivar changes unable to offset global greater risks and damages in
temperature increase production losses (5.4.4.1) tropical and arid regions (5.4.4.1)
At 1.5°C, 30,000 annual deaths
due to extreme heat with up to
90,000 annual deaths at 3C in
Soft: many adaptation measures will not be able to 2100 (***) (13.7.1) At 3°C,
fully mitigate overheating in buildings with high thermal comfort hours during
levels of global warming (***) (13.6.2.3) Hard: summer will decrease by as much
Human health in Europe at above 3°C, people and health systems unable to adapt as 74% in locations in southern
risk to heat (***) (13.6.2.3, 13.7.2, 13.7.4, 13.10.2.1, 13.8) Europe (***) (13.6.1.5)
Globally the impact of projected
climate change on temperature-
related mortality is expected to be
a net increase under RCP4.5 to
RCP8.5, even with adaptation,
Human health at risk to Soft: socio-economic constraints limit adaptation particularly for regions with warm
heat responses to extreme heat (7.4.2.6, Table 8.7) climates (****) (7.3.1, Table 8.7)
At RCP4.5, 25-50% of population
affected; at RCP8.5 more than
South Asian settlements at Soft and hard: At 4.5ºC, maximum temperature is 50% of population affected. At
risk to coastal flooding, expected to exceed survivability threshold across most 4.5ºC of warming, increase in
drought, sea level rise and of South Asia, particularly relevant for outdoor work heat-related deaths of 12.7% in
heatwaves (*) (Table 10.6) South Asia (*) (Table 10.6)
Do Not Cite, Quote or Distribute 16-51 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
Tourism in Europe reliant Soft: at 3°C, snowmaking as an adaptation measure Damages in European tourism
with larger losses in Southern
on snow at risk to higher limited by biophysical and financial constraints (***) Europe (***) (13.6.1.4)
levels of warming (13.6.1.4, 13.6.2.3) -
Rapidly growing
towns/cities and smaller
cities at risk to range of Soft: governance and financial constraints lead to
climate hazards limits in ability to adapt (6.3, 6.4)
1
2
3 16.4.3.2 Constraints Leading to Limits to Adaptation
4
5 Across regions and sectors, a range of constraints (Figure 16.8) are identified as leading to limits to
6 adaptation, particularly financial constraints and constraints related to governance, institutions and policy
7 (high confidence). While individual constraints may appear straightforward to address, the combination of
8 constraints interacting with each other leads to soft limits that are difficult to overcome (high confidence).
9 The interplay of many different constraints that lead to limits makes it difficult to categorize limits beyond
10 being either soft or hard.
11
12
13
14 Figure 16.8 Constraints associated with limits by region and sector. Data from (Thomas et al. 2021), based on 1682
15 scientific publications reporting on adaptation-related responses in human systems. See SM16.1 for methods.
16 Constraints are categorized as: (1) Economic: existing livelihoods, economic structures, and economic mobility; (2)
17 Social/cultural: social norms, identity, place attachment, beliefs, worldviews, values, awareness, education, social
18 justice, and social support; (3) Human capacity: individual, organizational, and societal capabilities to set and achieve
19 adaptation objectives over time including training, education, and skill development; (4) Governance, Institutions &
20 Policy: existing laws, regulations, procedural requirements, governance scope, effectiveness, institutional arrangements,
Do Not Cite, Quote or Distribute 16-52 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 adaptive capacity, and absorption capacity; (5) Financial: lack of financial resources; (6)
2 Information/Awareness/Technology: lack of awareness or access to information or technology; (7) Physical: presence
3 of physical barriers; and (8) Biologic/climatic: temperature, precipitation, salinity, acidity, and intensity and frequency
4 of extreme events including storms, drought, and wind. Insufficient data: there is not enough literature to support an
5 assessment (less than 5 studies available); Minor constraint: <20% of assessed literature identifies this constraint;
6 Secondary constraint: 20-50% of assessed literature identifies this constraint; Primary constraint: >50% of assessed
7 literature identifies this constraint
8
9
10 Table 16.4: Key constraints associated with limits to adaptation for regions
Region Key constraints associated with limits to adaptation
Africa Financial constraints inhibit implementation of a variety of adaptation strategies including
ecosystem-based adaptation (Section 9.11.4.2) and adoption of drought tolerant crops by farmers
(Section 9.12.3).
Information constraints (including limited climate science information), governance constraints
(such as communication disconnects between national, district and community levels) and human
capacity constraints (limited capacities to analyse threats and impacts) are identified as negatively
affecting the implementation of adaptation policies (Section 9.13.1).
Social/cultural constraints (social status, caste and gender) also affect adaptation in contexts with
deep-rooted traditions (Section 9.12.4).
Asia Governance, human capacity, financial and informational constraints commonly present barriers to
urban adaptation (Section 10.4.6.5).
Economic, governance, financial and informational constraints are related to both soft and hard
limits to adaptation against a range of hazards in South Asia (Box 10.7), while in West Asia,
physical constraints to heatwaves and drought have been associated with limits to adaptation (Box
10.7).
Australasia A range of constraints, including governance, information and awareness, social/cultural, human
capacity and financial have been identified as impeding adaptation action in the region (Section
11.7.2, Box 11.1).
Evidence of limits to adaptation are primarily for ecosystems (Section 11.7.2, 11.6) although
individuals and communities are also approaching soft limits due to social constraints (Chapter
11.7.2).
Central and Financial, governance, knowledge, biophysical and social/cultural constraints identified as most
South America significant for adaptation (Section 12.5, Table 12.3).
Soft limits are largely related to governance constraints, while evidence of hard limits is related to
biophysical constraints, such as glacier shrinking leading to loss of livelihoods and cultural values
(Section 12.5.3.4).
Europe Key constraints are identified as technical, biophysical, economic and social (Section 13.6.2.4).
For cities, settlements and key infrastructure, technical socio-economic and environmental &
regulatory constraints may lead to limits at a range of spatial scales (Figure 13.12)
Biophysical constraints may lead to limits to the ability of water saving and water efficiency
measures to prevent water insecurity under high warming scenarios (Section 13.2.2.2).
North America Social/cultural, governance, financial, knowledge and biophysical constraints are identified as
most significant for adaptation and leading to both soft and hard limits (Section 14.5.2.1, Section
14.6, Section 14.6.2.1, Table 14.8)
Small Islands Financial, governance, information/awareness, technological, cultural and human capacity
constraints are identified as affecting adaptation and leading to soft limits (Section 15.5.3, Section
15.5.4, Section 15.6.1, Section 15.6.3, Section 15.6.4).
Differences between constraints and soft limits in the small island context is marginal, with
policymakers in the Caribbean and Indian Oceans seeing these as synonymous (Section 15.6.1).
11
12
13 16.4.3.3 Climate Change Impacts, Financial Constraints and Limits to Adaptation
14
Do Not Cite, Quote or Distribute 16-53 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 Across regions and sectors, financial constraints are identified as significant and contributing to limits to
2 adaptation, particularly in low-to-middle income countries (high confidence) (Section 3.6.3, Section 4.7.2,
3 Section 5.14.3, Section 6.4.5, Section 7.4.2, Section 8.4.5, Section 12.5.1, Section 12.5.2, Section 15.6.1,
4 Section 15.6.3, Figure 16.8, Table 16.4, CCP2.4.2). Impacts of climate change may increase financial
5 constraints (high confidence) and contribute to soft limits to adaptation being reached (medium confidence).
6 Table 16.5 details climate impact observations that point to potentially substantial negative impacts on the
7 availability of financial resources for different regions.
8
9
10 Table 16.5: Evidence of climate change impacts affecting availability of financial resources. Asterisks indicate
11 confidence level *=low confidence, **=medium confidence, ***=high confidence.
Region Evidence of climate change impacts affecting availability of financial resources
Africa Negative consequences for economic growth and GDP growth rate from higher average
temperatures and lower rainfall (***) (Section 9.9.1.1, Section 9.9.2, Section 9.9.3)
Economic losses from damage to infrastructure in the energy, transport, water supply,
communication services, housing, health, and education sectors (observed) (Section 9.7.2.2,
Section 9.8.2)
Asia High coastal damages due to sea level rise (China, India, Korea, Japan, Russia) (***) (Section
10.4.6.3.4)
Decline in aquaculture production (Section 10.4.5.2.1)
Loss of coastal ecosystem services (Bangladesh) (Section 5.9.3.2.4)
Australasia Loss of wealth and negative impacts on GDP (Section 11.5.1.2, Section 11.5.2.2)
High disaster costs (observed in Australia, NZ) (Section 11.5.2.1)
Central and High costs of extreme events relative to GDP (observed in Guatemala, Belize) (Section 12.3.1.4)
South America Decrease in growth of total GDP per capita and total income and labour income from one standard
deviation in the intensity of a hurricane windstorm (Section 12.3.1.4)
Europe Negative combined effect of multiple risks on economy for Europe in total (**) (Section 13.9.1,
Section 13.10.2)
Negative combined effect of multiple risks on economy for Southern Europe (***) (Section 13.9.1,
Section 13.10.2)
High economic costs in agriculture and construction following heat waves and flooding (Section
6.2.3.2, Section 7.4.2.2.1)
North America Small but persistent negative economy wide effect on GDP (observed in the United States and
Mexico) (**) (Box 14.5)
Economic risks associated with high temperature scenarios (***) (Box 14.5)
Small but persistent positive economy wide effect on GDP (observed in Canada) (**) (Box 14.5)
Significant economic costs for urban, natural and ecosystem infrastructure (USA) (Section 6.2.5.9)
High economic damages for a subset of sectors from high warming (southern and southeastern US)
(Box 14.5)
Adverse effects on municipal budgets due to costly liabilities, and disruption of financial markets
(Box 14.5)
Small Islands High economic costs relative to GDP from extreme events, particularly tropical cyclones
(observed) (Section 15.3.4.1)
Negative long-term implications of extreme events for state budgets (Section 8.2.1.4)
Inundation of almost all port and harbour facilities (Caribbean) (Section 15.3.4.1)
12
13
14 At the national level, negative macroeconomic responses to climate change may limit the availability of
15 financial resources, impede access to financial markets and stunt economic growth (high confidence).
16 Economic growth has been shown to decline under higher temperatures (Burke et al., 2015; Kahn et al.,
17 2019, Section 16.5.2.3.4) and following extreme events (Hsiang and Jina, 2014; IMF, 2017), particularly for
18 medium- and low-income developing countries (Section 18.1). The most severe impacts of climate-related
19 disasters on economic growth per capita have been observed in developing countries, although authors note a
Do Not Cite, Quote or Distribute 16-54 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 publication bias in the reporting of negative effects (Klomp and Valckx, 2014). Substantial immediate output
2 losses and reduced economic growth due to extreme events have been observed both in the short- and long-
3 term (Section 16.2.3). Estimates of the duration of negative effects of climate-related disasters differ, with
4 some analyses suggesting that on average economies recover after two years (Klomp, 2016) and others
5 finding negative effects of cyclones to persist 15 20 years following an event (Hsiang and Jina, 2014; IMF,
6 2017). Rising climate vulnerability has also been shown to increase the cost of debt (Kling et al., 2018).
7 Rising climatic risks negatively affect developing countries' ability to access financial markets (Cevik and
8 Jalles, 2020) and their disclosure may result in capital flight (Cross-Chapter Box FINANCE in Chapter 17).
9 Overall, the direct and indirect economic effects of climate change represent a major risk to financial system
10 stability (Section 11.5.2). These risks and effects may further limit the availability of financial resources
11 needed to overcome constraints, in particular for developing countries.
12
13 Sectoral studies indicate that climate impacts will result in higher levels of losses and damages and decreases
14 in income, thereby increasing financial constraints (medium confidence). Yield losses for major agricultural
15 crops are expected in nearly all world regions (Figure 5.7). Decreases in estimated marine fish catch
16 potential and large economic impacts from ocean acidification are expected globally, leading to the risk of
17 revenue loss (Section 5.8.3). Losses of primary productivity and farmed species of shellfish are expected in
18 tropical and subtropical regions (Section 5.9.3.2.2). Economic losses have been observed in the power
19 generation sector and transport infrastructure (Section 10.4.6.3.8), including economic losses from floods in
20 urban areas (Section 4.2.4.5). However, some positive sectoral climate change impacts have been identified
21 for the timber and forestry sector (Section 5.6.2), for primary productivity and farmed species of shellfish in
22 high-latitude regions (Section 5.9.3.2.2) and agriculture in high-latitude regions (Section 5.4.1.1).
23
24 At the household or community level, climate impacts may increase financial constraints (high confidence).
25 Impacts on agriculture and food prices could force between 3 to 16 million people into extreme poverty
26 (Hallegatte and Rozenberg, 2017). Within-country inequality is expected to increase following extreme
27 weather events (Section 16.2.3.6 and Chapter 8). Households affected by climate-related extreme events may
28 be faced with continuous reconstruction efforts following extreme events (Adelekan and Fregene, 2015) or
29 declines in critical livelihood resources in the agriculture, fisheries and tourism sectors (Forster et al., 2014,
30 Section 3.5.1). Further erosion of livelihood security of vulnerable households creates the risk of poverty
31 traps, particularly for rural and urban landless (Section 8.2.1, Section 8.3.3.1), for example in Malawi and
32 Ethiopia (Section 9.9.3). Levels of labour productivity and economic outputs are projected to decrease as
33 temperatures rise particularly in urban areas (Section 6.2.3.1). At the same time, higher utilities demand
34 under higher urban temperatures exert additional economic stresses on urban residents and households.
35 Substantial, negative impacts on the livelihoods of over 180 million people are expected from changes to
36 African grassland productivity (Section 5.5.3.1). In Western Uzbekistan, farmers´ incomes are at risk of
37 declining (Section 10.4.5.3). For Small Island Developing States, loss of livelihoods is expected due to
38 negative climatic impacts on coastal environments and resources (Section 3.5.1). Negative effects on
39 households from extreme events can also persist in the long-term and in multiple dimensions. Exposure to
40 disasters during the first year of life significantly reduces the number of years of schooling, increases the
41 chances of being unemployed as an adult and living in a multidimensionally poor household (González et al.,
42 2021).
43
44
45 16.5 Key Risks Across Sectors and Regions
46
47 This section builds on the analogous chapter in AR5 (Oppenheimer et al., 2014) to refine the definition of
48 climate-related key risks (KRs) and criteria for identifying them (16.5.1), and describe a broad range of key
49 risks by sector and region as identified by the authors of WGII AR6 (Section 16.5.2, SM16.4). Based on this,
50 eight clusters of key risks (i.e., Representative Key Risks, RKRs) are identified and assessed in terms of the
51 conditions under which they would become severe. In addition, the section assesses variation in KRs and
52 RKRs by the level of global average warming, socio-economic development pathways, and levels of
53 adaptation, and illustrates the implications from resulting dynamics in all risk dimensions (hazard, exposure,
54 vulnerability) along a case study of densely populated river deltas (Section 16.5.3). Last, interactions among
55 RKRs are discussed (Section 16.5.4).
56
Do Not Cite, Quote or Distribute 16-55 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 16.5.1 Defining Key Risks
2
3 A key risk is defined as a potentially severe risk and therefore especially relevant to the interpretation of
4 dangerous anthropogenic interference (DAI) with the climate system, the prevention of which is the ultimate
5 objective of the UNFCCC as stated in its Article 2 (Oppenheimer et al., 2014). Key risks are therefore a
6 relevant lens for the interpretation of this policy framing. The severity of a risk is a context-specific
7 judgment based on a number of criteria discussed below. KRs are `potentially' severe because, while some
8 could already reflect dangerous interference now, more typically they may become severe over time due to
9 changes in the nature of hazards (or, more broadly, climatic impact-drivers (or, more broadly, climatic
10 impact-drivers, IPCC, 2021a) and/or of the exposure/vulnerability of societies or ecosystems to those
11 hazards. They also may become severe due to the adverse consequences of adaptation or mitigation
12 responses to the risk (on the former see Section 17.5.1; the latter is not assessed separately here, except as it
13 contributes to risks from climate hazards). Dangerous interferences in this chapter are considered over the
14 course of the 21st century.
15
16 KRs may be defined for a wide variety of systems at a range of scales. The broadest definition is for the
17 global human system or planetary ecological system, but KRs may also apply to regions, specific sectors or
18 communities, or to parts of a system rather than to the system as a whole. For example, the population at the
19 lower end of the wealth distribution is often impacted by climate change much more severely than the rest of
20 the population (Leichenko and Silva, 2014; Hallegatte and Rozenberg, 2017; Hallegatte et al., 2017; Pelling
21 and Garschagen, 2019).
22
23 KRs are determined not just by the nature of hazards, exposure, vulnerability, and response options, but also
24 by values, which determine the importance of a risk. Importance is understood here as the degree of
25 relevance to interpreting DAI at a given system's level or scale, and was an explicit criterion for identifying
26 key vulnerabilities and risks in AR5 (Oppenheimer et al., 2014). Because values can vary across individuals,
27 communities, or cultures, as well as over time, what constitutes a KR can vary widely from the perspective
28 of each of these groups, or across individuals. For example, ecosystems providing indirect services and
29 cultural assets such as historic buildings and archaeological sites may be considered very important to
30 preserve by some people but not by others; and some types of infrastructure, such as a commuter rail, may be
31 important to the well-being of some households but less so to others. Therefore, Chapter 16 authors do not
32 make their own judgements about the importance of particular risks. Instead, we highlight importance as an
33 overarching factor but identify and evaluate KRs based on four other criteria for what may be considered
34 potentially severe.
35
36 Magnitude of adverse consequences. Magnitude measures the degree to which particular dimensions of a
37 system are affected, should the risk materialize. Magnitude can include the size or extent of the system, the
38 pervasiveness of the consequences across the system (geographically or in terms of affected population), as
39 well as the degree of consequences. Consequences can be measured by a wide range of characteristics. For
40 example, risks to food security can be measured as uncertain consequences for food consumption, access, or
41 prices. The magnitude of these consequences would be the degree of change in these measures induced by
42 climate change and accounting for the interaction with exposure and vulnerability. In addition to
43 pervasiveness and degree of change, several other aspects can contribute to a judgement of magnitude,
44 although they refer to concepts that are difficult to capture and highly context-specific:
45 Irreversibility of consequences. Consequences that are irreversible, at least over long timescales, would be
46 considered a higher risk than those that are temporary. For example, changes to the prevailing ecosystem in a
47 given location may not be reversible on the decade to century scale.
48 Potential for impact thresholds or tipping points. Higher risks are posed by the potential for exceeding a
49 threshold beyond which the magnitude or rate of an impact substantially increases.
50 Potential for cascading effects beyond system boundaries. Higher risks are posed by those with the potential
51 to generate downstream cascading effects to other ecosystems, sectors or population groups within the
52 affected system and/or to another system, whether neighbouring or distant (Cross-Chapter Box INTEREG in
53 this Chapter).
54
55 Likelihood of adverse consequences. A higher probability of high-magnitude consequences poses a larger
56 risk a priori, whatever the scale considered. This probability may not be quantifiable, and it may be
57 conditional on assumptions about the hazard, exposure, or vulnerability associated with the risk.
Do Not Cite, Quote or Distribute 16-56 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1
2 Temporal characteristics of the risk. Risks that occur sooner, or that increase more rapidly over time,
3 present greater challenges to natural and societal adaptation. A persistent risk (due to the persistence of the
4 hazard, exposure, and vulnerability) may also pose a higher threat than a temporary risk due, for example, to
5 a short-term increase in the vulnerability of a population (e.g., due to conflict or an economic downturn).
6
7 Ability to respond to the risk. Risks are more severe if the affected ecosystems or societies have limited
8 ability to reduce hazards (e.g., for human systems, through mitigation, ecosystem management and possibly
9 solar radiation management); to reduce exposure or vulnerability through various human or ecological
10 adaptation options; or to cope with or respond to the consequences, should they occur.
11
12 The relative influence of these different criteria is case-specific and left to author judgment in the
13 identification of KRs (groups of authors in regional and sectoral chapters, see Supplementary Material Table
14 SM16.10) and the assessment of representative key risks (author teams, see Supplementary Material Table
15 SM16.10). But in general, the more criteria are met, the higher is the risk
16
17 16.5.2 Identification and Assessment of Key Risks and Representative Key Risks
18
19 16.5.2.1 Identification of Key Risks (KR)
20
21 The authors of the sectoral and regional chapters and Cross Chapter Papers of the WGII AR6 Report
22 identified more than 130 key risks (Table SM16..4). Authors were asked to rely on the above definition and
23 criteria to identify risks that could potentially become severe according to changes in the associated hazards,
24 the study systems' exposure and/or vulnerability; and important adaptation strategies that could reduce these
25 risks (see 16.B.2 for methodology). Wherever possible, identification is based on literature that includes
26 projected future conditions for all three components of risk and adaptation. Where literature was insufficient,
27 potential severity is based on current vulnerability and exposure to climate hazards and the expectation that
28 hazards will increase in frequency and/or intensity in the future. This approach is more limited in that it does
29 not consider future changes in exposure and vulnerability nor in adaptation, but has the benefit of being
30 grounded in observed experience.
31
32 Table SM16.4 indicates that climate change presents a wide range of risks across scales, sectors and regions
33 that could become severe under particular conditions of hazards, exposure, and vulnerability, which may or
34 may not occur. Some illustrations of the extent and diversity of KRs are provided here, and more detailed
35 assessment can be found in the Chapters referenced in the table.
36
37 Global scale KRs include threats to biodiversity in oceans, coastal regions, and on land, particularly in
38 biodiversity hotspots, as well as other ecological risks such as geographic shifts in vegetation, tree mortality,
39 reduction in populations, and reduction in growth (such as for shellfish). These ecological risks include
40 cascading impacts on livelihoods and food security. Global-scale risks also include risks to people, property,
41 and infrastructure from river flooding and extreme heat (particularly in urban areas), risks to fisheries (with
42 implications for living standards and food security), and some health risks from food-borne diseases as well
43 as psychopathologies.
44
45 Many KRs are especially prominent in particular regions or systems, or for particular subgroups of the
46 population. For example, coastal systems and small islands are a nexus of many KRs, including those to
47 ecosystems and their services, especially coral reefs; people (health, livelihoods); and assets, including
48 infrastructure. Risks to socio-ecological systems in polar regions are also identified as KRs, as are ecological
49 risks to the Amazon forest in South America and savannahs in Africa. For some regions risks from wildfire
50 are of particular concern, including in Australasia and North America. Vector-borne diseases are a particular
51 concern in Africa and Asia. Loss of cultural heritage is identified as a KR in Small Islands, Mountain
52 Regions, Africa, Australasia, and North America.
53
54 For many risks, low-income populations are particularly vulnerable to KRs. Climate-related impacts on
55 malnutrition and other forms of food insecurity will be larger for this group, along with small-holder farming
56 households and indigenous communities reliant on agriculture, and for women, children, the elderly, and the
57 socially isolated (Section 5.12). KRs in coastal communities are expected to affect low income populations
Do Not Cite, Quote or Distribute 16-57 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 more strongly, including through risks to livelihoods of those reliant on coastal fisheries. KRs related to
2 health are generally higher for low income populations less likely to have adequate housing or access to
3 infrastructure.
4
5 16.5.2.2 Identification of Representative Key Risks (RKR)
6
7 As in AR5 Oppenheimer et al. (2014), major clusters of KRs are further analysed, and here referred to as
8 `representative key risks' (RKRs). RKRs were defined in a three-step process (SM16.2.1). First, half of
9 Chapter 16 authors independently mapped the KRs in Table SM16.4 to a set of candidate RKRs. Second, all
10 Chapter 16 authors discussed the set of independent results and proposed a list of RKRs, considering scope
11 and overlap. Third, this proposal was discussed with a consultative group of about twenty WGII AR6 authors
12 from other chapters closely involved in the KR identification process, and a final list of 8 RKRs was
13 identified (Table 16.6).
14
15 The RKRs are intended to capture the widest variety of KRs to human or ecological systems with a small
16 number of categories that are easier to communicate and provide a manageable structure for further
17 assessment. They expand the scope of some AR5 KR clusters (e.g., on coasts, health, food, and water) and
18 add new ones (e.g., on peace and mobility). The RKRs encompass a diversity of types of systems, including
19 an example of a geographically defined system (RKR-A on coastal regions), ecosystem well-being and
20 integrity (RKR-B), a cross-cutting issue relevant to several outcomes of concern (RKR-C on critical
21 infrastructure), and several topics focused directly on aspects of human well-being and security (RKR-D to
22 RKR-H). This set of RKRs manages but does not eliminate overlap, instead providing alternative
23 perspectives on underlying key risks that sometimes include complementary views on common risks. For
24 example, the water security RKR highlights the many key risks mediated by water quantity or quality, which
25 are sometimes manifested as risk to food security (RKR-F) or health (RKR-E).
26
27
28 Table 16.6: Climate-related representative key risks (RKRs). The scope of each RKR is further described in the
29 assessments in Section 16.5.2.3. Relation to categories of overarching key risks identified in AR5 is provided for
30 continuity.
Code Representative Scope Relation to Sub-section
Key Risk AR5 assessment
overarching
key risks
for definitions,
refer to
(Oppenheimer
et al., 2014)
RKR-A Risk to low- Risks to ecosystem services, people, livelihoods and Contains key 16.5.2.3.1
lying coastal key infrastructure in low-lying coastal areas, and risk (i),
socio-ecological associated with a wide range of hazards, including overlaps with
systems sea level changes, ocean warming and acidification, key risks (iii)
weather extremes (storms, cyclones), sea ice loss, and (vii)
etc.
RKR-B Risk to Transformation of terrestrial and ocean/coastal Contained in 16.5.2.3.2
terrestrial and ecosystems, including change in structure and/or key risks (vii)
ocean functioning, and/or loss of biodiversity. and (viii)
ecosystems
RKR-C Risks associated Systemic risks due to extreme events leading to the Overlaps with 16.5.2.3.3
with critical breakdown of physical infrastructure and networks key risk (iii)
physical providing critical goods and services.
infrastructure,
networks and
services
Do Not Cite, Quote or Distribute 16-58 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
RKR-D Risk to living Economic impacts across scales, including impacts Broader version 16.5.2.3.4
standards on Gross Domestic Product (GDP), poverty, and of key risk (ii)
livelihoods, as well as the exacerbating effects of
impacts on socio-economic inequality between and
within countries.
RKR-E Risk to human Human mortality and morbidity, including heat- Broader version 16.5.2.3.5
health related impacts and vector-borne and water-borne of key risk (iv)
diseases.
RKR-F Risk to food Food insecurity and the breakdown of food systems Overlaps with 16.5.2.3.6
security due to climate change effects on land or ocean key risk (v)
resources.
RKR-G Risk to water Risk from water related hazards (floods and Overlaps with 16.5.2.3.7
security droughts) and water quality deterioration. Focus on key risk (iv)
water scarcity, water-related disasters and risk to
indigenous and traditional cultures and ways of life
RKR-H Risks to peace Risks to peace within and among societies from New 16.5.2.3.8
and to human
mobility armed conflict as well as risks to low-agency human
mobility within and across state borders, including
the potential for involuntarily immobile populations.
1
2
3 16.5.2.3 Assessment of Representative Key Risks
4
5 Each RKR was assessed by a team of 4 to 9 members drawn from Chapter 16, other WGII AR6 chapters,
6 and external contributing authors (16.B.3.1). The following subsections describe the scope of the category of
7 risk (underlying KR considered) and the approach to defining `severe' risks for each particular RKR. They
8 also assess the conditions in terms of warming (more broadly, climatic impact-drivers; (Ranasinghe et al.,
9 2021), exposure/vulnerability and adaptation under which the RKR would become severe. For each of these
10 dimensions, RKR teams considered generic levels ranging from High to Medium and Low. For warming
11 levels, in line with WG1 framing, High refers to climate outcomes consistent with RCP8.5 or higher, Low
12 refers to climate outcomes consistent with RCP2.6 or lower, and Medium refers to intermediary climate
13 scenarios. For reference, the full range of warming levels (across all climate models) associated with RCP8.5
14 for the 2081-2100 period is 3.0C to 6.2C; for RCP2.6 it is 0.9C to 2.3C; and for intermediate RCPs it is 1.8C
15 to 3.6C (Cross-Chapter Box CLIMATE in Chapter 1). For Exposure-Vulnerability, levels are determined by
16 the RKR teams relative to the range of future conditions considered in the literature, for example based on
17 the Shared Socioeconomic Pathways (SSPs) in which future conditions based on SSPs 1 or 5 represent Low
18 exposure or vulnerability and those based on SSPs 3 or 4 represent High exposure or vulnerability (O'Neill
19 et al., 2014; van Vuuren and Carter, 2014). For Adaptation, two main levels have been considered: High
20 refers to near maximum potential and Low refers to the continuation of today's trends. Despite being
21 intertwined in reality, Exposure-Vulnerability and Adaptation conditions are distinguished to help
22 understand their respective contributions to risk severity. Importantly, this assessment does not consider all
23 risks, but only those that can be considered severe given the definition and criteria presented in Section
24 16.5.1. The assessment does not exclude the possibility that severe risks are already observed in some
25 contexts, and considers projected risks through the end of this century.
26
27 Each RKR assessment followed a common set of guidelines (16.B.3) that included broad criteria for defining
28 severity (Section 16.5.1), consideration of complex risks and interactions within and across RKRs, and
29 consideration of risks across a range of scales, regions, and ecological and human development contexts. The
30 specific definition of severity within each RKR was determined by the author teams of that assessment,
31 applying different combinations of key risk criteria and metrics as judged appropriate in each case.
32 Definitions are transparent and use common criteria, but are nonetheless based on the respective author
33 team's judgment. Conclusions about severity and associated confidence statements are therefore conditional
34 on those definitions.
35
Do Not Cite, Quote or Distribute 16-59 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 Assessments are based on different types of evidence depending on the nature of the literature. In some
2 cases, quantitative projections of potential impacts are available. In others and as for KR identification, the
3 potential for severe risk is inferred from high levels of current vulnerability and the expectation that the
4 relevant climate hazards (CIDs) will increase in frequency or intensity in the future.
5
6 16.5.2.3.1 Risk to the integrity of low-lying coastal socio-ecological systems (RKR-A)
7 RKR-A considers climate change-related risks to low-lying coasts including their physical, ecological and
8 human components. Low-lying systems are those occupying land below 10 m of elevation that is contiguous
9 and hydrologically connected to the sea (McGranahan et al., 2007). The assessment builds on Key Risks
10 identified in chapters 3 and 15, Cross Chapter Paper 2 as well as in the SROCC (Magnan et al., 2019;
11 Oppenheimer et al., 2019). It highlights risks to (i) natural coastal protection and habitats; (ii) lives,
12 livelihoods, culture and well-being; and (iii) critical physical infrastructure; it therefore overlaps with several
13 other RKRs (Fig. 16.10 and 16.11) but within a coastal focus. It encompasses all latitudes and considers
14 multiple sources of climate hazards, including sea-level rise (SLR), ocean warming and acidification,
15 permafrost thaw, and sea-ice loss and changes in weather extremes.
16
17 Severe risks to low-lying coasts involve irreversible long-term loss of land, critical ecosystem services,
18 livelihoods, well-being or culture in relation to increasing combined drivers, including climate hazards and
19 exposure and vulnerability conditions. The definition depends on the local context because of variation in the
20 perception of tolerable risks and the limits to adaptation (Handmer and Nalau, 2019). Accordingly, a
21 qualitative range of consequences is presented here, in place of a quantitative global severe risk threshold.
22
23 The literature suggests that severe risks generally occur at the nexus of high levels and rates of
24 anthropogenic-driven change in climate hazards (16.2.3.2), concentrations of people and tangible and
25 intangible assets, non-climate hazards such as sediment mining and ecosystem degradation (3.4.2.1), and the
26 reaching of adaptation limits (16.4) (medium evidence, high agreement). In some Arctic communities and in
27 communities reliant on warm-water coral reefs, even 1.5-2ºC warming will lead to severe risks from loss of
28 ecosystem services (3.4.2.2; CCP6) (high confidence). Loss of land is already underway globally due to
29 accelerating coastal erosion and will be amplified by increased sea-level extremes and permanent flooding
30 (high confidence; Oppenheimer et al. 2019, Ranasinghe et al. 2021). Observed impacts of and projected
31 increases in high intensity extreme events (Ranasinghe et al. 2021) also provide evidence for severe risk to
32 occur on livelihoods, infrastructure and well-being (Section 16.5.2.3.3) by mid-century (high confidence).
33 Consequently, the combination of high warming, continued coastal development and low adaptation levels
34 will challenge the habitability of many low-lying coastal communities in both developing and developed
35 countries over the course of this century (low evidence, high agreement) (Duvat et al., 2021; Horton et al.,
36 2021). In some contexts, climate risks are already considered severe (medium evidence, medium agreement),
37 and in others, even lower warming will induce severe risks to habitability, which will not necessarily be
38 offset by ambitious adaptation (low evidence, medium agreement).
39
40 (i) Natural coastal protection and habitats -- Severe risks from the loss of shoreline protection from
41 reductions in wave attenuation (Beck et al., 2018, Section 3.5.5.1; Section 3.5.4.5) and sediment delivery
42 (3.4.2.5; 15.3.3) are already observed in some coastal systems (Section 16.2.3.1) and occur broadly even
43 with 1.5ºC of global warming (Hoegh-Guldberg et al., 2018a; Bindoff et al., 2019, Section 3.4.2). These
44 impacts are the consequence of warming and SLR on coastal ecosystems.
45
46 Warm-water coral reefs are at risk of widespread loss of structural complexity and reef accretion by 2050
47 under 1.5°C global warming (Section 3.4.2.1) (high confidence). Kelp forests may experience shifts in
48 community structure (Arafeh-Dalmau et al., 2019; Rogers-Bennett and Catton, 2019; Smale, 2020; Smith et
49 al., 2021) with >2°C of global warming especially at lower latitudes (Section 3.4.2.2) (high confidence). In
50 addition, depending on the local tide and sediment conditions, SLR associated with >1.5°C of global
51 warming (SSP1-2.6; 3.4.2.5) is sufficient to initiate shifts to alternate states in some seagrass and coastal
52 wetland systems (van Belzen et al., 2017; El-Hacen et al., 2018, Section 3.4.2.5, Cross-Chapter Box SLR in
53 Chapter 3), and submergence of some mangrove forests (3.4.2.5). A striking example of risks becoming
54 severe at higher levels of warming is the one of coral islands with low elevation (Section 15.3.4, Box 15.1):
55 the risk of loss of habitability transitions from Moderate-to-High under RCP2.6 for most island types (urban
56 and rural) to High-to-Very High under RCP8.5 (Duvat et al., 2021), even under a high adaptation scenario
Do Not Cite, Quote or Distribute 16-60 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 (Oppenheimer et al., 2019), partly due to declining sediment supply (Perry et al., 2018) and increased annual
2 flooding (Giardino et al., 2018; Storlazzi et al., 2018).
3
4 More broadly, about 28,000 km2 of land have been lost globally since the 1980s due to anthropogenic
5 factors (e.g., coastal structures, disruption of sediment fluxes) and coastal hazards (Mentaschi et al., 2018),
6 and an additional loss of 6000-17,000 km2 is estimated by the end of the century due to coastal erosion alone
7 associated with SLR in combination with other drivers (Hinkel et al., 2013).
8
9 (ii) Impacts to lives, livelihoods, culture and well-being -- In the absence of effective adaptation, changing
10 extreme and slow-onset hazards combined with anthropogenic drivers (e.g., increased population pressure at
11 the coast between +5% and +13.6% by 2100 compared to today, Jones and O'Neill, 2016) will lead to loss of
12 lives, livelihoods, health, well-being, and/or culture (McGregor et al., 2016; Pinnegar et al., 2019; Pugatch,
13 2019; Schneider and Asch, 2020; Thomas and Benjamin, 2020; McNamara et al., 2021) (high confidence).
14 Catastrophic examples that may foreshadow the future include Hurricane Sandy in 2012 (Strauss et al.,
15 2021)and super Typhoon Haiyan in 2013 (>6,000 deaths and inequities in access to safe housing; Trenberth
16 et al. 2015) (6.2.2, 6.3.5.1). Although there is no unique definition of `intolerable' loss, risks are generally
17 expected to become severe over this century (Tschakert et al., 2017; Dannenberg et al., 2019; Tschakert et
18 al., 2019). Globally, with High warming, 90 to 380 million more people will be exposed to annual flood
19 levels by the mid- and end-century, respectively, compared to 250 million people today (Kulp and Strauss,
20 2019; Kirezci et al., 2020), with potential implications on forced displacement or migration (Oppenheimer et
21 al., 2019; Wrathall et al., 2019; Hauer et al., 2020; Lincke and Hinkel, 2021, Section 16.5.2.3.9). Some of the
22 largest fish-producing and fish-dependent ecoregions have already experienced losses of up to 35% in
23 marine fisheries productivity due to warming (Free et al., 2019), and about 11% of the global population will
24 face increasing nutritional risks if current trajectories continue (Golden et al., 2016). While difficult to
25 measure, current climate-driven losses to (indigenous) knowledge, traditions (Tschakert et al., 2019; Pearson
26 et al., 2021) and well-being (Ebi et al., 2017; Cunsolo and Ellis, 2018; Jaakkola et al., 2018) indicate such
27 risk as already severe in some regions (low evidence, medium agreement), jeopardizing communities'
28 realization of their rights to food, health and culture. In the Arctic, climate-driven changes to ice and weather
29 regimes have substantially affected traditional coastal-based hunting and fishing activities (Fawcett et al.,
30 2018; Galappaththi et al., 2019; Huntington et al., 2020; Nuttall, 2020, CCP6), and where permafrost thaw,
31 SLR and coastal erosion are contributing to threatening cultural sites (Hollesen et al., 2018; Fenger-Nielsen
32 et al., 2020).
33
34 (iii) Critical physical infrastructure -- Severe risks are also illustrated through damages that lead to possibly
35 long-lasting disruption of key services like transportation as well as energy generation and distribution in
36 coastal areas (Section 16.5.2.3.3) under all RCPs (CCP2.2.3) and if no additional adaptation (medium
37 confidence). Critical transport infrastructure is already suffering from structural failures in polar regions, for
38 instance, due to permafrost thaw and increased erosion associated with ocean warming, storm surge flooding
39 and loss of sea ice (Melvin et al., 2017; Fang et al., 2018, Section 14.5.2.8, Section 16.2.3.2, CCP6). One
40 hundred airports are projected to be below mean sea-level in 2100 with 2°C of warming (i.e., 0.62 m SLR,
41 Yesudian and Dawson, 2021), including in small islands (Monioudi et al., 2018; Storlazzi et al., 2018) and
42 megacities. Projections show San Francisco International Airport, for instance, to be inundated by 2100
43 under the upper likely range of SLR in RCP8.5 (also considering subsidence trends, Shirzaei and Bürgmann,
44 2018). On the energy side, it is estimated that with 1.8m SLR, for example, four out of 13 US nuclear power
45 plant facilities will become exposed to storm surges and three others will be surrounded or submerged by
46 seawater (Jordaan et al., 2019; Jenkins et al., 2020).
47
48 16.5.2.3.2 Risk to terrestrial and ocean ecosystems (RKR-B)
49 This risk refers to transformations of terrestrial and ocean/coastal ecosystems that would include significant
50 changes in structure and/or functioning, and/or loss of a substantial fraction of species richness (commonly
51 used to indicate loss of biodiversity). These are sourced mainly from Chapters 2 and 3, CCP1, and reference
52 the 1.5C report, Chapter 4 from WGII AR5, and Chapter 4 from WGII AR4 Reports.
53
54 Severe adverse impacts on biodiversity include significant risk of species extinction (e.g., loss of a
55 substantial fraction (one tenth or more) of species from a local to global scale), mass population mortality
56 (>50% of individuals or colonies killed), ecological disruption (order-of-magnitude increases or abrupt
57 reductions of population numbers or biomass), shifts in ecosystem structure and function (order-of
Do Not Cite, Quote or Distribute 16-61 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 magnitude increases or abrupt decreases in cover and/or biomass of novel growth forms or functional types),
2 and/or a socio-economically material increase in environmental risk (e.g., destruction by wildfire) or socio-
3 economically material decline in goods and services (e.g., carbon stock losses, loss of grazing, loss of
4 pollination). Metrics relevant to Sustainable Development Goals are also germane.
5
6 A substantial proportion of biodiversity is at risk of being lost below 2ºC of global warming (Chapter 2), due
7 to range reductions and loss globally, with this risk amplified roughly three times in insular ecosystems and
8 biodiversity hotspots, due to the increased vulnerability of endemic species (Manes et al., 2021). High
9 latitude, high altitude, insular, freshwater, and coral reef ecosystems and biodiversity hotspots (Chapter 2,
10 Cross-Chapter Paper 1 on Biodiversity Hotspots) are at appreciable risk of substantial biodiversity loss due
11 to climate change even under Low warming (high confidence). These systems comprise a large fraction of
12 unique and endemic biodiversity, with species impacts often exacerbated by multiple drivers of global
13 change (Chapter 2, Chapter 3). Roughly one third of all known plant species are extremely rare, vulnerable
14 to climate impacts, and clustered in areas of higher projected rates of anthropogenic climate change (Enquist
15 et al., 2019). Much evidence shows increased risk of the loss of 10% or more of terrestrial biodiversity with
16 increasing anthropogenic climate change (Urban, 2015; Smith et al., 2018) (medium confidence), likely with
17 2ºC warming above pre-industrial level (Chapter 2), with consequent degradation of terrestrial, freshwater,
18 and ocean ecosystems (Oliver et al., 2015) and adverse impacts on ecosystem services (Pecl et al., 2017) and
19 dependent human livelihoods (Dube et al., 2016). Adverse impacts on biodiversity may show lagged
20 responses (Essl et al., 2015), and loss of a substantial fraction of species could occur abruptly,
21 simultaneously across multiple taxa, below 4ºC of global warming (Trisos et al., 2020).
22
23 Mass population-level mortality (>50% of individuals or colonies killed) and resulting abrupt ecological
24 changes can be caused by simple or compound climate extreme events, such as exceedance of upper thermal
25 limits by vulnerable terrestrial species (Fey et al., 2015), who also note reduced mass mortality trends due to
26 extreme low thermal events); marine heatwaves that can cause mortality, enhance invasive alien species
27 establishment, and damage coastal ecological communities and small-scale fisheries (high confidence)
28 (Section 3.4.2.7); and increased frequency and extent of wildfires that threaten populations dependent on
29 habitat availability (like Koala Bears, Lam et al., 2020). Abrupt ecological changes are widespread and
30 increasing in frequency (Turner et al., 2020), and include tree mortality due to insect infestation exacerbated
31 by drought, and ecosystem transformation due to wildfire (Vogt et al., 2020). Freshwater ecosystems and
32 their biodiversity are at high risk of biodiversity loss and turnover due to climate change (precipitation
33 change and warming, including warming of water bodies), due to high sensitivity of processes and life
34 histories to thermal conditions and water quality (Chapter 2) (high confidence). In marine systems,
35 heatwaves cause damages in coastal systems, including extensive coral bleaching and mortality (very high
36 confidence) (Section 3.4.2.1), mass mortality of invertebrate species (low to high confidence, depending on
37 system) (Sections 3.4.2.2, Section 3.4.2.5, Section 3.4.4.1), and abrupt mortality of kelp-forest (high
38 confidence) (Section 3.4.2.3) and seagrass-meadow habitat (high confidence) (Section 3.4.4.2). The
39 biodiversity of polar seas shows strong impacts of climate change on phenological timing of plankton
40 activity, Arctic fish species range contractions and species community change (Table 16.2) (high
41 confidence). Extreme weather events and storm surges exacerbated by climate change have severe and
42 sudden adverse impacts on coastal systems, including loss of seagrass meadows and mangrove forests (high
43 confidence) (see Section 3.4.2.7, Section 3.4.2.8, Cross-Chapter Box EXTREMES in Chapter 2).
44
45 Ecological disruption (order-of-magnitude increases or abrupt reductions of population numbers or biomass)
46 can occur due to unprecedented inter-species interactions with unpredictable outcomes in `novel ecosystems'
47 (Chapter 2) as species shift geographic ranges idiosyncratically in response to climatic drivers (Table 16.2).
48 Idiosyncratic geographic shifts are now observed in an appreciable fraction of species studied (Chapter 2,
49 Table 16.2). Commensal or parasitic diseases may infect immunologically naive hosts (e.g., chytrid fungus
50 in amphibians). Atypical disturbance regimes may be enhanced, for example, with the spread of flammable
51 plant species (e.g., du Toit et al., 2015), exacerbated by introduced species (e.g., Martin et al., 2015), thus
52 significantly increasing risk of loss and damage to infrastructure and livelihoods, ecological degradation, and
53 challenging existing management approaches.
54
55 Landscape- and larger-scale shifts in ecosystem structure and function (order-of magnitude increases or
56 abrupt decreases in cover and/or biomass of novel growth forms or functional types) are occurring in non-
57 equilibrium ecosystems (systems which exist in multiple states, often disturbance-controlled) in response to
Do Not Cite, Quote or Distribute 16-62 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 changing disturbance regime, climate and rising CO2 (high confidence) Woody plant encroachment has been
2 occurring in multiple ecosystems, including sub-tropical and tropical fire driven grassland and savanna
3 systems, upland grassland systems, arid grasslands and shrublands (high confidence), leading to large scale
4 biodiversity changes, albedo changes, and impacts on water delivery, grazing services and human livelihoods
5 (medium confidence). Expansion of grasses (alien and native) into xeric shrublands is occurring causing
6 increasing fire prevalence in previous fire free vegetation (CCP3). In tropical forests repeated droughts and
7 recurrence of large-scale anthropogenic fires increase forest degradation, loss of biodiversity and ecosystem
8 functioning (high confidence) (Anderson et al., 2018b; Longo et al., 2020). Accelerated growth rates and
9 mortality of tropical trees is also adversely affecting tropical ecosystem functioning (McDowell et al., 2018;
10 Aleixo et al., 2019). Projected changes in ecosystem functioning, such as via wildfire (Section 2.5.5.2), tree
11 mortality (Section 2.5.5.3) and woody encroachment under climate change (Chapter 2) would alter
12 hydrological processes, with adverse implications for water yields and water supplies (Sankey et al., 2017;
13 Robinne et al., 2018; Rodrigues et al., 2019; Uzun et al., 2020).
14
15 The loss of a substantial fraction of biodiversity globally, abrupt impacts like significant local biodiversity
16 loss and mass population mortality events, and ecological disruption due to novel species interactions have
17 been observed or are projected at global warming levels below 2ºC (Chapter 2 Table 2.S.4, Cross Chapter
18 Box: EXTREMES in Chapter 2, Section 2.4.4.3.1, Section 2.4.2.3.3) (medium confidence). Simple and
19 compound impacts of extreme climate events are already causing significant loss and damage in vulnerable
20 ecosystems, including through the facilitation of important global change drivers of ecological disruption and
21 homogenisation like invasive species (high confidence). Severe impacts on human livelihoods and
22 infrastructure, and valuable ecosystem services are all projected to accompany these changes. Adaptation
23 potential for many of these risks is low due to the projected rate and magnitude of change, and to the
24 requirement of significant amounts of land for terrestrial ecosystems (Hannah et al., 2020). Biodiversity
25 conservation efforts may be hampered due to climate change impacts on the effectiveness of protected areas,
26 with high sensitivity of effectiveness to forcing scenario (medium confidence). In addition, climate-related
27 risks to ecosystems pose challenges to ecosystem-based adaptation responses (`nature-based solutions')
28 (Section 2.1.3) (medium confidence).
29
30 16.5.2.3.3 Risk to critical physical infrastructure and networks (RKR-C)
31 RKR-C includes risks associated with the breakdown of physical infrastructure and networks which provide
32 goods and services considered critical to the functioning of societies. It encompasses infrastructure systems
33 for energy, water, transportation, telecommunications, health care and emergency response, as well as
34 compound, cascading and cross-boundary risks resulting from infrastructure interdependencies (Birkmann et
35 al., 2016; Fekete, 2019). Critical infrastructures such as transport or energy supply also play a central role in
36 coping with climate risks, especially in acute disaster situations in which the services of transport
37 infrastructure, communication technologies or electricity are particularly needed, despite the fact that these
38 very systems are themselves exposed to disaster impacts (Garschagen et al., 2016; Pescaroli et al., 2018).
39 The major hazards driving such risks are acute extreme events such as cyclones, floods, droughts or fires
40 (high confidence), but cumulative and chronic hazards such as sea level rise (SLR) are also considered.
41
42 RKR-C is considered severe when the functioning of critical infrastructure cannot be secured and maintained
43 against climate change impacts, resulting in the frequent and widespread breakdown of service delivery and
44 eventually a significant rise of detrimental impacts on people (lives, livelihoods and well-being), the
45 economy (including averted growth) or environment (disruption and loss of ecosystems) above historically
46 observed levels. Severity in this RKR is assessed on two levels for (i) direct impacts of climate change on
47 infrastructure assets and networks (e.g., amount of port infrastructure damaged or destroyed by SLR,
48 flooding and storms) on which most of the literature focuses, as well as (ii) indirect and cascading
49 downstream impacts to people, economy and environment (Markolf et al., 2019; Pyatkova et al., 2019;
50 Chester et al., 2020), for which attribution is more difficult and uncertainties tend to be much higher.
51 Overall, the literature with quantified assessments of climate change infrastructure risks remains to be less
52 extensive than for many other risks, particularly with regards to assessments focusing on the Global South.
53 While climate-related changes in hazards are widely considered in the literature, changes in future exposure
54 and vulnerability conditions are often not treated explicitly. In addition, the severity of infrastructure risks
55 also depends on future trends in the capacity to maintain, repair and rebuild infrastructure and adapt it to new
56 hazard intensities (medium evidence, high agreement). These are mostly not quantified in a forward-looking
Do Not Cite, Quote or Distribute 16-63 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 manner in the literature; however, damage projections (see below) indicate a rapidly rising demand for
2 investment, straining the financial capacity of countries (medium evidence, high agreement).
3
4 (i) Risks related to direct impacts on critical infrastructure would become severe with high warming, current
5 infrastructure development regimes and minimal adaptation (high confidence), and in some contexts even
6 with low warming, current vulnerability and no additional adaptation (medium confidence), with severity
7 defined as infrastructure damage and required maintenance costs exceeding multiple times the current levels.
8 Transport and energy infrastructure in coasts, polar systems and along rivers are projected to face a
9 particularly steep rise in risk, resulting in severe risk even under medium warming (high confidence). Risk in
10 relation to the increasing intensity and frequency of extreme events might become severe before the middle
11 of the century (medium confidence). Damages from multiple climate hazards to transport, energy, industry
12 and social infrastructure in Europe could increase tenfold by the 2080s, from 3.4 billion annually to date,
13 and 15-fold for transport infrastructure, under Medium warming (A1B, ~3°C by 2100) and with current
14 adaptation levels, even if no further extension of the infrastructure in exposed areas is considered (Forzieri et
15 al., 2018). Under High warming (RCP8.5) in 2100, the percent of roads in the United States that require
16 rehabilitation due to high temperatures and precipitation is expected to increase to 2333%, relative to 14%
17 in 2100 when no climate change is considered (Mallick et al., 2018). Projections of climate-induced changes
18 in exposure are an incomplete measure of risk but in the absence of other metrics can serve as a proxy for the
19 potential for severe impacts. In the circumpolar Arctic, 14.8% of critical infrastructure assets would be
20 affected by climate change under RCP8.5 by 2050, with lifecycle replacement costs projected to increase by
21 27.7% if infrastructure is to be preserved at current adaptation levels (Suter et al., 2019). Under RCP8.5, the
22 number of ports under high risk will increase from 3.8% in the present day to 14.4% by 2100, as a result of
23 increased coastal flooding and overtopping due to sea level rise, as well as the heat stress impacts of higher
24 temperatures (Izaguirre et al., 2021). In the UK under High warming (4ºC), the number of clean and
25 wastewater treatment sites located in the 1 in 75-year floodplain will increase by a third relative to today by
26 the 2080s under current vulnerability and adaptation levels (Dawson et al., 2018). A global assessment of
27 changing climate and water resources for electricity generation finds considerable reductions in usable
28 hydropower and thermoelectric capacity by 2050 for a range of warming scenarios from Low to High, with
29 absolute declines on average for most (6174%) of the world's hydropower resources and monthly
30 maximum reductions above 30% of usable capacity for over two-thirds of 1,427 thermoelectric power plants
31 worldwide (Van Vliet et al., 2016). Many studies find large technical potential for coordinated adaptation-
32 mitigation policies in the electricity sector to avoid a significant portion of projected climate change impacts
33 (e.g., a two-thirds reduction, and in some cases fully offset) (Ciscar and Dowling, 2014; Van Vliet et al.,
34 2016; Gerlak et al., 2018; Allen-Dumas et al., 2019).
35
36 (ii) Studies quantifying the indirect impacts of infrastructure failure on lives, livelihoods and economies are
37 still rare but emerging, suggesting that risks would become severe in many contexts globally with high
38 warming, current vulnerability and no additional adaptation (medium confidence). Severity in this context is
39 defined as the potential to disrupt the lives, livelihoods and well-being of a significantly increased proportion
40 of the population and to significantly forestall economic growth and development potential. Global risks to
41 air travel from SLR, expressed in terms of expected annual route disruptions, could increase by a factor of
42 between 17 and 69 by 2100 under the 1.5 C and the 95th percentile value of the RCP8.5 SLR scenario,
43 respectively (Yesudian and Dawson, 2021). By 2050, up to 185,000 airline passengers per year may be
44 grounded due to extreme heat (48°C) if no additional adaptation is taken, roughly 23 times more than today
45 (McKinsey Global Institute, 2020). In Africa, under RCP8.5 and without additional adaptation a 250%
46 increase in disruption time of the transport network is expected by 2050 due to extreme temperatures, a 76%
47 increase due to precipitation, and 1400% increase due to flooding (Cervigni et al., 2015). On the Dawlish
48 railway section (UK), the number of days with line restrictions are set to increase by up to 1170%, to as
49 many as 84120 per year by 2100 due to 0.8m SLR with High warming (Dawson et al., 2016). Next to the
50 limited number of projections or scenarios of indirect impacts, additional inferences from studies focusing on
51 past and current impacts can be drawn. Already today, climate-related impacts on transport and energy
52 infrastructure reach far beyond the direct impacts on physical infrastructure, triggering indirect impacts on,
53 for example, health and income (medium confidence). A case study of future flood hazard in Europe found
54 that the indirect impact of a power outage on the local economy is six to eight times greater than the direct
55 flood damage and asset repair costs, due to the interruption of daily economic activity (Karagiannis et al.,
56 2019). In low and middle-income countries, the annual costs from infrastructure disruptions reach up to 300
57 billion USD for firms and 90 billion USD for private households, with natural hazards such as floods being
Do Not Cite, Quote or Distribute 16-64 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 responsible for 10 to 70 % of these disruptions, depending on the sectors and regions (Hallegatte et al.,
2 2019). Power outages triggered by floods or droughts have also been found to have substantial health
3 implications, particularly amongst low-income populations (Klinger et al., 2014), and shown to impede
4 disaster recovery efforts and severely disrupt local economies (Karagiannis et al., 2019; Nicolas et al., 2019).
5 In addition, risks associated with infrastructure have the potential to become particularly severe when
6 hazard-driven infrastructure disruptions undermine the capacity of emergency response in disaster situations
7 (low evidence, high agreement). A study on the UK shows, for example, that even a small increase in minor
8 road flooding leads to a disproportionately high disruption of the efficacy of emergency services (Yu et al.,
9 2020). Similar risks have been found for rural areas, particularly in developing countries (Alegre et al.,
10 2020).
11
12 16.5.2.3.4 Risk to living standards (RKR-D)
13 This RKR includes risks to (i) aggregate economic output at the global and national levels, (ii) poverty, and
14 (iii) livelihoods, and their implications for economic inequality. It is informed by key risks identified by
15 regional and sectoral chapters. Risks are potentially severe as measured by the magnitude of impacts in
16 comparison to historical events or as inferred from the number of people currently vulnerable.
17
18 (i) Risks to aggregate economic output would become severe at the global scale with high warming and
19 minimal adaptation (medium confidence), with severity defined as the potential for persistent annual
20 economic losses due to climate change to match or exceed losses during the world's worst historical
21 economic recessions. With historically observed levels of adaptation, warming of ~4ºC may cause a 10-23%
22 decline in annual global GDP by 2100 relative to global GDP without warming, due to temperature impacts
23 alone (Burke et al., 2015; Kahn et al., 2019; Kalkuhl and Wenz, 2020). These magnitudes exceed economic
24 losses during the Great Recession (2008-2009, ~5% decline in global GDP, up to 15-18% in some countries)
25 and the COVID-19 pandemic (2020, ~3% decline globally, up to 10% in some countries) (IMF, 2020; IMF,
26 2021). Unlike past recessions, climate change impacts would occur continuously in every year. However,
27 smaller effects (1-8%) are found when using alternative methodologies (Diaz and Moore, 2017; Nordhaus
28 and Moffat, 2017; Kompas et al., 2018; Kalkuhl and Wenz, 2020), assuming less warming (Kahn et al.,
29 2019; Takakura et al., 2019), and assuming lower vulnerability and/or more adaptation (Diaz and Moore,
30 2017); this literature is comprehensively summarized in Cross-Working Group Chapter Box ECONOMIC.
31 Impacts at high levels of warming are particularly uncertain, as all methodologies require extrapolation and
32 insufficiently incorporate possible tipping elements in the climate system (Kopp et al., 2016).
33
34
35
36 Figure 16.9. Illustrative examples from individual studies of risks to living standards and the conditions under which
37 they could become severe. Selected studies are not representative of the literature, but provide examples of potentially
Do Not Cite, Quote or Distribute 16-65 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 severe risks to aggregate economic output, poverty, and livelihoods. High, medium, and low levels of warming,
2 exposure/vulnerability, and adaptation are defined as in Figure 16.10.
3
4
5 Annual economic output losses in developing countries could exceed the worst country-level losses during
6 historical economic recessions (medium confidence). Assuming global warming of ~4ºC by 2100, historical
7 adaptation levels, and high vulnerability, losses across sub-Saharan Africa may reach 12% of GDP by 2050
8 (Baarsch et al., 2020) and 80% by 2100 (Burke et al., 2015), and ~9% on average across developing
9 countries by 2100 (Acevedo et al., 2017). The largest estimates are debated and depend on assumptions
10 about development trends, adaptive capacity, and whether temperature impacts the level or growth rate of
11 economic activity (Kalkuhl and Wenz, 2020). Severe risks are more likely in (typically hotter) developing
12 countries because of nonlinearities in the relationship between economic damages and temperature (Burke et
13 al., 2015; Acevedo et al., 2017). These risks are highest in scenarios and countries with: a large portion of the
14 workforce employed in highly exposed industries (Acevedo et al., 2017); a high concentration of population
15 and economic activity on coastlines (Hsiang and Jina, 2014; Acevedo et al., 2017); and an increase in the
16 frequency or intensity of disasters triggered by natural hazards (Berlemann and Wenzel, 2018; Botzen et al.,
17 2019). Whether baseline economic growth may help avoid severe future risks is highly uncertain (Dell et al.,
18 2012; Burke et al., 2015; Acevedo et al., 2017; Deryugina and Hsiang, 2017).
19
20 (ii) Under medium warming pathways, climate change risks to poverty would become severe if vulnerability
21 is high and adaptation is low (limited evidence, high agreement). We define poverty in terms of absolute
22 consumption levels and define severity as tens to hundreds of millions of additional people in poverty
23 relative to the number without change (globally) or an absolute increase in the number of people living in
24 poverty compared to today (nationally or locally). This global impact is comparable to the effect of the 2007
25 food price shock (De Hoyos and Medvedev, 2009) and the 2020 COVID-19 pandemic (World Bank, 2020)
26 and can be compared to about 700 million in poverty in 2017, down from 1.9 billion in 1990 (World Bank,
27 2020).
28
29 In a high-vulnerability development pathway, climate change in 2030 could push 35-132 million people into
30 extreme poverty, in addition to the people already in poverty assuming climate is unchanged (disregarding
31 impacts from natural variability; Hallegatte and Rozenberg, 2017; Jafino et al., 2020). In a low warming
32 pathway, risks from mitigation costs could also be severe if no progressive redistribution from carbon pricing
33 revenues is applied (Soergel et al., 2021). At the national level there is limited evidence of climate change
34 causing an absolute increase in poverty (e.g., absolute increase of ~1-2%/yr through 2040, Montaud et al.,
35 2017). Potentially severe risks to poverty are also supported by (1) the observed impacts of past disasters
36 (Winsemius et al., 2018; Hallegatte et al., 2020; Rentschler and Melda, 2020) and previous crises such as
37 food price shocks (Ivanic and Martin, 2008) or current diseases (WHO, 2018) on poor people and on
38 poverty; (2) the expectation that these events will become more intense or frequent in some regions (WGI
39 Chapter 12, Ranasinghe et al., 2021); and (3) population growth and the low adaptive and coping capacities
40 of the poor (Leichenko and Silva, 2014; Huynh and Stringer, 2018; Thomas et al., 2020). This literature
41 provides indirect evidence that climate change will keep many people poor and may cause more than tens of
42 millions to fall into poverty (low evidence, high agreement).
43
44 (iii) Climate change poses severe risks to livelihoods at low levels of warming, high exposure/vulnerability,
45 and low adaptation in climate-sensitive regions, ecosystems, and economic sectors (high confidence), where
46 severity refers to the disruption of livelihoods for tens to hundreds of millions of additional people (Arnell
47 and Lloyd-Hughes, 2014; Liu et al., 2018). More widespread severe risks would occur at high levels of
48 warming (with high exposure/vulnerability and low adaptation) where there is additional potential for one or
49 more social or ecological tipping points to be triggered (Cai et al., 2015; Cai et al., 2016b; Kopp et al., 2016;
50 Steffen et al., 2018; Lenton et al., 2019), and for severe impacts on livelihoods to cascade from relatively
51 more climate-sensitive to relatively less climate-sensitive sectors and regions (medium confidence)
52 (Lawrence et al., 2020). Severity assessment is based on the current magnitude of exposure and vulnerability
53 across multiple social and ecological systems, projected future exposure and vulnerability, and the rate at
54 which hazard frequency or intensity is expected to increase (Otto et al., 2017; Roy et al., 2018; Li et al.,
55 2019, Section 8.5). Without effective adaptation measures, regions with high dependence on climate-
56 sensitive livelihoods particularly agriculture and fisheries in the tropics and coastal regions would be
57 severely impacted even at low levels of warming (high confidence) (Hoegh-Guldberg et al., 2018b; Roy et
Do Not Cite, Quote or Distribute 16-66 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 al., 2018). For example, it is estimated that 330396 million people could be exposed to lower agricultural
2 yields and associated livelihood impacts at warming between 1.5 and 2ºC (Byers et al., 2018). Risks to the
3 200 million people with livelihoods derived from small-scale fisheries would also be severe, given
4 sensitivity to ocean warming, acidification, and coral reef loss occurring beyond 1.5ºC (Cheung et al., 2018b;
5 Froehlich et al., 2018; Free et al., 2019; Barnard et al., 2021). Livelihoods in highly exposed locations, such
6 as small-island developing states, low-lying coastal areas, arid or semi-arid regions, the Arctic, and urban
7 informal settlements or slums, are particularly vulnerable (Ford et al., 2015c; Hagenlocher et al., 2018;
8 Ahmadalipour et al., 2019; Tamura et al., 2019). Within populations, the poor, women, children, the elderly,
9 and indigenous populations are especially vulnerable due to a combination of factors including gendered
10 divisions of paid and/or unpaid labour, as well as barriers in access to information, skills, services, or
11 resources (Bose, 2017; Thomas et al., 2019b; Anderson and Singh, 2020; Adzawla and Baumüller,
12 2021)(high confidence). Future structural transformation could moderate risk severity by improving adaptive
13 capacity, creating livelihoods in less climate-sensitive sectors, or by enabling sustainable migration to less
14 climate-sensitive locations (Henderson et al., 2017; Roy et al., 2018). However, successful risk moderation
15 would depend upon simultaneous avoidance of both climate change-related and mitigation-related (Doelman
16 et al., 2019; Fujimori et al., 2019; Doelman et al., 2020) or maladaptation-related risks (Magnan et al., 2016;
17 Benveniste et al., 2020; Schipper, 2020).
18
19 Climate change also could increase income inequality between countries (high confidence) as well as within
20 them (medium evidence, high agreement) that result from and exacerbate impacts on aggregate economic
21 activity, poverty, and livelihoods. Increasing inequality implies larger impacts on the least well-off, threatens
22 their ability to respond to climate hazards, compromises basic principles of fairness and established global
23 development goals, and potentially threatens the functioning of society and long-term progress (Roe and
24 Siegel, 2011; Cingano, 2014; van der Weide and Milanovic, 2018). There is evidence that warming has
25 slowed down the convergence in between-country income in recent decades (Diffenbaugh and Burke, 2019).
26 Future impacts may halt or even reverse this trend during this century due to high sensitivity of developing
27 economies (Burke et al., 2015; Pretis et al., 2018; Baarsch et al., 2020), although projections depend as much
28 or more on future socioeconomic development pathways and mitigation policies as on warming levels
29 (Takakura et al., 2019; Harding et al., 2020; Taconet et al., 2020). Within countries, studies that find adverse
30 impacts on low-income groups imply an increase in inequality (Hallegatte and Rozenberg, 2017; Hsiang et
31 al., 2017), although evidence for long-term climate impacts on within-country inequality at global scale
32 remains limited.
33
34 16.5.2.3.5 Risk to human health (RKR-E)
35 This RKR includes (i) mortality from heat, and morbidity and mortality from (ii) vector-borne diseases and
36 (iii) waterborne diseases. It builds on KRs identified primarily in Chapter 7 and health risks in regional
37 chapters.
38
39 A severe risk to health is the potential for a widespread, substantial worsening of health conditions due to
40 climate change. We measure severity in terms of the magnitude of mortality and morbidity. We consider a
41 severe mortality impact to be a sustained increase in the crude mortality rate (CMR) of more than about 2-4
42 deaths per 10,000 people per year, or 2-5% over the current background rate. This range of increase is
43 consistent with current mortality impacts with substantial global effects, including traffic fatalities (CMR of
44 1.6/10,000/yr, IHME) and the COVID-19 pandemic (4/10,000/yr, as of April, 2021, expressed as an
45 annualized rate (Ritchie, Hannah et al., 2021). We use these global rates as thresholds in all cases,
46 recognizing that they reflect substantial variation across regions and sub-populations (other points of
47 comparison are included in Table SM16.13). Morbidity impacts are measured in numbers of disease cases or
48 hospital admissions. We find that severe health impacts are projected to occur for particular sub-populations
49 and regions where vulnerability is currently high and is assumed to persist into the future; we focus our
50 assessment on these cases. In other cases, literature is either inadequate or does not support severe outcomes.
51
52 (i) Risks of heat-related mortality would become severe at global and regional scales with high levels of
53 warming and vulnerability (high confidence). Under these conditions (SSP3-8.5), accounting for adaptation,
54 heat mortality would increase the global CMR by up to 2-7/10,000/yr by 2100 (Carleton et al., 2020). For
55 example, the US would experience a CMR increase of 2-4/10,000/yr by the end of the century (medium
56 vulnerability without adaptation, and recent vulnerability with adaptation, respectively) (Weinberger et al.,
57 2017; Shindell et al., 2020). Also assuming no adaptation and recent vulnerability, every population of the
Do Not Cite, Quote or Distribute 16-67 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 world would experience an increase of 2-10 percentage points in the proportion of deaths attributable to heat
2 by the end of the century (RCP8.5). Harmful conditions for health are expected to increase in frequency and
3 intensity over all land areas along with the rising temperatures in the coming decades (Pal and Eltahir, 2016;
4 Russo et al., 2017; Ranasinghe et al., 2021; Saeed et al., 2021; Schwingshackl et al., 2021). Projections of
5 exposure are an incomplete measure of risk but suggest the potential for severe impacts. For example, the
6 percent of global population exposed to deadly heat stress would increase from today's 30% to 48-74% by
7 the end of the century depending on level of warming and population distribution (Mora et al., 2017).
8 Projected impacts are larger if exposure and/or vulnerability increases due to ageing of the population or
9 increased inequality (Weinberger et al., 2017; Chen et al., 2020a; IPCC, 2021a) and with limited adaptation
10 capacity (e.g., poor infrastructure, limited air conditioning, few medical and public health resources) (Table
11 SM16.4) (Carleton et al., 2020). Higher risks are also expected in urban areas due to hazard amplification
12 (i.e., urban heat island effect) and in highly dense settlements with other environmental hazards such as air
13 pollution (Zhao et al., 2018; Sera et al., 2019).
14
15 (ii) Risks of vector-borne disease would become severe with high warming and current vulnerability,
16 concentrated in children and in sensitive regions (medium confidence). Severity is defined by regionally
17 substantial numbers of additional malaria deaths, disease cases, and episodic hospitalisation demands (for
18 dengue).
19
20 With high warming, the CMR for malaria among children under the age of one year could increase by 5.2-
21 10.1/10,000/yr in Africa under current vulnerability levels. This estimate assumes a net increase of 70-130
22 million more people exposed to potential disease transmission due to climate change in a high warming
23 scenario (RCP8.5, end of century)(Caminade et al., 2014; Colón-González et al., 2018; Ryan et al., 2020),
24 representing a 14-27% increase in the current population at risk (Ryan et al., 2020), and assumes children
25 under 1 year of age are facing the same crude mortality in the future as for the African region today (Table
26 SM16.13). The largest increase is observed in Eastern Africa, where the population exposed could nearly
27 double by 2080 (Ryan et al., 2020) without accounting for population growth, driven mainly by changes
28 among previously unexposed populations at higher altitude areas (Colón-González et al., 2018). Actual
29 future disease burden of malaria will be highly sensitive to regional socio-economic development and the
30 effectiveness of malaria intervention programs.
31
32 For dengue, with high warming and current levels of vulnerability there could be as many as a doubling of
33 cases and hospital admissions per year globally, relative to today, driven by both warming and population
34 growth. These estimates are derived by assuming similar relative incidence rates as today (Shepard et al.,
35 2016) combined with projections of a more than doubling of the population exposed to potential disease
36 transmission by the end of the century in a high warming scenario (RCP8.5), although much of this increase
37 is driven by population growth (Colón-González et al., 2018; Monaghan et al., 2018; Messina et al., 2019).
38 There are around 3 billion people exposed to dengue today.
39
40 (iii) Climate change would lead to severe risks of morbidity and mortality caused by waterborne diseases,
41 particularly for diarrhoea in children in many lower- and middle-income countries (LMICs) and where
42 vulnerability remains high (medium confidence). The global CMR for diarrhoea is 1.98 for all ages, but
43 varies by region and age group, reaching as high as 53 for <1 year olds in Africa (Institute for Health Metrics
44 and Evaluation (IHME), 2021) . In these vulnerable populations even a small percentage increase can lead to
45 substantial additional morbidity and mortality. For example, assuming no change in vulnerability or
46 population, an increase in diarrhoea mortality of only 5% over 2019 baseline rates would create a severe risk
47 (CMR of 2.0) for children under the age of 1 in the WHO Africa (AFRO) region. This percent increase due
48 to climate change is plausible since diarrhoea incidence increases of 7% (95% confidence interval 3-10%)
49 are associated with a 1ºC increase in ambient temperature (WHO, 2014; Carlton et al., 2016), and diarrhoea
50 is positively associated with heavy rainfall and flooding events (Levy et al., 2016), expected in some regions
51 (WGI). Assuming vulnerability remains the same as today, mortality and morbidity rates would increase
52 equivalently.
53
54 However, risks will be highly dependent on development trajectories, given that waterborne disease
55 transmission is exacerbated by lack of clean drinking water and sanitation systems, inadequate food safety
56 and hygiene conditions, lack of flood and drought protections, and interactions with other risks such as
57 cholera outbreaks, food insecurity, and infrastructure damage. Climate change threatens the progress that has
Do Not Cite, Quote or Distribute 16-68 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 been made toward reducing the burden of diarrhea. For example, in Sub-Saharan Africa, while overall
2 diarrhea rates are expected to continue to decline (GBD 2016 Diarrhoeal Disease Collaborators, 2018),
3 warming in 2030 (relative to the late 20th century) is projected to lead to diarrheal deaths in children under
4 15 equivalent to a CMR increase of 0.56/10,000/yr (based on population projections for the region and age
5 group (UN, 2020)) (WHO, 2014). In China, by 2030 climate change could delay progress toward reducing
6 waterborne disease burden by 8-85 months (Hodges et al., 2014).This RKR includes (i) mortality from heat,
7 and morbidity and mortality from (ii) vector-borne diseases and (iii) waterborne diseases. It builds on KRs
8 identified primarily in Chapter 7 and health risks in regional chapters.
9
10 16.5.2.3.6 Risk to food security (RKR-F)
11 Climate change affects food security primarily through impacts on food production, including crops,
12 livestock, and fisheries, as well as disruptions in food supply chains, linked to global warming, drought,
13 flooding, precipitation variability and weather extremes (Myers et al., 2017; FAO et al., 2018; Mbow et al.,
14 2019). This RKR builds on Key Risks identified primarily in the Food, Fibre, and other Ecosystem Products
15 Chapter, some sectoral (Health), and regional (Africa, Australasia, Central and South America, North
16 America) chapters, as well as SR15, SRCCL and SROCC.
17
18 The severity of the risk to food security is defined here using a combination of criteria including the
19 magnitude and likelihood of adverse consequences, affecting 10s to 100s of millions of people, timing of the
20 risk and ability to respond to the risk. In this assessment, we use the number of undernourished people as a
21 proxy outcome of these dimensions and their multiple interactions.
22
23 Climate change will pose severe risks in terms of increasing the number of undernourished people, affecting
24 tens to hundreds of million people under High vulnerability and High warming, particularly among low-
25 income populations in developing countries (high confidence). Extreme weather events will increase risks of
26 undernutrition even on a regional scale, via spikes in food price and reduced income (high confidence) (FAO
27 et al., 2018, Hickey and Unwin, 2020; Mbow et al., 2019). The timing of these impacts and our ability to
28 respond to them vary based on the level of GHG emissions and Shared Socioeconomic Pathways (SSP).
29 Under a low vulnerability development pathway (SSP1), climate change starts posing a moderate risk to food
30 security above 1ºC of global warming (i.e., impacts become detectable and attributable to climate-related
31 factors), while beyond 2.5ºC the risk becomes high (widespread impacts on larger numbers or proportion of
32 population or area, but with the potential to adapt or recover) (Hurlbert et al., 2019). Under high
33 vulnerability-high warming scenario (i.e., SSP3-RCP6.0), up to 183 million additional people are projected
34 to become undernourished in low income countries due to climate change by 2050 (Mbow et al., 2019).
35 Climate-related changes in food availability and diet quality are estimated to result in a crude mortality rate
36 of about 54 deaths per million people with about 2ºC warming by 2050 (SSP2, RCP8.5), most of them
37 projected to occur in South and East Asia (67-231 deaths per million depending on the country) (Springmann
38 et al., 2016). In a medium vulnerability-high warming scenario (SSP2, RCP6.0), Hasegawa et al. (2018)
39 projects that the number of undernourished people increases by 24 million in 2050, compared to outcomes
40 without climate change and accounting for the CO2-fertilization effect. This number increases by around 78
41 million in a low warming scenario (RCP2.6) accounting for the impacts of both climate change and
42 mitigation policies. Caveats to these modelling studies are that most models (crop models in particular) are
43 designed for long-term change in climate but not suited to project the impacts of short-term extreme events.
44 The inclusion of adaptation measures into modeling estimates remains selective and partial.
45
46 Climate change risks of micronutrient deficiency will become severe in high vulnerability development
47 pathways and in the absence of societal adaptation, leading to hundreds of millions of additional people
48 lacking key nutrients for atmospheric CO2 levels above 500 ppm (high confidence) (Myers et al., 2017;
49 Nelson et al., 2018; Mbow et al., 2019). For example, concentration of many micronutrients (e.g.,
50 phosphorus, potassium, calcium, sulphur, magnesium, iron, zinc, copper, and manganese) can decrease by 5-
51 10% under atmospheric CO2 concentrations of 690 ppm (3.5ºC warming). The decline in zinc content is
52 projected to lead to an additional 150-220 million people affected by zinc deficiency with increases in
53 existing deficiencies in more than 1 billion people (Myers et al., 2017). Similarly, decrease in protein and
54 micronutrient content in rice due to a higher CO2 concentration (568 to 590 ppm) can lead to 600 million
55 people with rice as a staple at risk of micronutrient deficiency by 2050 (Zhu et al., 2018). Additionally, the
56 impact on protein content of increased CO2 concentration (> 500 ppm) can lead an additional 150 million
Do Not Cite, Quote or Distribute 16-69 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 people with protein deficiency by 2050 (within the total of 1.4 billion people with protein deficiency) in
2 comparison to the scenario without increased CO2 concentration (Medek et al., 2017).
3
4 16.5.2.3.7 Risk to water security (RKR-G)
5 Water security encompasses multiple dimensions: water for sanitation and hygiene, food production,
6 economic activities, ecosystems, water-induced disasters, and use of water for cultural purposes (Chapter 4;
7 Box 4.1; Section 4.6.1). Water security risks are a combination of water-related hazards such as floods,
8 droughts, and water quality deterioration, and exposure of vulnerable groups exposed to too little, too much,
9 or contaminated water. Reasons for these can include both environmental conditions and issues of safety and
10 access influenced by effectiveness of water governance (Sadoff et al., 2020). These are manifest through loss
11 of lives, property, livelihoods and culture, and impacts on human health and nutrition, ecosystems and water-
12 related conflicts which in turn can drive forced human displacement.
13
14 This RKR focuses on three types of risks with the potential to become severe: those associated with water
15 scarcity, those driven by water-related disasters, and those impacting indigenous and traditional cultures and
16 ways of life. Risk to water security constitutes a potentially severe risk because climate change could impact
17 the hydrologic cycle in ways that would lead to substantial consequences for the health, livelihoods,
18 property, and cultures of large numbers of people. For those associated with water scarcity, `severe' refers to
19 magnitude (number of people in areas where water scarcity falls below recognised thresholds for adequate
20 water supply per capita), along with the likelihood of unforeseen increases in water scarcity that outpace the
21 ability to prepare for the increased risk by putting in place new large-scale infrastructure within the required
22 timescale. For those associated with extreme events, `severe' refers to magnitude (numbers of people
23 affected, including deaths, physical health impacts including disease, mental health impacts, loss of
24 livelihoods, loss of or damage to property) and timing (for example, events coinciding with other stresses,
25 e.g., a pandemic occurring at a time when local infrastructures are weakened by an extreme weather event).
26 Important water-related extreme events include river flooding caused by heavy and/or prolonged rainfall,
27 glacial lake outburst floods, and droughts. For those impacting cultures, `severe' refers to the loss of key
28 aspects of traditional ways of life. This includes consequences of the above two key risks.
29
30 Risks associated with water scarcity have the potential to become severe based on projections of large
31 numbers of people becoming exposed to low levels of water availability per person, where `water
32 availability' includes fresh water in the landscape, including soil moisture and streamflows, available for all
33 uses including agriculture as a dominant sector. Approximately 1.6 billion people currently experience
34 `chronic' water scarcity, defined as the availability of less than 1000 m3 of renewable sources of fresh water
35 per person per year (Gosling and Arnell, 2016). In this context, we define a severe outcome as an additional
36 1 billion people experiencing `chronic' water scarcity, relating to all uses of water, representing an increase
37 of a magnitude comparable with current levels. The global number of people experiencing chronic water
38 scarcity is projected to increase by approximately 800 million to 3 billion for 2°C global warming, and up to
39 approximately 4 billion for 4°C global warming, considering the effects of climate change alone, with
40 present-day population (Gosling and Arnell, 2016). Severe outcomes are projected to occur even with no
41 changes in exposure: present-day exposure is defined here as `medium' since either an increase or decrease
42 in exposure could be possible. Vulnerability is not quantified in the literature assessed here, so in this
43 assessment it is considered that severe outcomes could occur with present-day levels of vulnerability, again
44 defined here as `medium'. Particularly severe outcomes (i.e., the high end of these ranges) are driven by
45 regional patterns of climate change bringing severe reductions in precipitation and/or high levels of
46 evapotranspiration in the most highly-populated regions, leading to very substantial reductions in water
47 availability compared to demand. There is strong consensus across models that water scarcity is projected to
48 increase across substantial parts of the world even though projections disagree on which specific areas would
49 see this impact. Moreover, a projected decrease in water scarcity in some regions does not prevent the
50 increase in water scarcity in other regions becoming severe. Hence there is high confidence that risks to
51 water scarcity have the potential to become severe due to climate change. Consequences of water scarcity
52 include potential competition and conflicts between water users (Vanham et al., 2018), damaging
53 livelihoods, hindering socio-economic development, and reducing human well-being, for example through
54 malnutrition resulting from inadequate water supplies leading to long-term health impacts such as child
55 stunting (Cooper et al., 2019). The avoidance of these consequences at high levels of water scarcity would
56 require transformational adaptations including large-scale interventions such as dams and water transfer
57 infrastructure (Greve et al., 2018). Since these require many years or even decades for planning and
Do Not Cite, Quote or Distribute 16-70 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 construction, and are also costly and irreversible and can potentially lead to lock-in and maladaptation, the
2 potential for inadequate policy decisions made in the context of high uncertainties in regional climate
3 changes brings the risk of a shortfall in adaptation. Around 2050, at approximately 2°C global warming, the
4 risk of a substantial adaptation shortfall and hence severe outcomes for water scarcity have a relatively high
5 likelihood across large parts of the southern USA and Mexico, northern Africa, parts of the Middle-East,
6 northern China, and southern Australia, as well as many parts of Northwest India and Pakistan (Greve et al.,
7 2018).
8
9 Risks associated with water-related extreme events and disasters have the potential to become severe based
10 on projections of large numbers of people or high values of assets being affected. The risks to people from
11 disasters can often only be quantified in terms of the hazard and exposure (the number of people affected),
12 rather than the full consequences such as number of deaths, injuries or other health outcomes, as these often
13 depend on complex or unpredictable factors such the effectiveness of emergency and humanitarian responses
14 or the access to healthcare. With approximately 50 million people per year currently affected by flooding
15 (Alfieri et al., 2017), we define severe outcomes as more than 100 million people affected by flooding. At
16 2°C global warming, between approximately 50 million and 150 million people are projected to be affected
17 by flooding, with figures rising to 110 million to 330 million at 4°C global warming. These projections
18 assume present-day population and no additional adaptation, so no changes in exposure. Increased flood risk
19 is projected by the WHO to lead to an additional 48,000 deaths of children under 5 years due to diarrhoea by
20 2030, with Sub-Saharan Africa impacted the most (WHO, 2014). Other consequences of floods that already
21 occur include deaths by drowning, loss of access to fresh water, vector-borne diseases, mental health
22 impacts, loss of livelihoods, and loss of or damage to property. Many of these consequences depend on the
23 vulnerability of individuals, households or communities to flooding impacts, for example through the
24 presence or absence of measures to safeguard health and livelihoods, such as through infrastructure services,
25 insurance or community support. The risks associated with these consequences could increase if there were
26 no local adaptations to counter the effect of increased levels of hazard by reducing exposure and/or
27 vulnerability. Climate-related changes to extreme events that would lead to these severe outcomes: increased
28 frequency and/or magnitude of river floods of flash floods due to heavy or long-lasting precipitation, rapid
29 snowmelt, or catastrophic failure of glacial lake moraine dams. These climate conditions are projected to
30 increase with global warming.
31
32 Risks to cultural uses of water can become severe if there are permanent loss of aspects of communities'
33 cultures due to changes in water, including loss of areas of ice or snow with spiritual meanings, loss of
34 culturally-important places of access to such places, and loss of culturally-important subsistence practices
35 including by indigenous people (Chapter 4). This includes mountain regions where changes in the
36 cryosphere are having profound impacts (CCP5). In these cases, severe outcomes would be defined locally
37 rather than globally. Communities that lost a dominant environmental characteristic deeply associated with
38 its cultural identity would be considered to be severely impacted. For example, due to the central role that
39 travel on sea ice plays in the life of Inuit communities, providing freedom and mental wellbeing, loss of sea
40 ice can be argued to represent environmental dispossession of these communities (Durkalec et al., 2015).
41 Traditional ways of life are therefore threatened and resulting changes would be transformative rather than
42 adaptive. Similarly, changes in streamflow affecting the availability of species for traditional hunting can
43 also negatively impact indigenous communities (Norton-Smith et al.). Such changes are already being seen
44 at current levels of warming, but studies remain somewhat limited in number, so this assessment is assigned
45 medium confidence due to medium evidence and medium agreement. WG1 conclude that it is virtually
46 certain that further warming will lead to further reductions in Northern Hemisphere snow cover and mass
47 loss in individual glacier regions is projected to be between approximately 30% and 100% by 2100 under
48 high-warming scenarios (Chapter 4). Streamflows are projected to change in most major river basins
49 worldwide by several tens of percent at 4°C global warming (Chapter 4).
50
51 There is strong potential for increases in water scarcity, flooding, loss of snow and ice and changes in water
52 bodies to lead to severe outcomes such as deaths from water-related diseases, drowning and starvation, long-
53 term health impacts arising from malnutrition and diseases, loss of property, loss of existence or access to
54 places of cultural significance, loss of livelihoods and loss of aspects of culture especially for indigenous
55 people with traditional lifestyles. The numbers of people affected are projected to range from hundreds of
56 millions to several billion, depending on the level of global warming and socio-economic futures. A key
57 aspect of the risk is the high uncertainty in future regional precipitation changes in many regions of high
Do Not Cite, Quote or Distribute 16-71 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 vulnerability, including the potential for large and highly-impactful changes, for which it may not be
2 possible to provide adaptation measures before they become needed, leading to a high likelihood of
3 adaptation deficits.
4
5 16.5.2.3.8 Risks to peace and to human mobility (RKR-H)
6 This RKR includes risks to peace within and among societies from armed conflict as well as risks to human
7 mobility, epitomized by involuntary migration and displacement within and across state borders and
8 involuntary immobility. Breakdown of peace and the inability of people to choose to move or stay challenge
9 core elements of human security (Adger et al., 2014). Risks to peace also inform the agency and viability of
10 mobility decisions. However, evidence does not indicate that human mobility constitutes a general risk to
11 peace.
12
13 Breakdown of peace, materialized as overt or covert violence across social and spatial scales, constitutes a
14 key risk because of its potential to cause widespread loss of life, livelihood, and wellbeing. Such impacts are
15 considered severe if they result in at least 1,000 excess battle-related deaths in a country in a year. This
16 threshold is consistent with the conventional definition of war (Pettersson and Öberg, 2020). However,
17 because armed conflict routinely causes significant material destruction, triggers mass displacement,
18 threatens health and food security, and undermines economic activity and living standards (Baumann and
19 Kuemmerle, 2016; FAO et al., 2017; de Waal, 2018), risks to peace can be considered severe also when
20 conflict has cascading effects on other aspects of wellbeing and amplifies vulnerability to other RKRs.
21 Beyond the magnitude of such impacts, the rapidity with which armed conflict can escalate and the
22 challenges of ending violence once it has broken out imply potentially very limited time and ability to
23 respond for populations at risk.
24
25 Mobility is a universal strategy for pursuing wellbeing and managing household risks (Section 7.2.6; Cross-
26 Chapter Box MIGRATE in Chapter 7,UN, 2018) and, where it occurs in a safe and orderly fashion, can
27 reduce social inequality and facilitate sustainable development (Franco Gavonel et al., 2021). Involuntary
28 mobility constitutes a key risk because it implies reduced human agency with high potential for significant
29 economic losses and non-material costs, an unequal gender burden, and amplified vulnerability to other
30 RKRs (Schwerdtle et al., 2018; Adger et al., 2020; Maharjan et al., 2020; Piggott-McKellar et al., 2020).
31 Climate change also may erode or overwhelm human capacity to use mobility as a coping strategy,
32 producing involuntarily immobile populations (Adams, 2016). A severe impact is when a large share of an
33 affected population is forcibly displaced or prevented from moving, relative to normal mobility patterns, at
34 local to global scale. However, because mobility may be a favourable mechanism for reducing risk or an
35 adverse outcome of risk, depending on the circumstances under which it occurs, it is not possible to specify a
36 simple quantitative threshold for when impacts become severe.
37
38 Complex causal pathways and lack of long-term projection studies presently prevent making confident
39 quantitative judgments about how risks to peace and human mobility will materialize in response to specific
40 warming levels, development pathways, and adaptation scenarios. Literature concludes with medium
41 confidence that risks to peace will increase with warming, with the largest impacts expected in weather-
42 sensitive communities with low resilience to climate extremes and high prevalence of underlying risk factors
43 (Theisen, 2017; Busby, 2018; Koubi, 2019; von Uexkull and Buhaug, 2021). However, climate-driven
44 impacts on societies will depend critically on future political and socioeconomic development trajectories
45 (limited evidence, high agreement), suggesting that risks due to climate change are relevant primarily for
46 highly vulnerable populations and for pessimistic development scenarios. Overall risks to peace may decline
47 despite warming if non-climatic determinants are reduced sufficiently in the future.
48
49 Regular human mobility will continue regardless of climate change but mobility-related risks will increase
50 with warming, notably in densely populated hazard-prone regions, in small islands and low-lying coastal
51 zones, and among populations with limited coping capacity (RKR-A; CCP2 2.2.2; Chapter 7) (high
52 confidence). Such risks can become severe even with limited levels of warming for populations with low
53 adaptive capacity and whose settlements and livelihoods are critically sensitive to environmental conditions
54 (medium evidence, high agreement). Likewise, risk of involuntary immobility could become severe for
55 highly vulnerable populations with limited resources, even with moderate levels of warming (limited
56 evidence, high agreement). Critically, population growth and shifting exposure will interact with warming to
57 shape these risks (Davis et al., 2018; Hauer et al., 2020; Robinson, 2020a). Although climate-driven human
Do Not Cite, Quote or Distribute 16-72 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 mobility generally does not increase risks to peace (medium confidence), armed conflict is a major driver of
2 forced displacement (high confidence).
3
4 Expert elicitation estimates that 4°C warming above pre-industrial levels will have severe and widespread
5 effects on armed conflict with 26% probability, assuming no change from present levels in non-climatic
6 drivers (Mach et al., 2019). That judgment refers to impacts that exceed the threshold for severity considered
7 here, suggesting that global warming of 4°C would produce severe risks to peace under present societal
8 conditions (low confidence). Future risks to peace will remain strongly influenced by socioeconomic
9 development (Hegre et al., 2016). A study of Sub-Saharan Africa that accounts for both temperature and
10 socioeconomic changes, 201565, concludes that determinants other than rising temperatures, notably
11 quality of governance, will remain most influential in shaping overall levels of violence even in the high-
12 warming RCP8.5 scenario (Witmer et al., 2017).
13
14 A larger empirical literature offers indirect evidence that climate change may produce severe risks to peace
15 within this century by demonstrating how climate variability and extremes affect contemporary conflict
16 dynamics, especially in contexts marked by low economic development, high economic dependence on
17 climate-sensitive activities, high or increasing social marginalization, and fragile governance (medium
18 confidence) (Chapter 7.2.7; Chapter 16.2, Schleussner et al., 2016a; Von Uexkull et al., 2016; Busby, 2018;
19 Harari and Ferrara, 2018; Ide et al., 2020; Scartozzi, 2020).
20
21 Climatic risks interact with economic, political, and social drivers to create risks to human mobility both
22 directly (through the threat of physical harm and destruction of property and infrastructure) and indirectly
23 (via adverse impacts on livelihood and wellbeing). Extreme weather events are leading causes of forced
24 displacement (Cross-Chapter Box MIGRATE in Chapter 7, IDMC, 2020). Projected increases in the
25 frequency and severity of extreme events (AR6 WGI Chapter 12, Ranasinghe et al., 2021) in combination
26 with future population growth in hazard-prone regions (e.g., Merkens et al., 2016) suggest that risks to
27 mobility will increase in response to future global warming (Robalino et al., 2015; Davis et al., 2018; Rigaud
28 et al., 2018). For example, moving from RCP2.6 to RCP8.5 (entailing 0.5°C additional global warming by
29 2050) is projected to increase internal migration by 2050 from 51 [31-72] million to 118 [92-143] million
30 people across South Asia, Latin America, and Africa (Rigaud et al., 2018), although those estimates are
31 principally comprised of migrants, whose decisions are also informed by non-climatic drivers, rather than
32 involuntarily displaced people. Global levels of flood displacement are estimated to increase by 50% with
33 each 1°C warming (Kam et al., 2021). Should future warming reduce adaptation options for vulnerable
34 populations (Chapter 16.4), a consequence may be higher levels of involuntary migration and immobility
35 (Grecequet et al., 2017; Otto et al., 2017). There is little evidence that climate-driven mobility negatively
36 affects peace (Brzoska and Fröhlich, 2016; Burrows and Kinney, 2016; Freeman, 2017; Petrova, 2021).
37
38 There is high agreement that even moderate levels of future SLR will severely amplify involuntary migration
39 and displacement in small islands and densely populated low-lying coastal areas in the absence of
40 appropriate adaptive responses (high confidence) (Hauer, 2017; IPCC, 2019b; Hauer et al., 2020; McMichael
41 et al., 2020, Section 15.3.4; Section 16.4). In some contexts climate change also may accelerate migration
42 toward high-exposure coastal areas (Bell et al., 2021). Under a high emissions RCP8.5 scenario (global
43 median 0.7m SLR by 2100), the number of people exposed to annual coastal flooding may more than double
44 by 2100 compared to present numbers (Kulp and Strauss, 2019). In USA alone, SLR of 0.9m could
45 potentially put 4.2 million people at risk of inundation by the end of this century (Hauer, 2017). However,
46 numbers of people exposed to SLR does not evenly translate to forcibly displaced populations (Hauer et al.,
47 2020). Ascertaining how many people will move forcibly or as adaptive response to SLR is inherently
48 challenging because of the complex and highly individual nature of migration decisions (Black et al., 2013;
49 Boas et al., 2019; Piguet, 2019; Bell et al., 2021). Implications of climate change for risks to human mobility
50 across borders are even harder to quantify and highly uncertain, due to unknown developments in legal and
51 political conditions that govern international migration (McLeman, 2019; Wrathall et al., 2019).
52
53 16.5.2.4 Synthesis of the Assessment of Representative Key Risks
54
55 Figure 16.10 provides a synthesis of the RKRs and the conditions that lead to severe risks over the course of
56 the 21st century, as assessed in Sections 16.5.2.3.1 to 16.5.2.3.8 (see Supplementary Table SM16.12 for
57 further description). It identifies sets of conditions -- defined by levels of warming, exposure/vulnerability,
Do Not Cite, Quote or Distribute 16-73 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 and adaptation -- that would produce severe risk with a particular level of confidence. The risks are of two
2 scopes: broadly applicable, meaning that the risks described by a particular KR or RKR would be severe
3 pervasively and even globally; and specific, meaning that these risks would apply to particular areas, sectors,
4 or groups of people.
5
6
7
8 Figure 16.10: Synthesis of the severity conditions for Representative Key Risks by the end of this century. The figure
9 does not aim to describe severity conditions exhaustively for each RKR, but rather to illustrate the risks highlighted in
10 this report (Sections 16.5.2.3.1 to 16.5.2.3.8). Colored circles represent the levels of warming (climate),
11 exposure/vulnerability, and adaptation that would lead to severe risks for particular key risks and RKRs. Each set of
12 three circles represents a combination of conditions that would lead to severe risk with a particular level of confidence,
13 indicated by the number of black dots to the right of the set, and for a particular scope, indicated by the number of stars
14 to the left of the set. The two scopes are `broadly applicable', meaning applicable pervasively and even globally, and
15 `specific', meaning applicable to particular areas, sectors, or groups of people. Details of confidence levels and scopes
16 can be found in Section 16.5.2.3. In terms of severity condition levels (see Section 16.5.2.3), for warming levels
17 (colored circles labeled `C' in the figure), High refers to climate outcomes consistent with RCP8.5 or higher, Low refers
18 to climate outcomes consistent with RCP2.6 or lower, and Medium refers to intermediary climate scenarios. Exposure-
19 Vulnerability levels are determined by the RKR teams relative to the range of future conditions considered in the
20 literature. For Adaptation, High refers to near maximum potential and Low refers to the continuation of today's trends.
Do Not Cite, Quote or Distribute 16-74 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 Despite being intertwined in reality, Exposure-Vulnerability and Adaptation conditions are distinguished to help
2 understand their respective contributions to risk severity.
3
4
5 Five main messages arise from this synthesis:
6
7 Severe risk is rarely driven by a single determinant (warming, exposure/vulnerability, adaptation), but rather
8 by a combination of conditions that jointly produce the level of pervasiveness of consequences,
9 irreversibility, thresholds, cascading effects, likelihood of consequences, temporal characteristics of risk and
10 the systems' ability to respond (medium to high confidence). In other words, climate risk is not a matter of
11 changing climatic impact drivers (CIDs) only, but of the confrontation between changing CIDs and changing
12 socio-ecological conditions.
13
14 In most of the RKRs, severe risk for broadly applicable situations requires high levels of warming or
15 exposure/vulnerability, or low adaptation. In many cases, it is associated with several of these conditions
16 occurring simultaneously (e.g., high warming and high vulnerability). Examples include low-lying coastal
17 areas (RKR-A; medium confidence), loss of livelihoods (RKR-D; medium confidence) or armed conflicts
18 (RKR-H; low confidence).
19
20 High warming and exposure/vulnerability combined with low adaptation is however not necessarily required
21 to lead to severe risk, and various other sets of conditions can lead to such an outcome. For example:
22 Without high levels of warming -- This is especially the case for terrestrial and marine ecosystems (RKR-B)
23 and water security (RKR-G) for which even medium to low levels of warming will generate severe risk,
24 depending on the processes considered (e.g., mass population-level mortality and ecological disruption for
25 ecosystems). This is also the case when more specific situations are considered, for example in the case of
26 (in)voluntary mobility of vulnerable populations with limited resources (RKR-H), and for some critical
27 infrastructure in already highly exposed and vulnerable contexts (RKR-C).
28 With high levels of adaptation -- High levels of adaptation will not necessarily avoid severe risk, as is
29 illustrated by the cases of coral-dependent and arctic coastal communities (RKR-A), some terrestrial and
30 marine ecosystems (RKR-B), and water scarcity and the cultural uses of water (RKR-G).
31
32 All RKR assessments indicate that risks are higher in high vulnerability development pathways, and in some
33 cases high vulnerability can occur in high income societies. Examples include the possibility of increasing
34 coastal settlement and the location of critical infrastructure in highly exposed locations (RKR-A, RKR-C)
35 including to floods (RKR-G) and risks to terrestrial and marine ecosystems (RKR-B). The assessment
36 therefore show that depending on socioeconomic trends especially in terms of equity, social justice and
37 income sustainability, as well as on the ability to shift towards more climate resilient economic and
38 settlement systems (e.g., at the coast), higher income societies also are at serious risk of being substantially
39 affected in the decades-to-century to come.
40
41 In terms of the time frames, most of the RKRs conclude that severe risks to many dimensions (ecosystems,
42 health, etc.) are expected to occur by the end of the 21st century and across the globe. Some RKRs however
43 highlight that severe risk could occur far earlier, e.g. as soon as a warming level of 1.5°C or 2°C is reached,
44 which means potentially well before mid-century (IPCC, 2021a). In some cases, risks are already considered
45 severe, for example after major climatic events such as tropical storms (RKR-A).
46
47 16.5.3 Variation of Key Risks Across Levels of Global Warming, Exposure and Vulnerability, and
48 Adaptation
49
50 This section builds on Sections 16.5.1 and 16.5.2 as well as on additional literature to illustrate how
51 consequences associated with KRs and RKRs are projected to vary with three types of determinants: global
52 average warming level, as a proxy for associated changes in climate hazards (climatic impact-drivers, CIDs,
53 Ranasinghe et al., 2021); socio-economic development pathway, as a means of capturing alternative future
54 exposure and vulnerability conditions; and level of adaptation to reflect the extent to which successful
55 adaptation is implemented. While these three dimensions are partly intertwined e.g., warming and
56 adaptation scenarios are constrained by development pathways (Chapter 18) this section assesses the
57 influence of each dimension separately (Section 16.5.3.2 to Section 16.5.3.4) to highlight how sensitivity
Do Not Cite, Quote or Distribute 16-75 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 varies across these dimensions for different KRs and RKRs. We then bring the dimensions together in an
2 illustrative example (large deltas; Section 16.5.3.5).
3
4 16.5.3.1 Warming Level, Including Risks Avoided by Mitigation
5
6 Studies illustrating sensitivity to warming level typically do so by contrasting projected impacts for the same
7 socioeconomic conditions but different climate pathways or temperature levels, often based on
8 Representative Concentration Pathways (RCPs) (van Vuuren and Carter, 2014). We refer to future climate
9 conditions either based on their global average warming level or as a `high warming' scenario (based on
10 RCP8.5), medium warming (RCP4.5 or RCP6.0), or low warming (RCP2.6 or 1.5ºC scenarios). Because
11 some of these scenarios assume no or minimal mitigation (RCP8.5, RCP6.0) while others do (RCP4.5,
12 RCP2.6), differences in outcomes between them reflect risks avoided by mitigation (assuming consistent
13 socioeconomic assumptions).
14
15 Some ecological risks (Chapter 2) are particularly sensitive to warming. For example, warm-water coral
16 reefs are already experiencing High risk levels and are expected to face Very High risks under 1.5ºC of
17 global warming (Hoegh-Guldberg et al., 2018a; Bindoff et al., 2019). Some societal risks, such as human
18 mortality due to extreme heat, also are sensitive to warming. A medium warming scenario (relative to high
19 warming) reduces projected global average mortality due to heat from seven deaths per 10,000 people per
20 year (7/10,000/yr) by 2100 to ~1/10,000/yr, assuming high vulnerability societal conditions (Carleton et al.,
21 2020). At the national level, without considering adaptation, reductions in a broader measure of mortality are
22 projected across a range of countries including Colombia, the Philippines, and several in Europe (Guo et al.,
23 2018), and exposure of the US population to high mortality heatwaves is reduced by nearly half (Anderson et
24 al., 2018a). Without considering changes in exposure or vulnerability, warming of 1.5-2ºC (compared to 4-
25 5ºC) reduces global mortality impacts from an increase of 2.1-13.0% to 0.1-2.2% (Gasparrini et al., 2017;
26 Vicedo-Cabrera et al., 2018a) and impacts in China from up to 4/10,000/yr (Weinberger et al., 2017) to 0.3-
27 0.5/10,000/yr (Wang and Hijmans, 2019).
28
29 A low warming scenario (relative to high warming) reduces aggregate economic impacts from around 7% of
30 global GDP to less than 1% (Takakura et al., 2019), and changes impacts on the number of people suffering
31 from hunger from an increase (by 7-55 million) to a decrease (by up to 6 million) (Janssens et al., 2020).
32 Low versus high warming also reduces the coastal population at risk of flooding due to SLR from tripling by
33 2100 (relative to today) to doubling (Kulp and Strauss, 2019, Section 16.5.2.3.2). The SROCC estimates that
34 SLR risks are reduced from Moderate-to-High to Moderate for large tropical agricultural deltas and resource-
35 rich megacities, and from High and Very high to Moderate-to-High for Arctic human communities and
36 Urban atoll islands, respectively (Oppenheimer et al., 2019).
37
38 Higher levels of warming are projected to also generate higher income inequality between countries (e.g.,
39 Pretis et al., 2018; Takakura et al., 2019) as well as within them (Hallegatte et al., 2016) even though other
40 drivers will be more important (Section 16.5.2.3.5). Similarly, climate and weather events are expected to
41 play an increasing role in shaping risks to peace (medium agreement, low evidence) and migration (high
42 agreement, medium evidence) in the future, but uncertainty is high due to complex causal pathways and non-
43 climate factors likely dominate outcomes (Section 16.5.2.3.9). There is high agreement that future SLR will
44 amplify levels of forced migration from small islands and low-lying coastal areas in the absence of
45 appropriate adaptive responses (Oppenheimer et al., 2019).
46
47 A synthesis of risk assessments in the recent IPCC Special Reports (Magnan et al., 2021) concludes that an
48 integrated measure of today's global climate risk level will increase by the end of this century by two- to
49 four-fold under a low and high warming, respectively (based on aggregated scores developed in the study).
50 An additional comparison of risk levels under +1.5 °C and +2 °C suggests that every additional 0.5 °C of
51 global warming will increase the risk level by about a third.
52
53 16.5.3.2 Exposure and Vulnerability Trends
54
55 Development pathways describe plausible alternative futures of societal change and are critical to future
56 risks because they affect outcomes of concern both through non-climate and climate-related channels (very
57 high confidence).
Do Not Cite, Quote or Distribute 16-76 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1
2 Studies illustrating sensitivity to development pathways typically do so by contrasting projected impacts for
3 the same climate pathway or temperature level but different levels of socioeconomic exposure and
4 vulnerability, for example based on Shared Socioeconomic Pathways (SSPs) (O'Neill et al., 2014; Van
5 Vuuren et al., 2014). Or, they infer sensitivity to future development pathways based on differences in
6 impacts across current populations with different levels of exposure or vulnerability. We refer to future
7 conditions based on SSPs 1 or 5 as `low exposure' or `low vulnerability' conditions, and those based on
8 SSPs 3 or 4 as `high exposure' or `high vulnerability' conditions (O'Neill et al., 2014; van Vuuren and
9 Carter, 2014).
10
11 A wide range of climate change impacts depend strongly on development pathway (high confidence), A low
12 (relative to high) exposure future, determined by limited population growth and urbanization, results in about
13 30% fewer people exposed to extreme heat globally (Jones et al., 2018b) and about 50% fewer in Africa
14 (Rohat et al., 2019a), similar to the effect of a medium vs. high level of global warming. Low exposure
15 conditions also reduce the fraction of the population in Europe at very high risk of heat stress from 39% to
16 11% (Rohat et al., 2019b). Demographic differences lead to a reduction in the global population exposed to
17 mosquitos acting as viral disease vectors by more than half (Monaghan et al., 2018) and exposure to wildfire
18 risk by nearly half (Knorr et al., 2016).
19
20 Studies are increasingly going beyond exposure to incorporate future vulnerability, finding that it is often the
21 dominant determinant of risk (high confidence). A low (relative to high) vulnerability future reduces the risk
22 to global poverty by an order of magnitude, robustly across approaches that account for macroeconomic
23 growth, structural change in the economy, inequality, and access to infrastructure services (Hallegatte and
24 Rozenberg, 2017), or for the exposure of vulnerable populations to multi-sector climate-related risks (Byers
25 et al., 2018). A low (relative to high) vulnerability future also reduces the global mean number of
26 temperature-attributable deaths in 2080-2095 due to enteric infections by an order of magnitude (from
27 >80,000 to <7000; (Chua et al., 2021)). Low future socioeconomic vulnerability to flooding reduces global
28 fatalities and economic losses by 69-96% (Jongman et al., 2015). Low vulnerability as measured by
29 indicators including per capita GDP, education, governance, water demand, and storage potential reduces
30 water insecurity by a factor of three (Koutroulis et al., 2019). A scenario with reduced barriers to trade
31 reduces the number of people at risk of hunger due to climate change by 64% (Janssens et al., 2020).
32 Structural transformation of the economy (shift of the workforce from highly exposed sectors such as
33 agriculture and fishing to less exposed sectors such as services) lowers GDP impact projections by 25-30%
34 in today's developing countries by 2100 (Acevedo et al., 2017).
35
36 The IPCC SRCCL supports the importance of societal conditions to climate-related risk (Hurlbert et al.,
37 2019), concluding that risks of water scarcity in drylands (i.e., desertification), land degradation and food
38 insecurity are close to High3 beginning at 1.5ºC under high vulnerability conditions (SSP3), but remain close
39 to Moderate up to slightly above 2ºC for low vulnerability conditions (SSP1). Specifically, risk of water
40 scarcity in drylands (i.e., desertification) at 1.5ºC warming is reduced in low vulnerability (relative to high
41 vulnerability) conditions from High to Medium. Similarly, under a 2ºC warming, risk is reduced from High
42 to Moderate for food security and High to Moderate-to-High for land degradation.
43
44 While climate change will increase risk to society and ecosystems, future exposure and vulnerability
45 conditions will also greatly impact outcomes of concern directly. Global economic damages to coastal assets
46 from tropical cyclones are projected to increase by more than 300% due to coastal development alone, a
47 much larger effect than projected climate change impacts through 2100 even in RCP8.5 (Gettelman et al.,
48 2018). Similarly, global crop prices are more than three times more sensitive to alternative assumptions
49 about changes in production technologies and demand than to alternative climate outcomes (Ren et al.,
50 2016). Future water scarcity is driven mainly by both demographic change and socioeconomic changes
51 affecting water demand and management. A measure of between-country inequality (Gini coefficient) would
52 decline by more than 50% this century in low vulnerability conditions, but would double in a high
53 vulnerability future (Crespo Cuaresma, 2017), outweighing the effect of climate (Taconet et al., 2020).
3 The IPCC distinguishes between four qualitative risk levels, from Undetectable (risks that are undetected), to Moderate (detectable
with at least medium confidence), High (significant and widespread) and Very high (very high probability of severe risks and
significant irreversibility or persistence of impacts).
Do Not Cite, Quote or Distribute 16-77 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 Similarly, the global prevalence of armed conflict will roughly double this century in a high vulnerability
2 future, whereas it will drop by half in a low vulnerability future (Hegre et al., 2016). In Sub-Saharan Africa,
3 assumptions about governance and political rights are estimated to be far more important to the future risk of
4 violent conflict than climate change (Witmer et al., 2017).
5
6 16.5.3.3 Climate Adaptation Scenarios
7
8 One approach to understand adaptation benefits for risk reduction is to contrast projected impacts for the
9 same climate and development conditions but different levels of adaptation. For example, global-scale
10 coastal protection studies considering both RCPs and SSPs suggest that under a given RCP, the total flooded
11 area may be reduced by 40% by using 1-m height dykes, compared to a no-adaptation baseline (Tamura et
12 al., 2019). The global cost of SLR over the 21st century can be lowered by factor of two to four if local cost-
13 benefit decisions consider migration an adaptation option, in addition to hard protection (Lincke and Hinkel,
14 2021). Under a low warming scenario, it is estimated that adaptation (i.e. changes in crop variety and
15 planting dates) could reduce the total number of people at risk of hunger globally by about 4%, and by about
16 10% in a high warming scenario Hasegawa et al. (2014). Impacts on heat-related mortality would be cut
17 from 10 to 7 deaths per 10,000 people per year in 2100 by adaptation actions beyond those assumed to be
18 driven by income growth (Carleton et al., 2020). In a regional example, proactive adaptation efforts on
19 infrastructure (especially roads, runways, buildings, and airports) in Alaska, USA, could reduce damage-
20 related expenditure by 45% under medium or high warming (Melvin et al., 2017).
21
22 Another approach infers the potential future effectiveness of adaptation based on current sensitivity of
23 impacts to interventions. For example, the future disease burden of malaria is likely to be highly dependent
24 on the future development of health services, deployment of malaria programs and adaptation. Investments
25 in water and sanitation infrastructure are also recognized to have the potential to reduce severe risks of
26 waterborne disease, although these improvements likely need to provide transformative change (Cumming et
27 al., 2019). The potential for severe risks may also be substantially reduced through the development of
28 vaccines for specific enteric diseases (Riddle et al., 2018), although most current vaccines target viral
29 pathogens, incidence for which tends to be inversely correlated with ambient temperature (Carlton et al.,
30 2016). In addition, international migration as well as forced movement of people across borders will be
31 influenced by developments in legal and political conditions (McLeman, 2019; Wrathall et al., 2019), but the
32 fact that these developments are unknown strongly limits any forecasts on the magnitude of adaptation
33 benefits (Section 16.5.2.3.9).
34
35 Last, there is growing concern that even ambitious adaptation efforts will not eliminate residual risks from
36 climate change (Section 16.4.2). A synthesis of risk assessments in the recent IPCC Special Reports
37 (Magnan et al., 2021) concludes that high societal adaptation is expected to reduce the aggregated score the
38 proxy used in the study of global risk from anthropogenic climate change by about 40% under all RCPs by
39 the end of the century, compared to risk levels projected without adaptation. It however also shows that even
40 for the lowest warming scenario a residual risk one-third greater than today's risk level would still remain
41 (with a doubling of today's aggregated score under the high emission scenario).
42
43 16.5.3.4 Illustration: Risk and Adaptation Pathways in Densely Populated and Agricultural Deltas
44
45 Large deltas, which are very dynamic risk hotspots of global importance and interest (Wigginton, 2015; Hill
46 et al., 2020; Nicholls et al., 2020), serve well to illustrate how risk pathways develop over time, determined
47 by climatic as well as non-climatic risk drivers as well as by adaptation. Deltas occupy less than 0.5% of the
48 global land area but host over 5% of the global population (Dunn et al., 2019) and contribute major fractions
49 of food production in many world regions (Kuenzer et al., 2020). Future risk in these areas is heavily driven
50 by climate change but also greatly depends on past, current and future socio-economic changes which
51 influence future trends in exposure, vulnerability and adaptive capacity of natural and human systems (high
52 confidence) (Oppenheimer et al., 2019). From a risk perspective, trends over the past decades have been
53 unfavourable for many deltas, as most of them have experienced a simultaneous intensification of hazards,
54 rise in exposure and stagnation or only limited reduction in vulnerability, particularly in low income
55 countries (high confidence) (Day et al., 2016; Tessler et al., 2016; Loucks, 2019; Oppenheimer et al., 2019;
56 Hill et al., 2020).
57
Do Not Cite, Quote or Distribute 16-78 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 16.5.3.4.1 Hazard trends in deltas
2 Deltas face multiple interacting hazards, many of which over the past decades have been intensified by local
3 and regional anthropogenic developments (e.g., the construction of dams, groundwater extraction, or
4 agricultural irrigation practices) and most of which are expected to be exacerbated by climate change (high
5 confidence) (Giosan et al., 2014; Tessler et al., 2015; Tessler et al., 2016; Arto et al., 2019; Oppenheimer et
6 al., 2019). The most important hazards include sea level rise (SLR), inundation, salinity intrusion, cyclones,
7 storms and erosion, many of which occur in combination. The potential for flooding and inundation depends
8 on the relative sea level rise (RSLR) which results from global and regional SLR as well as local subsidence
9 within the deltas. Subsidence caused by natural and human drivers (mainly compaction and groundwater
10 extraction) is currently the most important cause for RSLR in many deltas and can exceed the rate of
11 climate-induced SLR by an order of magnitude (Oppenheimer et al., 2019). But in higher warming scenarios
12 the relative importance of climate-driven SLR is expected to increase over time (Oppenheimer et al., 2019).
13 In a global study covering 47 major deltas and assessing future trends of sediment delivery across four RCPs,
14 three SSPs (1,2,3) and a projection of future dam construction, Dunn et al. (2019) find most deltas (33 out of
15 the 47) will experience a mean decline of 38% in sediment flux by the end of the century when considering
16 the average of the scenarios. Nienhuis et al. (2020) find in a global assessment that some deltas have gained
17 land through increased sediment load (e.g., through deforestation), but recent land gains are unlikely to be
18 sustained if SLR continues to accelerate. According to the latest assessments, it is virtually certain that
19 global mean sea level will continue to rise over the 21st century, with sea level rise by 2100 likely to reach
20 0.28-0.55 m in a an SSP1-1.9 and 0.63-1.01 m in an SSP5-8.5 scenario relative to 1995-2014 (IPCC, 2021a).
21 The combined effects of local subsidence and GMSL rise result in a significant increase in the potential for
22 inundation of low-lying deltas across all RCPs, with some variation according to regional sea level change
23 rates, without significant further adaptation measures (very high confidence).
24
25 In terms of salt-water intrusion and salinization, global comparative studies are still lacking but the general
26 processes are well understood (e.g., White and Kaplan, 2017)) and research on individual deltas is on the
27 rise. In the Mekong Delta of Vietnam, one of the main rice producing deltas globally, salinity intrusion has
28 been observed to extend around 15 km inland during the rainy season and around 50 km during the dry
29 season (Gugliotta et al., 2017), resulting in rice yield losses of up to 4 tons per hectare per year (Khat et al.,
30 2018). SLR, along with the expansion of dams and dry season irrigation upstream, is expected to further
31 increase the salinity intrusion into the delta. This creates additional risk for food production as rice and other
32 crops might be pushed beyond their adaptation limits in terms of salt tolerance, potentially affecting many of
33 the 282,000 agriculture-based livelihoods in the Mekong Delta and increasing the pressure for cost-intensive
34 adaptation (Smajgl et al., 2015). Genua-Olmedo et al. (2016) find for the Ebro that in high scenario (RCP8.5,
35 and SLR of almost 1m by 2100), SLR-induced salinity intrusion will lead to almost a doubling of salinity
36 levels and a decrease of mean rice productivity by over 20% in a high SLR scenario with almost 1 meter of
37 SLR by the end of the century.
38
39 16.5.3.4.2 Exposure trends in deltas
40 Next to the trends in hazards, future exposure of and in deltas is shaped particularly by the increase of
41 population and infrastructure and the intensification of land use. Over the recent years, the population has
42 been rising in major deltas, roughly along with overall national population trends (Szabo et al., 2016). In
43 2017, 339 million people lived in deltas with a high exposure to flooding, cyclones and other coastal hazards
44 (Edmonds et al., 2020). Over 40% of the global population exposed to flooding from tropical cyclones lived
45 in deltas, more than 90% of which in developing countries and emerging economies (ibid.). Looking into the
46 future, population in low elevation coastal zones is expected to increase by 2050 across all SSPs with
47 diverging developments in the second half of the century, and at the end of the century will reach well over 1
48 billion people in SSP3 (Jones and O'Neill, 2016; Merkens et al., 2016). A major part of this population is
49 expected to reside in deltas with large cities or mega-urban agglomerations such as the Pearl River Delta,
50 China. One of the first studies using the SSP-RCP framework on the delta scale suggests a strong increase in
51 intensive agricultural land by the middle of the century in three SSPs (2, 3, 5) in the Volta Delta, Ghana,
52 whilst the Mahanadi, India, and the Ganges-Brahmaputra-Meghna do not show a significant further increase
53 (Kebede et al., 2018). Hence, the amount of population and infrastructure as well as agricultural land is
54 expected to rise further under certain SSPs, further increasing the exposure to future climate hazards.
55
Do Not Cite, Quote or Distribute 16-79 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 16.5.3.4.3 Vulnerability trends in deltas
2 Deltas are characterized by multifaceted vulnerabilities of their environment and human populations. Over
3 200 indicators are being used in the literature to characterize and analyse vulnerability in deltas, spanning
4 social, ecological and economic aspects (Sebesvari et al., 2016). However, only a few studies model or
5 dynamically assess trends in vulnerability, particularly for the future, at global scale, or take a comparative
6 approach. But overall, a global trend assessment suggests that social vulnerability to climate hazards has
7 been improving over the past years in all world regions hosting major deltas apart from Oceania, yet with
8 emerging economies and developing countries in Africa showing less improvement than the Americas, Asia
9 and Europe (Feldmeyer et al., 2017). An analysis of 48 major deltas finds that vulnerability therefore is a less
10 dominant source of future increase in risk than exposure (Haasnoot et al., 2012). However, case study
11 research from individual deltas suggests that delta populations, particularly those with agriculture-based
12 livelihoods, have seen more limited vulnerability reduction due in particular to the impacts of environmental
13 hazards, stress and disasters (high confidence). In the Mekong Delta, for instance, the strong economic
14 growth since the beginning of Vietnam's reform process has not led to a reduction of vulnerability across the
15 board for all socio-economic groups (Garschagen, 2015). Rather, issues such as widespread landlessness or
16 continued poverty have maintained and, in some respect, increased social vulnerability.
17
18 16.5.4 RKR Interactions
19
20 Multiple feedbacks between individual risks exist that have the potential to create cascades (WEF, 2018;
21 Weyer, 2019 p. 680; Simpson et al., 2021) and then to amplify systemic risks and impacts far beyond the
22 level of individual RKRs (medium confidence). Scientific research however remains limited on whether such
23 interactions would result in increasing or decreasing the initial impact(s), and hence risk severity across
24 systems. Given the scope of this chapter on increasing risk severity, here we focus on assessing RKR
25 interactions that lead to increasing risk. Drawing directly on RKR assessments (16.5.2.3.2 to 16.5.2.3.9), this
26 section cites those assessments rather than primary literature. The arrows in Figure 16.11 are derived from a
27 qualitative analysis by three authors of Chapter 16 of the material provided by chapters on KRs and RKR
28 assessments (Section 16.5.2.3), and do not result from any systematic and quantitative approach as done in
29 some recent studies (e.g., WEF, 2018; Yokohata et al., 2019).
30
31 Interactions at the RKR level (Figure 16.11, Panel A) Climate change will combine with pre-existing
32 socioeconomic and ecological conditions (grey blocks on the left hand-side of Panel A in Figure 16.10) to
33 generate direct and second-order effects (black plain arrows) both on the structure and/or functioning of
34 ecosystems (RKR-B) and on some natural processes such as the hydrologic cycle (RKR-G) for example.
35 This then translates into implications not only for biodiversity, but also for natural resources that support
36 livelihoods, which will in turn affect food security (especially food availability; RKR-F), water security
37 (especially access to adequate quantities of acceptable quality water; RKR-G) and the living standards of
38 already vulnerable groups and aggregate economic outputs at the global level (RKR-D). Climatic impact
39 drivers (CID; IPCC, 2021a) will also directly affect infrastructure that are critical to ensure some basic
40 conditions for economies to function (RKR-C), e.g., through transportation within and outside the country,
41 energy production and international trade. Such disturbances to socioecological systems and economies pose
42 climate-related risks to human health (RKR-E) as well as to peace and mobility (RKR-H). Indeed, while
43 health is concerned with direct influence of climate change, e.g., through hotter air temperatures impacting
44 morbidity and mortality or the spatial distribution of disease vectors such as mosquitos, it is also at risk of
45 being stressed by direct and secondary climate impacts on living standards, food security and water security
46 (RKR-D, RKR-F, RKR-G, respectively). Increased poverty, increased hunger and limited access to drinkable
47 water are well-known drivers of poor health conditions. The role of impact cascades is even more prominent
48 in the case of peace and mobility (RKR-H), even though the scientific literature does not conclude on any
49 clear and direct climate influence on armed conflict and human migration. Rather, climate-induced
50 degradation of natural resources that are vital for subsistence agriculture and fisheries, transformational and
51 long-term consequences on livelihoods (e.g., new risks, increasing precarious living conditions, gendered
52 inequity, etc.), as well as erosion of social capital due to exacerbated tension within and between
53 communities, are considered among the main drivers of armed conflicts and forced displacement, therefore
54 highlighting links with water security (RKR-G) and living standards (RKR-D), for example.
55
56 RKR assessments also suggest that some feedback effects are at work (arrows moving from the right to the
57 left in Panel A) that contribute to the potentially long-lasting effects of climate risks. RKR-H assessment for
Do Not Cite, Quote or Distribute 16-80 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 example states that there is robust evidence that major armed conflicts routinely trigger mass displacement,
2 threaten health and food security, and undermine economic activity and livelihoods, often with lasting
3 negative consequences for living standards and socioeconomic development, therefore linking back to risks
4 to living standards (RKR-D), human health (RKR-E) and food security (RKR-F).
5
6 Interactions at the KR level (Figure 16.11, Panel B) Panel B illustrates risk connections at the Key Risk
7 level (Section 16.5.2.1) and as described in RKR assessments (Section 16.5.2.3). To only take one example
8 here, risk to livelihoods and economies is influenced by the loss of ecosystem services (RKR-B) and the loss
9 or breakdown of critical infrastructures (RKR-C), as well as it influences risks to human lives and health
10 (RKR-E), food and water security (RKR-F, RKR-G), poverty (RKR-D) and peace and mobility (RKR-H).
11 As a third-order sequence, RKR assessments show that increased risk to peace and mobility affects lives and
12 health as well as food security, which in turn threaten livelihoods and economies.
13
14 The above suggests that some vicious cycle effects play a central role in explaining impact processes.
15 Cascading effects can indeed lead to cumulative risks that partly feed various drivers of the emergence of
16 severe risks (Section 16.5.1), such as the acceleration of ecosystem degradation, or the reaching of thresholds
17 and irreversible states in human systems at a decade-to-century time horizon (e.g., when permanent
18 inundation questions the habitability of some low-lying coasts; RKR-A). The extent and duration of risk
19 cascades are however expected to substantially vary depending on warming levels and development
20 pathways, both separately (Section 16.5.3) and when combined (Section 16.6.1 and 16.6.2) (Fig. 16.10).
21
22 In addition, RKR assessments converge to suggest that regions that are already experiencing climate change
23 impacts will experience severe impact cascades first (e.g., RKR-F), because they are in areas (i) that face
24 development constraints and associated challenges such as poverty, inequity and social discrimination for
25 example, and (ii) where climate change projections are the most intense for the next decades. That is
26 especially a concern for Africa (RKR-F, RKR-G), Asia and Latin America (Chapters 9, 10 and 12). RKR-E
27 concludes for example that the likelihood of severe risks to human health is especially high for highly
28 susceptible populations, particularly the poor and otherwise marginalized. RKR assessments however
29 emphasize that middle- and high-income regions are also to be considered at serious risk because climate
30 change is accelerating at the global level (IPCC, 2021a), and because critical dimensions are exposed to
31 severe risks such as major transportation (e.g., international airports) and energy (e.g., nuclear power plants)
32 infrastructure for instance (RKR-C), and because of the interconnectedness of economies.
33
34 Finally, all RKR assessments suggest that enhanced adaptation has the potential to contain such feedback
35 effects and cascading processes more broadly, and reduce the duration of the impacts on the system as a
36 whole. There are however knowledge gaps on such a potential, as well as on the nature of impact cascades
37 (positive, negative, neutral, mixed).
38
39
Do Not Cite, Quote or Distribute 16-81 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
Panel A - Interactions across the eight Representative Key Risk level
Climatic impact-drivers
(CID) *
Recent development Ecosystems Food Health
trends and ecological (& ecosystem security
conditions (exposure and Peace and
services) Water mobility
vulnerability conditions) security
Infrastructure
Population growth Living
Settlement trends Standards
Socioecon. inequalities
Resources use
Indigen./local knowlegde
Etc.
Panel B - Illustration of interactions at the Key Risk level
(e.g. from ecological risk to key dimensions for human societies)
Climatic impact-drivers N.B.: Trends in exposure and
(CID) * vulnerability conditions, as
represented in the grey box in
Panel A, are not represented as
such in Panel B, but contribute to
all risk considered in this figure
Species Loss of life- Loss/breakdown Water
extinction supporting of infratsructure- Food
& ecological ecosystem
disruption based service
delivery Well-being
Changes in
habitats and Livelihoods and Income
biodiversity economies inequality
(all latitudes, (supply chains,
land and ocean)
aggregate economic
outputs, etc.)
Other things that Lives and Peace from
societies value health armed
(intangible assets,
Poverty conflicts
landscapes, places,
identity, etc.)
Migration and forced
displacements
(within/across state borders)
* CIDs are physical climate system conditions (e.g., means, events, extremes) that affect an element of
society or ecosystems. Indiced changes are system-dependent and can be detrimental, beneficial, neutral,
or a mixture of each (see IPCC WG1 contribution to AR6, Summary for Policy Makers).
Risk cascades ** Representative key Risks
Across key risks
Climate-driven A (Low-lying coasts) E (Human health)
** As suggested across RKR assessments; B (Ecosystems) F (Food security)
illustrative rather than comprehensive, C (Infrastructure) G (Water security)
D (Living standards) H (Peace and mobility)
and qualitative rather than quantitative
1
2 Figure 16.11: Illustration of some connections across key risks. Panel A describes all the cross-RKR risk cascades that
3 are described in RKR assessments (Sections 16.5.2.3.2 to 16.5.2.3.9). Panel B builds on Section 16.5.2.2 and Table
4 SM16.4 to provide an illustration of such interactions at the Key Risk level, e.g. from ecological risk to key dimensions
Do Not Cite, Quote or Distribute 16-82 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 for human societies. The arrows are representative of interactions as qualitatively identified in this chapter; they do not
2 result from any quantitative modelling exercise.
3
4
5 [START CROSS-WORKING GROUP BOX SRM HERE]
6
7 Cross-Working Group Box SRM: Solar Radiation Modification
8
9 Authors: Christopher H. Trisos (South Africa), Oliver Geden (Germany), Sonia I. Seneviratne (Switzerland),
10 Masahiro Sugiyama (Japan), Maarten van Aalst (The Netherlands), Govindasamy Bala (India), Katharine J.
11 Mach (USA), Veronika Ginzburg (Russia), Heleen de Coninck (The Netherlands), Anthony Patt
12 (Switzerland)
13
14 Proposed Solar Radiation Modification Schemes
15
16 This cross-working group box assesses Solar Radiation Modification (SRM) proposals, their potential
17 contribution to reducing or increasing climate risk, as well as other risks they may pose (categorised as risks
18 from responses to climate change in the IPCC AR6 risk definition in 1.2.1.1), and related perception, ethics
19 and governance questions.
20
21 SRM refers to proposals to increase the reflection of shortwave radiation (sunlight) back to space to
22 counteract anthropogenic warming and some of its harmful impacts (de Coninck et al., 2018) (Cross-chapter
23 Box 10; WG1 Chapter 4 and Chapter 5). A number of SRM options have been proposed, including:
24 Stratospheric Aerosol Interventions (SAI), Marine Cloud Brightening (MCB), Ground-Based Albedo
25 Modifications (GBAM), and Ocean Albedo Change (OAC). Although not strictly a form of SRM, Cirrus
26 Cloud Thinning (CCT) has been proposed to cool the planet by increasing the escape of longwave thermal
27 radiation to space and is included here for consistency with previous assessments (de Coninck et al., 2018).
28 SAI is the most-researched proposal. Modeling studies show SRM could reduce surface temperatures and
29 potentially ameliorate some climate change risks (with more confidence for SAI than other options), but
30 SRM could also introduce a range of new risks.
31
32 There is high agreement in the literature that for addressing climate change risks SRM cannot be the main
33 policy response to climate change and is, at best, a supplement to achieving sustained net zero or net
34 negative CO2 emission levels globally (de Coninck et al., 2018; MacMartin et al., 2018; Buck et al., 2020;
35 National Academies of Sciences and Medicine, 2021b). SRM contrasts with climate change mitigation
36 activities, such as emission reductions and CDR, as it introduces a `mask' to the climate change problem by
37 altering the Earth's radiation budget, rather than attempting to address the root cause of the problem, which
38 is the increase in GHGs in the atmosphere. In addition, the effects of proposed SRM options would only last
39 as long as a deployment is maintained-- e.g. requiring ca. yearly injection of aerosols in the case of SAI as
40 the lifetime of aerosols in the stratosphere is 1-3 years (Niemeier et al., 2011) or continuous spraying of sea
41 salt in the case of MCB as the lifetime of sea salt aerosols in the atmosphere is only about 10 days--which
42 contrasts with the long lifetime of CO2 and its climate effects, with global warming resulting from CO2
43 emissions likely remaining at a similar level for a hundred years or more (MacDougall et al., 2020) and long-
44 term climate effects of emitted CO2 remaining for several hundreds to thousands of years (Solomon et al.,
45 2009).
46
47 Which scenarios?
48
49 The choice of SRM deployment scenarios and reference scenarios is crucial in assessment of SRM risks and
50 its effectiveness in attenuating climate change risks (Keith and MacMartin, 2015; Honegger et al., 2021).
51 Most climate model simulations have used scenarios with highly stylized large SRM forcing to fully
52 counteract large amounts of warming in order to enhance the signal-to-noise ratio of climate responses to
53 SRM (Kravitz et al., 2015; Sugiyama et al., 2018a; Tilmes et al., 2018; Krishna-Pillai et al., 2019).
54
55 The effects of SRM fundamentally depend on a variety of choices about deployment (Sugiyama et al.,
56 2018b), including: its position in the portfolio of human responses to climate change (e.g., the magnitude of
57 SRM used against the background radiative forcing), governance of research and potential deployment
Do Not Cite, Quote or Distribute 16-83 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 strategies, and technical details (latitude, materials, and season, among others, see WG1 Chapter 4.6.3.3).
2 The plausibility of many SRM scenarios is highly contested and not all scenarios are equally plausible
3 because of socio-political considerations (Talberg et al., 2018b), as with, for example, CDR (Fuss et al.,
4 2014; Fuss et al., 2018). Development of scenarios and their selection in assessments should reflect a diverse
5 set of societal values with public and stakeholder inputs (Sugiyama et al., 2018a; Low and Honegger, 2020),
6 as depending on the focus of a limited climate model simulation, SRM could look grossly risky or highly
7 beneficial(Pereira and al., 2021).
8
9 In the context of reaching the long-term global temperature goal of the Paris Agreement, there are different
10 hypothetical scenarios of SRM deployment: early, substantial mitigation with no SRM, more limited or
11 delayed mitigation with moderate SRM, unchecked emissions with total reliance on SRM, and regionally
12 heterogeneous SRM. Each scenario presents different levels and distributions of SRM benefits, side effects,
13 and risks. The more intense the SRM deployment, the larger is the likelihood for the risks of side effects and
14 environmental risks (e.g., Heutel et al., 2018). Regional disparities in climate hazards may result from both
15 regionally-deployed SRM options such as GBAM, and more globally uniform SRM such as SAI (Jones et
16 al., 2018a; Seneviratne et al., 2018b). There is an emerging literature on smaller forcings of SAI to reduce
17 global average warming, for instance, to hold global warming to 1.5°C or 2°C alongside ambitious
18 conventional mitigation (Jones et al., 2018a; MacMartin et al., 2018), or bring down temperature after an
19 overshoot (Tilmes et al., 2020). If emissions reductions and CDR are deemed insufficient, SRM may be seen
20 by some as the only option left to ensure the achievement of the Paris Agreement's temperature goal by
21 2100.
22
23
24 Table Cross-Working Group Box SRM.1: SRM options and their potential climate and non-climate impacts
25 Description, potential climate impacts, potential impacts on human and natural systems, and termination effects of a
26 number of SRM options: Stratospheric Aerosol Interventions (SAI), Marine Cloud Brightening (MCB), Ocean Albedo
27 Change (OAC), Ground-Based Albedo Modifications (GBAM), and Cirrus Cloud Thinning (CCT).
SRM option SAI MCB OAC GBAM CCT
Description Injection of Spraying sea Increase Whitening roofs, Seeding to
reflective salt or other surface albedo changes in land use promote
Potential climate aerosol particles particles in of the ocean management (e.g., nucleation of
impacts other directly into the marine clouds, (e.g., by no-till farming, cirrus clouds,
than reduced stratosphere or a making them creating bioengineering to reducing
warming gas which then more reflective microbubbles make crop leaves optical
converts to or placing more reflective), thickness and
aerosols that reflective foam desert albedo cloud lifetime
reflect sunlight on the surface) enhancement, to allow more
covering glaciers outgoing
Change Change in land- Change in with reflective longwave
precipitation and sea contrast in land-sea sheeting radiation to
runoff temperature and contrast in escape to
pattern; reduced precipitation, temperature Changes in regional space
temperature and regional and precipitation Changes in
precipitation precipitation and precipitation, pattern, regional temperature
extremes; runoff changes regional , extremes and and
precipitation precipitation regional circulation precipitation
reduction in and runoff pattern, altered
some monsoon changes. regional water
regions; cycle, increase
decrease in in sunlight
direct and reaching the
increase in surface
diffuse sunlight
at surface;
changes to
stratospheric
Do Not Cite, Quote or Distribute 16-84 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
dynamics and
chemistry;
potential
delay in ozone
hole recovery;
changes in
surface ozone
and UV
radiation
Potential impacts Changes in crop Changes in Unresearched Altered Altered
on human and yields, changes regional ocean photosynthesis, photosynthesis
natural systems in land and productivity, carbon uptake and and carbon
ocean ecosystem changes in crop side effects on uptake
productivity, yields, reduced biodiversity
acid rain (if heat stress for
using sulphate), corals, changes
reduced risk of in ecosystem
heat stress to productivity on
corals land, sea salt
deposition over
land
Termination Sudden and Sudden and Sudden and GBAM can be Sudden and
effects sustained sustained sustained maintained over sustained
termination termination termination several years termination
References (also would result in would result in would result in without major would result in
see main text of rapid warming, rapid warming, rapid warming. termination effects rapid
this box) and abrupt and abrupt Magnitude of because of its warming.
changes to water changes to water termination regional scale of Magnitude of
cycle. cycle. depends on the application. termination
Magnitude of Magnitude of degree of Magnitude of depends on the
termination termination warming termination depends degree of
depends on the depends on the offset. on the degree of warming
degree of degree of warming offset. offset.
warming offset. warming offset. Evans et al.
Tilmes et al. Latham et al. (2010) Crook Zhang et al. (2016); Storelvmo and
(2018) (2012) Ahlm et et al. (2015a) Field et al. (2018); Herger (2014)
Simpson et al. al. (2017) Stjern Seneviratne et al. Crook et al.
(2019) Visioni et et al. (2018) (2018a) Davin et al. (2015a)
al. (2017) (2014) Crook et al. Jackson et al.
(2015a) (2016)
Gasparini et
al. (2020)
Duan et al.
(2020)
1
2
3 SRM risks to human and natural systems and potential for risk reduction
4
5 Since AR5, hundreds of climate modelling studies have simulated effects of SRM on climate hazards
6 (Kravitz et al., 2015; Tilmes et al., 2018). Modelling studies have shown SRM has the potential to offset
7 some effects of increasing GHGs on global and regional climate, including the increase in frequency and
8 intensity of extremes of temperature and precipitation, melting of Arctic sea ice and mountain glaciers,
9 weakening of Atlantic meridional overturning circulation, changes in frequency and intensity of tropical
10 cyclones, and decrease in soil moisture (WG1, Chapter 4). However, while SRM may be effective in
11 alleviating anthropogenic climate warming either locally or globally, it would not maintain the climate in a
12 present-day state nor return the climate to a pre-industrial state (climate averaged over 1850-1900, See WG1
13 Chapter 1, Box 1.2) in all regions and in all seasons even when used to fully offset the global mean warming
14 (high confidence); WG1 Chapter 4}. This is because the climate forcing and response to SRM options are
15 different from the forcing and response to GHG increase. Because of these differences in climate forcing and
16 response patterns, the regional and seasonal climates of a world with a global mean warming of 1.5 or 2°C
Do Not Cite, Quote or Distribute 16-85 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 achieved via SRM would be different from a world with similar global mean warming but achieved through
2 mitigation MacMartin et al. JGR2019}. At the regional scale and seasonal timescale there could be
3 considerable residual climate change and/or overcompensating change (e.g., more cooling, wetting or drying
4 than just what's needed to offset warming, drying or wetting due to anthropogenic greenhouse gas
5 emissions), and there is low confidence in understanding of the climate response to SRM at the regional
6 scale (WG1, Chapter 4).
7
8 SAI implemented to partially offset warming (e.g., offsetting half of global warming) may have potential to
9 ameliorate hazards in multiple regions and reduce negative residual change, such as drying compared to
10 present-day climate, that are associated with fully offsetting global mean warming (Irvine and Keith, 2020),
11 but may also increase flood and drought risk in Europe compared to unmitigated warming (Jones et al.,
12 2021). Recent modelling studies suggest it is conceptually possible to meet multiple climate objectives
13 through optimally designed SRM strategies (WG1, Chapter 4). Nevertheless, large uncertainties still exist for
14 climate processes associated with SRM options (e.g. aerosol-cloud-radiation interaction) (WG1, Chapter 4)
15 (Kravitz and MacMartin, 2020).
16
17 Compared with climate hazards, many fewer studies have examined SRM risks--the potential adverse
18 consequences to people and ecosystems from the combination of climate hazards, exposure and
19 vulnerability--or the potential for SRM to reduce risk (Curry et al., 2014; Irvine et al., 2017). Risk analyses
20 have often used inputs from climate models forced with stylized representations of SRM, such as dimming
21 the sun. Fewer have used inputs from climate models that explicitly simulated injection of gases or aerosols
22 into the atmosphere, which include more complex cloud-radiative feedbacks. Most studies have used
23 scenarios where SAI is deployed to hold average global temperature constant despite high emissions.
24
25 There is low confidence and large uncertainty in projected impacts of SRM on crop yields due in part to a
26 limited number of studies. Because SRM would result in only a slight reduction in CO2 concentrations
27 relative to the emission scenario without SRM (Chapter 5, WG1), the CO2 fertilization effect on plant
28 productivity is nearly the same in emissions scenarios with and without SRM. Nevertheless, changes in
29 climate due to SRM are likely to have some impacts on crop yields. A single study indicates MCB may
30 reduce crop failure rates compared to climate change from a doubling of CO2 pre-industrial concentrations
31 (Parkes et al., 2015). Models suggest SAI cooling would reduce crop productivity at higher latitudes
32 compared to a scenario without SRM by reducing the growing season length, but benefit crop productivity in
33 lower latitudes by reducing heat stress (Pongratz et al., 2012; Xia et al., 2014; Zhan et al., 2019). Crop
34 productivity is also projected to be reduced where SAI reduces rainfall relative to the scenario without SRM,
35 including a case where reduced Asian summer monsoon rainfall causes a reduction in groundnut yields (Xia
36 et al., 2014; Yang et al., 2016). SAI will increase the fraction of diffuse sunlight, which is projected to
37 increase photosynthesis in forested canopy, but will reduce the direct and total available sunlight, which
38 tends to reduce photosynthesis. As total sunlight is reduced, there is a net reduction in crop photosynthesis
39 with the result that any benefits to crops from avoided heat stress may be offset by reduced photosynthesis,
40 as indicated by a single statistical modeling study (Proctor et al., 2018). SAI would reduce average surface
41 ozone concentration (Xia et al., 2017) mainly as a result of aerosol-induced reduction in stratospheric ozone
42 in polar regions, resulting in reduced downward transport of ozone to the troposphere (Pitari et al., 2014;
43 Tilmes et al., 2018). The reduction in stratospheric ozone also allows more UV radiation to reach the surface.
44 The reduction in surface ozone, together with an increase in surface UV radiation, would have important
45 implications for crop yields but there is low confidence in our understanding of the net impact.
46
47 Few studies have assessed potential SRM impacts on human health and wellbeing. SAI using sulfate aerosols
48 is projected to deplete the ozone layer, increasing mortality from skin cancer, and SAI could increase
49 particulate matter due to offsetting warming, reduced precipitation and deposition of SAI aerosols, which
50 would increase mortality, but SAI also reduces surface-level ozone exposure, which would reduce mortality
51 from air pollution, with net changes in mortality uncertain and depending on aerosol type and deployment
52 scenario (Effiong and Neitzel, 2016; Eastham et al., 2018; Dai et al., 2020). However, these effects may be
53 small compared to changes in risk from infectious disease (e.g., mosquito-borne illnesses) or food security
54 due to SRM influences on climate (Carlson et al., 2020). Using volcanic eruptions as a natural analog, a
55 sudden implementation of SAI that forced the ENSO system may increase risk of severe cholera outbreaks in
56 Bengal (Trisos et al., 2018; Pinke et al., 2019). Considering only mean annual temperature and precipitation,
57 SAI that stabilizes global temperature at its present-day level is projected to reduce income inequality
Do Not Cite, Quote or Distribute 16-86 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 between countries compared to the highest warming pathway (RCP8.5) (Harding et al., 2020). Some
2 integrated assessment model scenarios have included SAI (Arino et al., 2016; Emmerling and Tavoni, 2018;
3 Heutel et al., 2018; Helwegen et al., 2019; Rickels et al., 2020) showing the indirect costs and benefits to
4 welfare dominate, since the direct economic cost of SAI itself is expected to be relatively low (Moriyama et
5 al., 2017; Smith and Wagner, 2018). There is a general lack of research on the wide scope of potential risk or
6 risk reduction to human health, wellbeing and sustainable development from SRM and on their distribution
7 across countries and vulnerable groups (Carlson et al., 2020; Honegger et al., 2021).
8
9 SRM may also introduce novel risks for international collaboration and peace. Conflicting temperature
10 preferences between countries may lead to counter-geoengineering measures such as deliberate release of
11 warming agents or destruction of deployment equipment (Parker et al., 2018). Game-theoretic models and
12 laboratory experiments indicate a powerful actor or group with a higher preference for SRM may use SAI to
13 cool the planet beyond what is socially optimal, imposing welfare losses on others although this cooling does
14 not necessarily imply excluded countries would be worse off relative to a world of unmitigated warming
15 (Ricke et al., 2013; Weitzman, 2015; Abatayo et al., 2020). In this context counter-geoengineering may
16 promote international cooperation or lead to large welfare losses (Heyen et al., 2019; Abatayo et al., 2020).
17
18 Cooling caused by SRM would increase the global land and ocean CO2 sinks (medium confidence), but this
19 would not stop CO2 from increasing in the atmosphere or affect the resulting ocean acidification under
20 continued anthropogenic emissions (high confidence) (WG1 Chapter 5).
21
22 Few studies have assessed potential SRM impacts on ecosystems. SAI and MCB may reduce risk of coral
23 reef bleaching compared to global warming with no SAI (Latham et al., 2013; Kwiatkowski et al., 2015), but
24 risks to marine life from ocean acidification would remain, because SRM proposals do not reduce elevated
25 levels of anthropogenic atmospheric CO2 concentrations. MCB could cause changes in marine net primary
26 productivity by reducing light availability in deployment regions, with important fishing regions off the west
27 coast of South America showing both large increases and decreases in productivity (Partanen et al., 2016;
28 Keller, 2018).
29
30 There is large uncertainty in terrestrial ecosystem responses to SRM. By decoupling increases in atmospheric
31 greenhouse gas concentrations and temperature, SAI could generate substantial impacts on large-scale
32 biogeochemical cycles, with feedbacks to regional and global climate variability and change (Zarnetske et
33 al., 2021). Compared to a high CO2 world without SRM, global-scale SRM simulations indicate reducing
34 heat stress in low latitudes would increase plant productivity, but cooling would also slow down the process
35 of nitrogen mineralization which could decrease plant productivity (Glienke et al., 2015; Duan et al., 2020).
36 In high latitude and polar regions SRM may limit vegetation growth compared to a high CO2 world without
37 SRM, but net primary productivity may still be higher than pre-industrial climate (Glienke et al., 2015).
38 Tropical forests cycle more carbon and water than other terrestrial biomes but large areas of the tropics may
39 tip between savanna and tropical forest depending on rainfall and fire (Beer et al., 2010; Staver et al., 2011).
40 Thus, SAI-induced reductions in precipitation in Amazonia and central Africa are expected to change the
41 biogeography of tropical ecosystems in ways different both from present-day climate and global warming
42 without SAI (Simpson et al., 2019; Zarnetske et al., 2021). This would have potentially large consequences
43 for ecosystem services (Chapter 2 and Chapter 9). When designing and evaluating SAI scenarios, biome-
44 specific responses need to be considered if SAI approaches are to benefit rather than harm ecosystems.
45 Regional precipitation change and sea salt deposition over land from MCB may increase or decrease primary
46 productivity in tropical rainforests (Muri et al., 2015). SRM that fully offsets warming could reduce the
47 dispersal velocity required for species to track shifting temperature niches whereas partially offsetting
48 warming with SAI would not reduce this risk unless rates of warming were also reduced (Trisos et al., 2018;
49 Dagon and Schrag, 2019). SAI may reduce high fire risk weather in Australia, Europe and parts of the
50 Americas, compared to global warming without SAI (Burton et al., 2018). Yet SAI using sulfur injection
51 could shift the spatial distribution of acid-induced aluminum soil toxicity into relatively undisturbed
52 ecosystems in Europe and North America (Visioni et al., 2020). For the same amount of global mean
53 cooling, SAI, MCB, and CCT would have different effects on gross and net primary productivity because of
54 different spatial patterns of temperature, available sunlight, and hydrological cycle changes (Duan et al.,
55 2020). Large-scale modification of land surfaces for GBAM may have strong trade-offs with biodiversity
56 and other ecosystem services, including food security (Seneviratne et al., 2018a). Although existing studies
Do Not Cite, Quote or Distribute 16-87 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 indicate SRM will have widespread impacts on ecosystems, risks and potential for risk reduction for marine
2 and terrestrial ecosystems and biodiversity remain largely unknown.
3
4 A sudden and sustained termination of SRM in a high CO2 emissions scenario would cause rapid climate
5 change (high confidence; WG1 Chapter 4). More scenario analysis is needed on the potential likelihood of
6 sudden termination (Kosugi, 2013; Irvine and Keith, 2020). A gradual phase-out of SRM combined with
7 emission reduction and CDR could avoid these termination effects (medium confidence) (MacMartin et al.,
8 2014; Keith and MacMartin, 2015; Tilmes et al., 2016). Several studies find that large and extremely rapid
9 warming and abrupt changes to the water cycle would occur within a decade if a sudden termination of SAI
10 occurred (McCusker et al., 2014; Crook et al., 2015b). The size of this `termination shock' is proportional to
11 the amount of radiative forcing being masked by SAI. A sudden termination of SAI could place many
12 thousands of species at risk of extinction, because the resulting rapid warming would be too fast for species
13 to track the changing climate (Trisos et al., 2018).
14
15 Public perceptions of SRM
16
17 Studies on the public perception of SRM have used multiple methods: questionnaire surveys, workshops, and
18 focus group interviews (Burns et al., 2016; Cummings et al., 2017). Most studies have been limited to
19 Western societies with some exceptions. Studies have repeatedly found that respondents are largely unaware
20 of SRM (Merk et al., 2015). In the context of this general lack of familiarity, the publics prefer carbon
21 dioxide removal (CDR) to SRM (Pidgeon et al., 2012), are very cautious about SRM deployment because of
22 potential environmental side effects and governance concerns, and mostly reject deployment for the
23 foreseeable future. Studies also suggest conditional and reluctant support for research, including proposed
24 field experiments, with conditions of proper governance (Sugiyama et al., 2020). Recent studies show that
25 the perception varies with the intensity of deliberation (Merk et al., 2019), and that the public distinguishes
26 different funding sources (Nelson et al., 2021). Limited studies for developing countries show a tendency for
27 respondents to be more open to SRM (Visschers et al., 2017; Sugiyama et al., 2020), perhaps because they
28 experience climate change more directly (Carr and Yung, 2018). In some Anglophone countries, a small
29 portion of the public believes in chemtrail conspiracy theories, which are easily found in social media
30 (Tingley and Wagner, 2017; Allgaier, 2019). Since researchers rarely distinguish different SRM options in
31 engagement studies, there remains uncertainty in public perception.
32
33 Ethics
34
35 There is broad literature on ethical considerations around SRM, mainly stemming from philosophy or
36 political theory, and mainly focused on SAI (Flegal et al., 2019). There is concern that publicly debating,
37 researching and potentially deploying SAI could involve a `moral hazard', with potential to obstruct ongoing
38 and future mitigation efforts (Morrow, 2014; Baatz, 2016; McLaren, 2016), while empirical evidence is
39 limited and mostly at the individual, not societal, level (Burns et al., 2016; Merk et al., 2016; Merk et al.,
40 2019). There is low agreement whether research and outdoors experimentation will create a `slippery slope'
41 toward eventual deployment, leading to a lock-in to long-term SRM, or can be effectively regulated at a later
42 stage to avoid undesirable outcomes (Hulme, 2014; Parker, 2014; Callies, 2019; McKinnon, 2019).
43 Regarding potential deployment of SRM, procedural, distributive and recognitional conceptions of justice
44 are being explored, (Svoboda and Irvine, 2014; Svoboda, 2017; Preston and Carr, 2018; Hourdequin, 2019).
45 With the SRM research community's increasing focus on distributional impacts of SAI, researchers have
46 started more explicitly considering inequality in participation and inclusion of vulnerable countries and
47 marginalized social groups (Flegal and Gupta, 2018; Whyte, 2018; Táíwò and Talati, 2021), including
48 considering stopping research (Stephens and Surprise, 2020; National Academies of Sciences and Medicine,
49 2021a). There is recognition that SRM research has been conducted predominantly by a relatively small
50 number of experts in the Global North, and that more can be done to enable participation from diverse
51 peoples and geographies in setting research agendas and research governance priorities, and undertaking
52 research, with initial efforts to this effect (e.g., Rahman et al., 2018), noting unequal power relations in
53 participation could influence SRM research governance and potential implications for policy (Whyte, 2018;
54 Táíwò and Talati, 2021) (Winickoff et al., 2015; Frumhoff and Stephens, 2018; Biermann and Möller, 2019;
55 McLaren and Corry, 2021; National Academies of Sciences and Medicine, 2021b)
56
57 Governance of research and of deployment
Do Not Cite, Quote or Distribute 16-88 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1
2 Currently, there is no dedicated, formal international SRM governance for research, development,
3 demonstration, or deployment (see WG3 Chapter 14). Some multilateral agreements--such as the UN
4 Convention on Biological Diversity or the Vienna Convention on the Protection of the Ozone Layer--
5 indirectly and partially cover SRM, but none is comprehensive and the lack of robust and formal SRM
6 governance poses risks (Ricke et al., 2013; Talberg et al., 2018a; Reynolds, 2019a). While governance
7 objectives range broadly, from prohibition to enabling research and potentially deployment (Sugiyama et al.,
8 2018b; Gupta et al., 2020), there is agreement that SRM governance should cover all interacting stages of
9 research through to any potential, eventual deployment with rules, institutions, and norms (Reynolds,
10 2019b). Accordingly, governance arrangements are co-evolving with respective SRM technologies across
11 the interacting stages of research, development, demonstration, and--potentially--deployment (Rayner et
12 al., 2013; Parker, 2014; Parson, 2014). Stakeholders are developing governance already in outdoors research;
13 for example, for MCB and OAC experiments on the Great Barrier Reef (McDonald et al., 2019). Co-
14 evolution of governance and SRM research provides a chance for responsibly developing SRM technologies
15 with broader public participation and political legitimacy, guarding against potential risks and harms relevant
16 across a full range of scenarios, and ensuring that SRM is considered only as a part of a broader portfolio of
17 responses to climate change (Stilgoe, 2015; Nicholson et al., 2018). For SAI, large-scale outdoor
18 experiments even with low radiative forcing could be transboundary and those with deployment-scale
19 radiative forcing may not be distinguished from deployment, such that (MacMartin and Kravitz, 2019) argue
20 for continued reliance on modeling until a decision on whether and how to deploy is made, with modeling
21 helping governance development. For further discussion of SRM governance see Chapter 14, WG3.
22
23 [END CROSS-WORKING GROUP BOX SRM HERE]
24
25
26 16.6 Reasons for Concern Across Scales
27
28 This section builds on Section 16.5 which identifies and assesses key risks (KRs) and representative key
29 risks (RKRs), including conditions contributing to their severity (i.e., Figure 16.10), in two ways. First, we
30 consider those risks in the context of the global goal for sustainable development which can be impacted, as
31 expressed in the United Nations 2030 Agenda for Sustainable Development and the Sustainable
32 Development Goals (SDGs). This discussion supports further assessment in Chapter 18 on sustainable
33 system transitions and climate resilient development pathways. Second, the potential global consequences
34 are then elaborated in an updated assessment of five globally aggregated categories of risk, designated as
35 Reasons for Concern (RFCs), that evaluates risk accrual by global warming level.
36
37 16.6.1 Key Risks and Sustainable Development
38
39 The United Nations 2030 Agenda for Sustainable Development, and the Sustainable Development Goals
40 (SDGs) (UN, 2015), since 2015, have become an important vision for the United Nations member countries
41 (Chimhowu, 2019) as well as for corporations to contribute towards sustainable growth (UNDP et al., 2016;
42 Ike et al., 2019; van der Waal and Thijssens, 2020). Climate change risks, as embodied in the RKR and
43 RFCs, can affect attainment of the SDGs and have consequences for lives and livelihoods (related to SDGs
44 1, 4, 8 and 9), health and well-being (related to SDGs 2, 3 and 6), ecosystems and species (related to SDGs
45 6, 14 and 15), economic (related to SDGs 1, 8 and 12), social and cultural assets (related to SDGs 5, 10, 11,
46 16 and 17), services including ecosystem services (related to SDGs 6, 7, 11, 12, 14 and 15), and
47 infrastructure (related to SDGs 6, 7, 9, 11 and 12). This section assesses the level of linkages between key
48 risks with sustainable development, in terms of the SDG targets and indicators. This informs on the key risks
49 which are most relevant to consider with respect to the attainment of the SDGs.
50
51 16.6.1.1 Links Between Key Risks and SDGs
52
53 Within the AR6 cycle, the three IPCC Special Reports have all considered the relationships between climate
54 change impacts and actions and the SDGs. SR15 discussed priorities for sustainable development in relation
55 to climate adaptation efforts (Section 5.3.1, SR15); synergies and trade-offs of climate adaptation measures
56 (Section 5.3.2, SR15); and the effect of adaptation pathways towards a 1.5ºC warmer world (Section 5.3.3
57 SR15). The SRCCL considered impacts of desertification on SDGs 1 (no poverty), 2 (zero hunger), 13
Do Not Cite, Quote or Distribute 16-89 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 (climate), 15 (life on land), and 5 (gender) (IPCC, 2019a, Figure 3.9). Trade-offs and synergies between
2 SDGs 2 (zero hunger) and 13 (climate action) at the global level were recognised (IPCC, 2019a, Section
3 5.6.6, Figure 5.16). Various integrated response options, interventions and investments were also evaluated
4 within the SDG framework (IPCC, 2019a, Section 6.4.3). The SROCC (Chapter 5) concluded that climate
5 change impacts on the ocean, overall, will negatively affect achieving the SDGs with 14 (life below water)
6 being most relevant (Singh et al., 2019).
7
8 Many linkages between SDG 13 (climate action) and other SDGs have been identified (very high
9 confidence), (Blanc, 2015; Kelman, 2015; Northrop et al., 2016; Hammill and Price-Kelly, 2017; ICSU,
10 2017; Mugambiwa and Tirivangasi, 2017; Dzebo et al., 2018; Major et al., 2018; Nilsson et al., 2018;
11 Sanchez Rodriguez et al., 2018). In addition, interactions between different climate change actions and
12 SDGs, and interactions among SDGs themselves, have also been assessed (Nilsson et al., 2016; IPCC, 2018;
13 McCollum et al., 2018; Fuso-Nerini et al., 2019; IPCC, 2019b; Cernev and Fenner, 2020). The Cross-
14 Chapter Box GENDER in Chapter 18 assessment indicates the importance of gender considerations in
15 achieving success and benefits in adaptation efforts. Aligning climate change adaptation to the SDGs could
16 bring potential co-benefits, increased efficiency in funding, and reduce the gap between adaptation planning
17 and implementation (very high confidence) (IPCC, 2018; Sanchez Rodriguez et al., 2018; IPCC, 2019b;
18 IPCC, 2019a).
19
20 Progress towards meeting the SDGs has been recognized to be able to reduce global disparities and support
21 more climate resilient development pathways (IPCC WGII AR5, Chapter 13, p. 818; discussed further in
22 Chapter 18). Nevertheless, we are still lagging in achieving the 2030 Goals (OECD, 2019; Sachs et al.,
23 2021), and this affects societal vulnerability, readiness and risk response capacities (IPCC, 2019a, Chapters
24 6, 7, Chapters 6 and 8, this report). We assess the risk literature for linkages between key risks (grouped by
25 RKRs) and the indicators of the SDGs (UN, 2015) using text analysis (details in Supplementary Material
26 SM16.5) to identify the potential level of effect of different risks on the SDGs. Some 940 documents were
27 analysed. The SDG status is associated with projected climate hazards, also called climatic impact-drivers
28 (CID) (Ranasinghe et al., 2021) (panel a), and RKRs (panel c), summarising hazard and exposure with
29 vulnerability aspects, as expressed by challenges in achieving the SDGs (panel d), on a regional level (Figure
30 16.12).
31
32
Do Not Cite, Quote or Distribute 16-90 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1
2 Figure 16.12: Linkages between the projected climatic impact-drivers (CIDs) by region, Sustainable Development
3 Goals (SDGs) by region, and the Representative Key Risks (RKRs).
4
5
6 16.6.1.2 Results, Implications and Gaps
7
Do Not Cite, Quote or Distribute 16-91 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 Linkages between the 17 SDGs and the eight RKRs (Figure 16.15 bottom left panel) are mapped to the
2 regional SDG status (Figure 16.15 bottom right panel) and related to the climate hazards (CIDs) (Figure
3 16.15 top left panel). Interconnections between climate hazards (CIDs) and RKRs are complicated by the
4 possibility of concurrent weather events, extremes and longer term trends. Risks are compounded by existing
5 vulnerabilities (Iwama et al., 2016; Thomas et al., 2019b; Birkmann et al., 2021) and cascading
6 consequences (Pescaroli and Alexander, 2015; Pescaroli and Alexander, 2018; Yokohata et al., 2019) (see
7 for example Sections 3.4.3.5, 5.12, 6.2.6, 7.2.2.2) as well as interactions. The level of challenges faced in
8 attaining the SDGs is one metric for assessing vulnerability and lack of capacity to manage risks (Cernev and
9 Fenner, 2020). Other metrics are also available (Parker et al., 2019; Garschagen et al., 2021b; Birkmann et
10 al., 2022). From Figure 16.12, aside from SDG13 (climate action), the strongest connections and risk
11 challenges are with zero hunger (SDG2), sustainable cities and communities (SDG11), life below water
12 (SDG14), decent work and economic growth (SDG8), no poverty (SDG1), clean water and sanitation
13 (SDG6) and good health and well-being (SDG3) (high confidence). Other SDGs have strong linkages with
14 specific RKRs, for example, terrestrial and marine ecosystems with Life on land (SDG15); infrastructure
15 (RKR-C) with Industry, innovation and infrastructure (SDG9) and Affordable and clean energy (SDG7);
16 living standards (RKR-D) with Gender equality (SDG5); and peace and mobility (RKR-H) with Peace,
17 justice and strong institutions (SDG 16) (high confidence).
18
19 On a global scale, priority areas for regions can be evaluated from the intersection of climate hazards, risks
20 and the level of challenges in SDG attainment (Moyer and Hedden, 2020; Sachs et al., 2021). The greatest
21 linkages and effects on the SDGs will be due to risks to water (RKR-G), living standards (RKR-D), coastal
22 socio-ecological systems (RKR-A) and Peace and human mobility (RKR-H) (high confidence) (details in
23 Supplementary Material SM16.5).
24
25 In particular, coastal socio-ecological systems (RKR-A), living standards (RKR-D), food security (RKR-F),
26 water security (RKR-G) and peace and mobility (RKR-H), have strong linkages with SDG 2 (zero hunger),
27 for which there are significant to major challenges for all regions (high confidence). Almost all the RKRs are
28 strongly linked to SDGs 8 (decent work and economic growth), and 11 (sustainable cities and communities)
29 (high confidence), where regions such as Africa, Asia, and Central and South America face significant to
30 major challenges in attaining targets. All regions also face major to significant challenges affecting SDGs 14
31 (life below water) and 15 (life on land), which relate to terrestrial and ocean ecosystems (RKR-B) (high
32 confidence).
33
34 The analysis of RKR linkages to SDGs is also useful in identifying gaps and susceptibilities, especially for
35 developing future climate resilient development targets. This aspect is discussed further in Chapter 18. Gaps
36 may arise as SDG targets and indicators are not specifically focused on systems affected by climate change
37 risks or impacts. For example, in the SRCCL Section 7.1.2 Hurlbert et al. (2019), noted the absence of an
38 explicit goal for conserving fresh-water ecosystems and ecosystem services in the SDGs. Such gaps (Tasaki
39 and Kameyama, 2015; Guppy et al., 2019) are inevitable as the current SDG targets and indicators focus on
40 overall sustainable development. As another example, projected increases in frequency and intensity of hot
41 temperature extremes are likely to result in increased heat-related illness and mortality, yet heat extremes are
42 not called out as an SDG indicator under SDGs 3 (good health and well-being) nor 13 (climate action). The
43 gaps on climate-related metrics for impacts on health are just beginning to be evaluated (Lloyd and Hales,
44 2019, see also Section 7.1.6). The current SDG 13 (climate action) targets also do not specifically track the
45 possibility of differential impacts on society from disasters and extreme weather events (RFC2). For
46 example, the first indicator (13.1.1.1), `Number of deaths, missing persons and directly affected persons
47 attributed to disasters per 100,000 population', does not include any requirement for disaggregated data,
48 unlike several other socio-economic and population SDG indicators, making it difficult to track the different
49 effects that climate-related disasters are expected to have on men, women, and children across different
50 segments of society, relevant for distributional impacts (RFC3) (see also Section 8.3, Cross-Chapter Box
51 GENDER in Chapter 18). The risk consequences identified and discussed in each RKR (Section 16.5.2)
52 provide useful entry points for identifying indicators and metrics for monitoring and evaluating specific
53 impacts of key climate change risks. In addition, the sector and region chapters have considered various
54 adaptation responses relevant to the SDGs (see for example, Sections 3.6, 4.7.5, 5.13.3, 8.2.1.6, 10.6.1,
55 13.11.4, 14.6.3) with relevant metrics for evaluation.
56
Do Not Cite, Quote or Distribute 16-92 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 In summary, key risks, and the consequences arising from them, are directly linked to and will affect specific
2 indicators of the SDGs (high confidence). They also will be indirectly linked to, and thus affect, the SDGs
3 overall, due to the interactions between the key risks (Section 16.5) and between the SDGs themselves (very
4 high confidence). These results support previous findings that climate change impacts pose a risk to
5 achieving sustainability (Ansuategi et al., 2015; Chirambo, 2016; ICSU, 2017; Pradhan et al., 2017; Gomez-
6 Echeverri, 2018; IPCC, 2018; IPCC, 2019b; IPCC, 2019a; Cernev and Fenner, 2020). Not all observed or
7 expected consequences arising from the key risks are fully captured by the SDG indicators, and nor were
8 they designed to be. Therefore, for monitoring and assessing the climate risk impacts, it is useful to consider
9 specific, climate change impact indicators and metrics (Enenkel et al., 2020) to capture any realised impacts.
10
11 In the near term, the strength of connection between the RKRs and the SDGs, with respect to existing SDG
12 challenges, indicate probable systemic vulnerabilities and issues in responding to climatic hazards (UN-
13 IATFFD, 2019; Leal Filho et al., 2020; Weaver et al., 2020; Tiedemann et al., 2021) (high confidence). In
14 the medium to long term (associated with global warming levels of between 2°C and 2.7°C under SSP2-45
15 scenario), if such vulnerabilities and challenges cannot be substantially reduced, the hazards and risks
16 resulting from the projected climate hazards (CIDs) (Figure 16.12b, c) will further stress systems relevant for
17 sustainable development, based on current experience of the COVID-19 pandemic (UN-IATFFD, 2021, see
18 also Cross-Chapter Box COVID in Chapter 7; Section 8.2, Section 8.3) (medium confidence, based on
19 medium evidence, high agreement).
20
21 The potential impacts of the various climate hazards, the occurrence of extreme events, and the projected
22 trends of climate hazards, give rise to complex risks for ecological and human systems, which are
23 compounded by the exposure, vulnerability and sustainability challenges faced in different regions of the
24 world. The potential global consequences are elaborated in the next section which describes the framework
25 and approach for the assessment of the five Reasons for Concern.
26
27 16.6.2 Framework and Approach for Assessment of RFCs and Relation to RKRs
28
29 The `Reasons for Concern' (RFC) framework communicates scientific understanding about accrual of risk in
30 relation to varying levels of warming for five broad categories: risk associated with (1) unique and
31 threatened systems, (2) extreme weather events, (3) distribution of impacts, (4) global aggregate impacts, and
32 (5) large-scale singular events (Smith et al., 2001; Mastrandrea and Schneider, 2004; Schneider and
33 Mastrandrea, 2005). The RFC framework was first developed during the Third Assessment Report (Smith et
34 al., 2001) along with a visual representation of these risks as `burning embers' figures, and this assessment
35 framework has been further developed and updated in subsequent IPCC reports including AR5 (IPCC, 2014;
36 Oppenheimer et al., 2014) and the recent IPCC Special Reports (SR15 2018; SRCCL 2019; SROCC 2019).
37
38 Relationship between RKRs and RFCs
39 RFCs reflect risks aggregated globally that together inform the interpretation of dangerous anthropogenic
40 interference with the climate system. The five RFC categories are maintained as previously defined for
41 consistency with earlier assessments. Compared to the synthesis of risk across RKRs in Section 16.5, we
42 note that the RKRs and RFCs are complementary methods that aggregate individual risks into different but
43 interconnected categories (Figure 16.13).
44
45
Do Not Cite, Quote or Distribute 16-93 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1
2 Figure 16.13 Interconnections between the Key Risks, Representative Key Risks and the Reasons for Concern
3
4
5 We draw important distinctions between RFC and RKR. First, RFCs assess risks that might be of global
6 concern, while RKRs also include risks that may be of concern only locally or for specific population groups
7 (Figure 16.13). RFCs focus on the full range of increasing risk, and locate transitions between four
8 categories of risk: undetectable, moderate, high, and very high. RKRs focus on severe risks, and attempt to
9 elaborate when/where severe impacts may occur. RKR assessments focus on the conditions under which
10 some risks would become severe over the course of this century, while RFCs evaluate changes in risk levels
11 against gradual increase in temperature levels. The RKR analysis used specific definitions of severity
12 including quantified thresholds where possible, and this is distinct from the approach based on the combined
13 elements of risk used in the RFC expert elicitation process. Severity as defined in the RKRs is associated
14 with high or very high risk levels but does not align precisely with either of those categories, and a further
15 difference arises from a more explicit emphasis on irreversibility and adaptation limits in the very high risk
16 category in the RFCs. Thus RKR and RFC neither map directly to one another in terms of content, nor in
17 terms of the response metric.
18
19 The treatment of vulnerability and adaptation is different in the RKR and RFC assessments. The RKR
20 assessment considered specifically three alternative levels of vulnerability, whereas the RFC process did not
21 explicitly differentiate risk by level of vulnerability. Therefore, the global warming levels at which the
22 various RKR assessments identify risk of severe impacts are not directly comparable to risk transitions
23 identified in the RFC assessments. In addition, RKRs consider implications of low vs. high adaptation in
24 order to illustrate the potential role of ambitious adaptation efforts to limit risk severity; RFCs consider risks
25 in a no/low adaptation scenario only, although there is some discussion of the potential role of adaptation in
26 assessing the transition to very high risk. Last, both RKRs and RFCs focus on the 21st century scale, though
27 recognizing risk will continue to increase after 2100, but treat this timing issue differently: RKRs assess
28 severe risks over the course of this century and distinguish risks that are already severe, that will become
29 severe by the mid-century, or that will become severe by the end of the century; while RFCs assess risk level
30 irrespective of their timing, but according to different temperature levels.
31
Do Not Cite, Quote or Distribute 16-94 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 Many of the elements of risk which contribute to RKRs also contribute to risk within one or more RFCs. In
2 turn, elements of risk within some RFCs, such as extreme weather and changes in the earth system contribute
3 to risk within one or more RKR. Hence RFCs may incorporate elements of many different RKRs, and vice
4 versa. There are therefore common elements between some particular RKRs and RFCs: for example, risks to
5 terrestrial and ocean ecosystems (RKR-B) contributes strongly to RFC1 (Unique and Threatened Systems)
6 and RFC4 (Global Aggregate Impacts); while RFC2 (extreme weather events) has implications for all RKRs,
7 including direct linkages with critical physical infrastructure, networks and services (RKR-C). Furthermore,
8 risks emerging from the interaction of RKRs also contribute to the RFCs, but are only qualitatively described
9 in Section 16.5.4. For example, the effects of risks to terrestrial and ocean ecosystems (RKR-A) affect living
10 standards and equity (RKR-C), as does the associated decline in ecosystem services which then impacts
11 livelihoods (RKR-D).
12
13 Elicitation Methodology
14 The method used to develop judgments on levels of risk builds on the approach described in WGII AR5
15 Chapter 19 (Oppenheimer et al., 2014) and outlined in more detail in (O'Neill et al., 2017), while integrating
16 advances in the AR6 SRs including expert judgment (SRCCL, Zommers et al., 2020). We provide further
17 details on the underlying judgements of risk level compared to previous assessments by indicating key risk
18 criteria associated with each judgement: magnitude of adverse consequences, likelihood of adverse
19 consequences, temporal profile of the risk, and ability to respond to the risk (Section 16.5.1). The definitions
20 of risk levels used to make the expert judgements are presented in Table 16.7 (Section 16.5.1).
21
22
23 Table 16.7: Definition of Risk Levels for Reasons for Concern.
Level Definition
Undetectable (White) No associated impacts are detectable and attributable to climate change.
Moderate (Yellow) Associated impacts are both detectable and attributable to climate change with at least
medium confidence, also accounting for the other specific criteria for key risks.
High (Red) Severe and widespread impacts that are judged to be high on one or more criteria for
assessing key risks.
Very High (Purple) Very high risk of severe impacts and the presence of significant irreversibility or the
persistence of climate-related hazards, combined with limited ability to adapt due to the
nature of the hazard or impacts/risks.
24
25
26 A brief summary of the framework that was used to carry out the risk assessment, synthesis and expert
27 elicitation is presented here and details are provided in Supplementary Material SM16.5. Expert judgements
28 about the qualitatively defined levels of risk (i.e., undetectable, moderate, high, and very high) reached at
29 various levels of global average warming are informed by evidence of observed impacts illustrated in
30 Section 16.2 and variations in individual key risks under different scenarios of climate change,
31 socioeconomics and adaptation effort in Section 16.5. We follow the methodological advances from SRCCL
32 Chapter 7 (Hurlbert et al., 2019), which used an expert elicitation protocol for developing the burning
33 embers (Zommers et al., 2020). Specifically, we used expert participants from within the AR6 author team
34 and a protocol based on the modified Delphi technique (Mukherjee et al., 2015) and the Sheffield Elicitation
35 Framework (Oakley and O'Hagan, 2010; Gosling, 2018). This approach (Figure 16.14) includes a two-round
36 elicitation process with a first round of independent anonymous judgements about the global warming level
37 at which risk levels transition from one to the next, and a final round of group discussion and deliberation to
38 develop consensus. The results are then reported and additional references made to findings from other
39 relevant chapters in this report, and reviewed by authors who had not participated in the elicitation as part of
40 independent appraisal.
41
42
Do Not Cite, Quote or Distribute 16-95 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1
2 Figure 16.14 Expert elicitation approach for assessment of RFC risk level transitions. A more detailed description of
3 the methodology used in this elicitation is provided in the Supplementary Material (SM16.5).
4
5
6 The resulting risk transition or `ember' diagram illustrates the progression of socio-ecological risk from
7 climate change as a function of global temperature change, taking into account the exposure and
8 vulnerability of people and ecosystems, as assessed by literature-based expert judgment. Section 16.6.3
9 presents these diagrams for each Reason for Concern, providing information about the most important
10 literature-based evidence that experts used to make their judgements. Similar assessments for selected
11 individual KRs are discussed in Chapters 2, 7, 9, 12, 13, and 14.
12
13 Representation of Warming Levels
14 The RFC assessment reflects the latest understanding of warming reported in WGI AR6. Global surface
15 temperature was 1.09 [0.95 to 1.20]ºC higher in 20112020 than 18501900, with stronger warming over
16 land (1.59 [1.34 to 1.83]ºC) than over the ocean (0.88 [0.68 to 1.01]ºC) (WGI AR6 Cross Chapter Box 2.3
17 Table 1, Eyring et al. in Gulev et al., 2021). Warming levels are commonly reported and studied in the
18 impacts literature using two scales of spatially averaged temperature rise, global surface air temperature
19 (GSAT), commonly produced by General Circulation Models (GCMs) when projecting climate changes, and
20 global mean surface temperature (GMST), commonly used in empirical studies. Both have the same
21 reference point of pre-industrial of 18501900. The ember diagrams presented here use GSAT, which is
22 consistent with most literature of projected risk (largely based on the output of climate models). To the
23 extent that the embers also draw on the observed impacts literature using GMST, this potential variation is
24 minimal as the average levels of GSAT and GMST have been shown to match closely (for further discussion
25 on this see Cross-Chapter Box CLIMATE in Chapter 1). Hence the diagrams are presented with a single y-
26 axis representing global temperature change, generally referring to global temperature rise irrespective of
27 when it occurs: however, the majority of the literature assessed considers alternative levels of warming
28 during the twenty-first century. For example, a warming level of 2ºC might occur in the 2050s, the 2080s, or
29 in 2100 (see next section).
30
31 Furthermore, climate-related hazards associated with each of the RFCs are assessed in WGI AR6 Cross-
32 Chapter Box 12.1 Table 1 (Tebaldi et al., 2021) which synthesizes information from various chapters of
33 WGI on 35 such hazards according to global warming levels (GWLs) to inform understanding of their
34 potential changes and associated risks with temperature levels in general.
35
36 Temporal dimension
37 When are the risks shown in the embers projected to occur? The issues associated with assessing transient
38 risks are discussed in Chapter 3, SR15 (IPCC, 2018). Some of the literature, however, does explore the
39 dynamics within human and natural systems (i.e., the way in which systems respond when a transient level
40 of warming is first reached and then further, how they continue to develop if that transient level of warming
41 is then maintained indefinitely). We note that this important factor is captured in the RFC assessment (and
42 ember diagrams), since the timing of risk accrual is one of the criteria for the assessment of the level of risk
43 (16.5.1). Risks that are known to evolve only over very long-time scales contribute less to the level of risk
Do Not Cite, Quote or Distribute 16-96 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 than those which are known to occur rapidly. This is because sea level rise also depends on the dynamics of
2 global warming, including the rate of change of radiative forcing, and time lags of several decades, including
3 between atmospheric and ocean warming, and in reaching equilibrium sea level state (Oppenheimer et al.,
4 2019; Fox-Kemper et al., 2021). However, longer-term risks that would arise if those transient temperatures
5 were maintained are also included, and this is particularly important in RFC5 (large scale singular events).
6 Note that risks that take place over a very long timescale are considered to be of lower concern than more
7 imminent risks. However, changes of very large magnitude can still be very important even if far away in
8 time, especially if these changes are irreversible (or reversible only on extremely long time scales) (see
9 Section 16.5.1).
10
11 Although the embers do not indicate the decade in which certain risks are projected to occur, clearly this
12 depends strongly on the level of mitigation action as well as the degree of adaptation. Hence, the ember
13 diagram (Figure 16. 14) is shown alongside a graphic illustrating possible global temperature time series
14 emerging from alternative future scenarios assessed by WGI AR6 which imply different levels of mitigation
15 effort. For example, in a scenario with a high level of mitigation effort (SSP1-1.9) reaching net zero
16 emissions in the 2050s, it is extremely likely that global warming remains below 2°C and more than 50%
17 likely that it will remain below 1.6°C (AR6 WGI 4.3.1.1, Meinshausen et al., 2020). On the other hand, a
18 level of 2°C warming is extremely likely to be exceeded during the 21st century under the three scenarios
19 assessed by WGI AR6 in which greenhouse gas emissions do not fall below current levels before mid-
20 century (i.e., SSP2-4.5, SSP3-7.0, SSP-8.5) (WGI AR6 4.3.1.1, Lee et al., 2021). WGI AR6 has assessed that
21 `global surface temperature averaged over 20812100 is very likely to be higher by 1.0°C1.8°C under the
22 lowest CO2 emission scenario considered in this report (SSP1-1.9) and by 3.3°C5.7°C under the highest
23 CO2 emission scenario (SSP5-8.5)'. However, almost all scenarios assessed by IPCC AR6 WGI reach 1.5°C
24 global warming level in the early 2030s (WGI AR6 SPM, IPCC, 2021a).
25
26 Temperature overshoot
27 The concept of temperature overshoot, defined as `exceedance of a specified global warming level followed
28 by a decline to or below that level during a specified period of time' is a relevant consideration for this RFC
29 risk assessment; however, the effect of overshoot has not explicitly been considered in the burning ember
30 assessment due to the limited literature basis. However, despite the lack of directly assessed overshoot
31 scenarios, the current literature provides several salient examples of irreversible changes that are projected to
32 occur once global temperatures reach a particular level. For example, coral reefs are unable to survive
33 repeated bleaching events that are too close together, leading to irreversible loss of the reefs even if
34 bleaching were to cease (see Section 16.6.3.1 RFC1). Species extinction is irreversible, and Chapter 2
35 assesses that at ~1.6°C, >10% of species are projected to become endangered as compared with >20% at
36 ~2.1°C (median) representing high and very high biodiversity risk, respectively (medium confidence)
37 (Section 2.5.4). Similarly, WGI AR6 finds that `Over the 21st century and beyond, abrupt and irreversible
38 regional changes in the water cycle, including changes in seasonal precipitation, streamflow and aridity,
39 cannot be excluded'. Thus, information about irreversibility provides information about the potential
40 outcome of temperature overshoot scenarios. Other types of losses, such as loss of human or species life, are
41 irreversible even if the loss process ceases in the future. The less resilient a system is, the more likely it is to
42 suffer irreversible damage during a temperature overshoot; the more resilient it is, the more likely it is to be
43 able to withstand the overshoot or recover afterwards. Very high levels of risk, as assessed here in the
44 Reasons for Concern, are associated with a wide range of criteria for risk assessment including
45 irreversibility. Whilst not all very high risks are irreversible, in general risks reaching a very high level
46 include a component of irreversible risks that would persist during and after an overshooting of a given
47 temperature level.
48
49 Risks associated with socioeconomic development, mitigation and mal-adaptation
50 The ember diagrams in Figure 16.14 capture only the risks arising from exposure of vulnerable socio-
51 ecological systems to climatic hazards across a range of socioeconomic futures. They do not capture any risk
52 component arising solely from changes in population or level of development. Importantly, they also do not
53 capture additional risks that may arise from the human response to climate change, including climate change
54 mitigation or unintended negative consequences of adaptation-related responses (i.e., maladaptation)
55 (Section 17.5.1). Such risks are discussed in SRCCL Chapter 7, for example, adverse effects of the very
56 large-scale use of land and water for primary bioenergy production on food production and biodiversity
57 (Hurlbert et al., 2019). Contributions of mitigation or maladaptation to risk can be important, however, and
Do Not Cite, Quote or Distribute 16-97 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 are discussed further in the context of specific RFCs in Section 16.6.3. In general, such components of risk
2 are difficult to quantify, and can be minimised by good design of climate change mitigation and adaptation.
3 Thus, the effect is excluded from the ember diagrams to allow a more clear representation of the accrual of
4 climate change risk with global warming.
5
6 Emergent Risk
7 AR5 Oppenheimer et al. (2014) defined `emergent risk' as a risk that arises from the interaction of
8 phenomena in a complex system. While emergent risk is a relevant consideration for this RFC risk
9 assessment, this type of risk has not been explicitly accounted for in the burning ember assessment due to the
10 limited literature basis. Unlike known or identified risks, emergent risks are characterized by the uncertainty
11 of consequences and/or probabilities of occurrence. The International Risk Governance Council (IRGC)
12 suggests three categories of emergent risks: 1) high uncertainty and a lack of knowledge about potential
13 impacts and interactions with risk-absorbing systems; 2) increasing complexity, emergent interactions and
14 systemic dependencies that can lead to non-linear impacts and surprises; and 3) changes in context (for
15 example social and behavioural trends, organisational settings, regulations, natural environments) that may
16 alter the nature, probability and magnitude of expected impacts. Feedback processes between climatic
17 change, human interventions involving mitigation and adaptation actions, and processes in natural systems
18 can be classified as emergent risks if they pose a threat to human security.
19
20 16.6.3 Global Reasons for Concern
21
22 In this section we present the results of the expert elicitation in the form of the burning embers diagram,
23 alongside a description of the recent literature and scientific evidence for each of the RFCs in turn. The
24 consensus transition values are illustrated in Figure 16.14, an updated version of the burning embers diagram
25 that describes the additional risk due to climate change for each RFC when a temperature level is reached
26 and then sustained or exceeded. (Table SM16.18 in Supplementary Material SM16.6 presents the consensus
27 values of the transition range and median estimate in terms of global warming level by risk level for each of
28 the five RFC embers). The shading of each ember provides a qualitative indication of the increase in risk
29 with temperature, and we retain the color scheme employed in the most recent versions of this figure, where
30 white, yellow, red, and purple indicate undetectable, moderate, high and very high additional risk,
31 respectively. These transitions were assessed under conditions of low to no adaptation compared to today, in
32 accordance with definitions provided in 16.3 (i.e., adaptation consists of fragmented, localized, incremental
33 adjustments to existing practices), though the effect of adaptation on risk for individual RFCs and related
34 literature is discussed further below.
35
36 The following subsections present the expert assessment and judgments made during the elicitation process
37 to identify consensus transition values for each RFC. The description of these transitions is further extended
38 with additional references to findings from underlying chapters in this report, and reviewed by Chapter 16
39 authors as part of independent appraisal. No changes were made to the transition values assessed through the
40 expert elicitation.
41
42
Do Not Cite, Quote or Distribute 16-98 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1
2 Figure 16.15: The dependence of risk associated with the Reasons for Concern (RFCs) on the level of climate change,
3 updated by expert elicitation and reflecting new literature and scientific evidence since AR5 and SR15. (a) Global
4 surface air temperature (GSAT), relative to 1995-2014 (left axis) and pre-industrial, 1850-1900 (right axis) (WGI AR6
5 Figure 4.2a, (Lee et al., 2021)). (b) Embers are shown for each RFC, assuming low to no adaptation (i.e., adaptation is
6 fragmented, localized, incremental adjustments to existing practices). The horizontal line denotes the present global
7 warming of 1.2°C (WMO, 2020) which is used to separate the observed, past impacts below the line from the future
8 projected risks above it. RFC1 Unique and threatened systems: ecological and human systems that have restricted
9 geographic ranges constrained by climate related conditions and have high endemism or other distinctive properties.
10 Examples include coral reefs, the Arctic and its indigenous people, mountain glaciers and biodiversity hotspots. RFC2
11 Extreme weather events: risks/impacts to human health, livelihoods, assets and ecosystems from extreme weather
12 events such as heatwaves, heavy rain, drought and associated wildfires, and coastal flooding. RFC3 Distribution of
13 impacts: risks/impacts that disproportionately affect particular groups due to uneven distribution of physical climate
14 change hazards, exposure or vulnerability. RFC4 Global aggregate impacts: impacts to socio-ecological systems that
15 can be aggregated globally into a single metric, such as monetary damages, lives affected, species lost or ecosystem
16 degradation at a global scale. RFC5 Large-scale singular events: relatively large, abrupt and sometimes irreversible
17 changes in systems caused by global warming, such as ice sheet disintegration or thermohaline circulation slowing.
18 Comparison of the increase of risk across RFCs indicates the relative sensitivity of RFCs to increases in GSAT. The
19 levels of risk illustrated reflect the judgments of IPCC author experts from WGI and WGII.
20
21
22 16.6.3.1 Unique and Threatened Systems (RFC1)
23
24 This RFC addresses the potential for increased damage to or irreversible loss of a wide range of physical,
25 biological, and human systems that are unique (i.e., restricted to relatively narrow geographical ranges and
26 have high endemism or other distinctive properties) and are threatened by future changes in climate (Smith et
27 al., 2001; Smith et al., 2009; Oppenheimer et al., 2014). The specific examples of such systems given in
28 previous IPCC assessment reports has remained broadly consistent, with AR4 including `coral reefs, tropical
29 glaciers, endangered species, unique ecosystems, biodiversity hotspots, small island states, and indigenous
30 communities' (Smith 2009), AR5 including `a wide range of physical, biological, and human systems that
31 are restricted to relatively narrow geographical ranges' and `are threatened by future changes in climate'
Do Not Cite, Quote or Distribute 16-99 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 (Smith et al., 2001), while SR15 Chapter 3 included `ecological and human systems that have restricted
2 geographic ranges constrained by climate related conditions and have high endemism or other distinctive
3 properties. Examples include coral reefs, the Arctic and its indigenous people, mountain glaciers and
4 biodiversity hotspots'. In this cycle, we retain the definition used in SR15 as most explicit and inclusive of
5 the previous definitions.
6
7 AR5 (Oppenheimer et al., 2014) assessed the transition from undetectable to moderate risk for RFC1 to lie
8 below recent global temperatures (1986-2005, which at the time was considered to correspond to a global
9 warming level of 0.6°C above pre-industrial levels; AR6 WGI now considers this time period of 1986-2005
10 to correspond to a global warming or approximately 0.7°C). At that time, there was at least medium
11 confidence in attribution of a major role for climate change for impacts on at least one each of ecosystems,
12 physical systems, and human systems within this RFC. SR15 Section 3.5.2.1 (Hoegh-Guldberg et al.,
13 2018b), concurred with high confidence that the transition to moderate risk had already occurred before the
14 time of writing.
15
16 The transitions here are informed by these assessments, along with the assessment in Chapter 2 on species
17 high extinction risk and on ecosystem transitions. It also draws substantially from information in Cross-
18 Chapter Paper 1 and Table SM16.22 on risks to unique and threatened biological systems. Some unique and
19 threatened systems, such as coral reefs and sea-ice dependent ecosystems, were already showing attributable
20 impacts with high confidence (see table 16.1, Cross-Chapter Paper 1 and Chapter 2) based on data collected
21 in the mid to latter 20th century, when global warming of 0.5°C above pre-industrial levels had taken place,
22 as noted already in AR3. In this AR6 assessment, the temperature range for the transition from undetectable
23 to moderate risk is still located at a median value of 0.5°C above pre-industrial levels, with very high
24 confidence. Since impacts were first detected in coral reef systems in the 1980s when warming of ~0.4°C of
25 global warming had occurred (SR15 Chapter 3), this provides the temperature at which the transition begins.
26 The September Arctic sea ice volume has declined by 55-65% between 1979 and 2010 (AR6 WGI, Schweiger et
27 al., 2019) as global warming increased from around 0.36°C in 1979 to around 0.9°C in 2010. These provide
28 evidence of a start to the transition from undetectable to moderate risk at 0.4°C above pre-industrial levels.
29 Recent evidence of observed impacts on mountaintop ecosystems, sea ice dependent species, and of range
30 shifts in multiple ecosystems during 1990-2000, which AR6 WGI now assesses as corresponding to a global
31 warming of 0.69°C (see WGI AR6 Cross-Chapter Box 2.3, Figure 1, Gulev et al., 2021) provides evidence
32 for an upper limit to this transition of 0.7°C with very high confidence. Overall, the transition is located at a
33 median of 0.5°C with lower and upper limits of 0.4 and 0.7°C respectively with very high confidence.
34
35 AR5 assessed the transition from moderate to high risk to lie around 1°C above 19862005 levels (which
36 corresponded at that time to 1.6°C above pre-industrial levels but has been reassessed by AR6 WGI to
37 correspond to 1.7°C) to reflect projected `increasing risk to unique and threatened systems, including Arctic
38 sea ice and coral reefs, as well as threatened species as temperature increases over this range.' SR15
39 relocated the transition slightly from 1.6°C to 1.5°C, owing to increased literature projecting the effects of
40 climate change upon Arctic sea ice and new literature assessing projected impacts of climate change on
41 biodiversity at 1.5°C warming.
42
43 In this AR6 assessment, the transition from moderate to high is based on the high level of observed impacts,
44 and the areas projected to begin undergoing major transformations by 1.5ºC (see CCP1, Chapter 2 and
45 SR15). A substantial number of unique and threatened systems are assessed to be in a high risk state owing
46 to the influence of anthropogenic climate change by the 2000-2010 period, when global warming had
47 reached approximately 0.85ºC (range 0.7-1ºC) (see WGI AR6 Cross-Chapter Box 2.3, Gulev et al., 2021)
48 using the 1995-2014 figure as a proxy for 2000-2010).
49
50 The most prominent example of a system assessed to be already in a high risk state is that of coral reefs,
51 which are already degrading rapidly. Observed impacts on coral reefs increased significantly during 2014-
52 2017 (Table 16.2, corresponding to a global warming of about 0.9ºC). This includes mass bleaching in the
53 Indian Ocean in 1998, 2010, 2015 and 2016 when bleaching intensity exceeded 20% in surveyed locations in
54 the western Indian Ocean, eastern Indian Ocean and western Indonesia. In the tropical Pacific Ocean,
55 climate-driven mass bleaching was reported in all countries in the region, with most bleaching reports
56 coinciding with 2014-2017 marine heatwaves. 50% of coral within shallow-water reefs of the northern and
57 central two-thirds of the Great Barrier Reef were killed in 2015/16. Subsequent coral recruitment in 2018
Do Not Cite, Quote or Distribute 16-100 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 was reduced to only 11% of the long-term average, representing an unprecedented shift in the ecology of the
2 northern and middle sections of the reef system to a highly degraded state. A second key example are sea ice
3 dependent systems in the Arctic. During August-October of 2010-2019, corresponding to a global warming of
4 about 0.9°C, average Arctic sea-ice area has declined in area by 25% relative to 1979-1988 (high confidence, AR6
5 WGI, Figure 9.13). September Arctic sea ice volume has declined by about 72 % between 1979 and 2016, with
6 the latter deemed a conservative estimate (AR6 WGI, Schweiger et al., 2019).
7
8 Other important examples of observed impacts on unique ecosystems that indicate that risks are already at a
9 high level (Table SM16.22) include mass tree mortalities, now well recorded in multiple unique forest and
10 woodland ecosystems around the world. Sections 2.4.3.3 and 2.4.5 report that between 1945 and 2007,
11 drought-induced tree mortality (sometimes associated with insect damage and wildfire) has caused the
12 mortality of up to 20% of trees in western North America, the African Sahel, and North Africa, linked to a
13 warming of 0.3-0.9ºC above pre-industrial levels, and is implicated in more than 100 other cases of drought-
14 induced tree mortality in Africa, Asia, Australia, Europe, and North and South America (high confidence).
15 Species in biodiversity hotspots already show changes in response to climate change (CCP1, high
16 confidence). Román-Palacios and Wiens (2020) attribute local extinctions of several taxonomic groups
17 between the latter 20th century and 2003-2012, (corresponding to warming of less than 0.85ºC) to climate-
18 change related temperature extremes for up to 44% (0-75%) of species. Widespread declines of up to 35% in
19 the species richness of the unique pollinator group, bumble bees, between 1901 - 1974 and 2000 - 2014 are
20 also attributed to climate change, via increasing exceedance of their thermal tolerance limits across Europe
21 and North America (Soroye et al., 2020). The first extinctions attributed to climate change have been now
22 detected with the present 1.2ºC warming including that of the Bramble Cays Melomys (Melomys rubicola), a
23 sub-species of the lemuroid ringtail possum (Hemibelideus lemuroides), and golden toad (Incilius
24 periglenes) (Chapter 2). An increasing frequency or unprecedented occurrence of mass animal mortality due
25 to climate-change enhanced heat waves have also been observed in recent years on more than one continent,
26 including temperature vulnerable terrestrial birds and mammals in South Africa and Australia (Ratnayake et
27 al., 2019; McKechnie et al., 2021). There have also been 90% declines in sea ice dependent species such as
28 sea lions and penguins in the Antarctic (Table 16.2). A strong effect of climate change on the observed
29 contraction of ranges of polar fish species and strong expansion of ranges of arcto-boreal or boreal fish was
30 observed between 2004 and 2012 Frainer et al. (2017). Even if current human driven habitat loss is excluded,
31 many hotspots are projected to cease to be refugia (i.e., to remain climatically suitable for >75% of the
32 species they contain which have been modelled), at 1.0-1.5ºC (Cross-Chapter Paper 1).
33
34 Based on observed and modelled impacts to unique and threatened systems, including in particular coral
35 reefs, sea ice dependent systems, and biodiversity hotspots, AR6 assesses that the transition to high risks for
36 RFC1 have already occurred at a median level of 0.9ºC, with a lower bound at 0.7ºC and an upper bound at
37 the present day level of global warming of 1.2ºC (WMO, 2020) (very high confidence).
38
39 Identification of the transition to very high risk is associated by definition with the reaching of limits to
40 natural and/or societal adaptation. Adaptation which occurs naturally is already included in the risk
41 assessment, but experts also discussed the effect of additional human-planned adaptation in reducing risk
42 levels in RFC1. This additional adaptation could help species to survive in situ despite a changing climate
43 (for example by reducing current anthropogenic stresses such as over harvesting), or facilitate the ability of
44 species to shift geographic range in response to changes in climate, and the potential benefits of nature-based
45 solutions and restoration (see Cross-Chapter Box NATURAL and Section 2.6.5.1 in Chapter 2).
46
47 When considering planned adaptation, the main option often considered in terrestrial ecosystems is the
48 expansion of the protected area network, which is broadly beneficial in increasing the resilience of
49 ecosystems to climate change (e.g., Hannah et al., 2020). However, this action is not effective if the unique
50 and threatened systems in question reach a hard limit to adaptation (as in the case of the loss of Arctic
51 summer sea ice, the submergence of a small island, the contraction and elimination of a species' climatic
52 niche from a mountaintop, or the degradation of a coral reef) (Section 16.4). Furthermore, adaptation
53 benefits deriving from restoration rapidly diminish with increasing temperature (Cross-Chapter Paper 1).
54 One study quantifies how land management (in terms of protecting existing ecosystems or restoring lost
55 ones) might reduce extinctions in biodiversity hotspots or globally significant terrestrial biodiversity areas
56 more generally (Warren et al., 2018b). Whilst the latter suggests that substantial benefits can result globally
57 in terrestrial systems, allowing less unique systems to persist at higher levels of warming but only under a
Do Not Cite, Quote or Distribute 16-101 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 high adaptation scenario in which globally applied terrestrial ecosystem restoration and protected area
2 expansion takes place, this is less likely for many of the unique and threatened terrestrial systems which are
3 more vulnerable than the globally significant biodiversity areas treated in that study (which excludes coral
4 reefs and Arctic sea ice dependent systems). Such high levels of adaptation globally are likely infeasible
5 owing to competition for land use with food production (Pörtner et al., 2021). Novel targeted adaptation
6 interventions for coral reefs such as artificial upwelling and local radiation management show some promise
7 for reducing the adverse effects of thermal stress and resulting coral bleaching Condie et al. (2021), but are
8 far from implementation (Sawall et al., 2020; Kleypas et al., 2021). Larger benefits in this RFC could
9 theoretically accrue only if adaptation action became ubiquitous and extensive, which experts considered
10 infeasible at the scales required. Small island communities are confronted by socio-ecological limits to
11 adaptation well before 2100, especially those reliant on coral reef systems for their livelihoods, even for a
12 low emissions pathway (Chapter 3) (high confidence). At warming levels beyond 1.5°C, the potential to
13 reach biophysical limits to adaptation due to limited water resources are reported for Small Islands (medium
14 confidence) and unique systems dependent on glaciers and snowmelt (Chapter 4) (medium confidence).
15
16 AR5 assessed with high confidence that the transition from high to very high risks for RFC1 to lie around
17 2°C above 19862005 levels (then considered to correspond to 2.6°C above pre-industrial levels) to reflect
18 the very high risk to species and ecosystems projected to occur beyond that level as well as limited ability to
19 adapt to impacts on coral reef systems and in Arctic sea ice-dependent systems. Using the additional
20 literature which became available on projected risks to Arctic sea ice, biodiversity and ecosystems at 1.5°C
21 vs 2°C warming above pre-industrial levels, SR15 assessed that the transition from high to very high risks in
22 RFC1 lay between 1.5°C and 2°C above pre-industrial levels.
23
24 In AR6, risks are considered to start to transition from high to very high risks above 1.2°C warming (present
25 day, WMO, 2020), with a median value of 1.5°C, owing in particular to the observation of a present day
26 onset of ecosystem degradation in coral reefs, which are projected in the SR15 report `to decline by a further
27 7090% at 1.5ºC (very high confidence)' . The literature for projected increases in risk to other unique and
28 threatened systems and their limited ability to adapt above 2ºC warming is substantial and robust and the
29 confidence level in very high risk remains high. At 2ºC, 18% of 34,000 insects are projected to lose >50%
30 climatically determined geographic range, as compared with 6% at 1.5ºC (Warren et al., 2018a). The risk of
31 species extinction increases with warming in all climate change projections, for all native species studied in
32 biodiversity hotspots (Cross-Chapter Paper 1, high confidence), being roughly threefold greater for endemic
33 than more widespread species for global warming of 3°C above pre industrial levels than 1.5°C) (Manes et
34 al., 2021, Cross-Chapter Paper 1) (medium confidence). The Arctic is projected to be practically ice free in
35 September in some years for global warming of between 1.5 and 2ºC (WGI AR6 Section 9.3.1.1, Fox-
36 Kemper et al., 2021), undermining the persistence of ice dependent species such as polar bears, ringed seals
37 and walrus (Meredith et al., 2019), and adversely affecting indigenous communities. Warming of 1.5ºC is
38 also assessed (Chapter 3) to reduce the habitability of small islands, due to the combined impacts of several
39 key risks (high confidence). Hence the transition from high to very high risk in these systems is assessed to
40 occur with high confidence beginning at 1.2ºC, passing through a median value of 1.5°C, and completing
41 (i.e. reaching its upper bound) at 2ºC warming.
42
43 16.6.3.2 Extreme Weather Events (RFC2)
44
45 This RFC addresses the risks to human health, livelihoods, assets and ecosystems from extreme weather
46 events such as heatwaves, heavy rain, drought and associated wildfires, and coastal flooding (Hoegh-
47 Guldberg et al., 2018b). Previous assessments of this RFC have focused mainly on changes to the hazard
48 component of the risk, using the projected increase in hazard as an indicator of higher risk. However, in AR6
49 an expanding (although still smaller) body of evidence now allows also incorporation of the exposure and/or
50 vulnerability components of risk and, to a limited extent, their trends.
51
52 AR5 identified a transition from undetectable to moderate risk below `recent' temperatures (i.e., during
53 1986-2005, which then corresponded to a global warming of 0.6ºC above pre-industrial levels). SR15
54 Section 3.5.2.2 (Hoegh-Guldberg et al., 2018b), concluded that differences of 0.5ºC in global warming led to
55 detectable changes in extreme weather and climate events on the global scale and for large regions. IPCC
56 WGI AR6 Chapter 11 confirms this assessment and concludes that `new evidence strengthens the conclusion
57 from SR15 that even relatively small incremental increases in global warming (+0.5°C) cause statistically
Do Not Cite, Quote or Distribute 16-102 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 significant changes in extremes on the global scale and for large regions'. Substantial literature is available
2 for comparisons at +1.5ºC vs +2ºC of global warming, but the conclusions are assessed to also apply at lower
3 global warming levels and smaller increments of global warming given the identified linearity of regional
4 responses of several extremes in relation to global warming(Seneviratne et al., 2016; Wartenburger et al.,
5 2017; Tebaldi and Knutti, 2018) and the identification of emergence of global signals in climate extremes for
6 global warming levels as small as 0.1°C (Seneviratne and Hauser, 2020, WGI AR6, Chapter 11, Figure 11.8;
7 WGI Cross-Chapter Box 12.1). Further analyses are consistent with this assessment, based on model
8 simulations (Fischer and Knutti, 2015; Schleussner et al., 2017; Kirchmeier-Young et al., 2019a; Seneviratne
9 and Hauser, 2020) and observational evidence (Zwiers et al., 2011; Dunn et al., 2020). A global warming of
10 +0.5ºC above pre-industrial conditions corresponds approximately to climate conditions in the 1980s
11 (Chapter 2, Figure 2.11), a time frame at which detectable changes in some extremes were established at the
12 global scale based on observations (Dunn et al., 2020). Heat-related mortality has also been assessed to have
13 increased considerably because of climate change (Ebi et al., 2021; Vicedo-Cabrera et al., 2021). The onset,
14 and also median location of the transitions of risk (Figure 16.15) from undetectable to moderate, is therefore
15 considered to be 0.5ºC. Further strong new evidence shows that changes in extremes emerged during the
16 1990s and 2000s (Dunn et al., 2020) by which time +0.7ºC of global warming had taken place (IPCC SR15,
17 Chapter 1; WGI AR6, Chapter 2). In AR5 Section 19.6.3.3 (Oppenheimer et al., 2014), a transition to
18 moderate risk was assessed to have taken place at the then `recent' global warming level of 0.6ºC, with high
19 confidence. Owing to the increase in evidence, there is now very high confidence that the median value of
20 the transition from undetectable to moderate risk is at 0.5ºC and led by heat extremes, with the lower
21 estimate set at 0.5°C as well, and upper estimate at 0.7°C.
22
23 Further evidence of more recent observed changes in extreme weather and climate events, and their potential
24 for associated adverse consequences across many aspects of society and ecosystems, has continued to accrue
25 (WGI AR6 Chapter 11; WGI AR6 Chapter 12). Since a necessary condition for `moderate' levels of risk is
26 the detection and attribution of observed impacts, the following text provides an overview of some salient
27 examples of this evidence. In particular, WGI AR6 Chapter 11 (Seneviratne et al., 2021) concludes that some
28 recent hot extreme events that happened in the past decade (2010s) would have been extremely unlikely to
29 occur without human influence on the climate system. Global warming in that decade reached approximately
30 1.09°C on average (IPCC WGI AR6 Chapter 2).
31
32 Assessment of a high level of risk requires a higher level of magnitude, severity and spatial extent of the
33 risks. Events prior to that already had substantial impacts such as the 2003 European heatwave (IPCC SREX
34 Chapter 9). Examples of impactful events in the early 2010s (at ca. 0.95°C of global warming, (WGI AR6
35 Chapter 2, Gulev et al., 2021) include the 2010 Russian heatwave (Barriopedro et al., 2011) and the 2010
36 Amazon drought (Lewis et al., 2011). Later impactful events include, among others, the 2013 heatwave in
37 eastern China (Sun et al., 2014), the 2017 tropical cyclone Harvey (Risser and Wehner, 2017; Van
38 Oldenborgh et al., 2017), and the 2018 concurrent north hemisphere heatwaves in Europe, North America
39 and Asia (Vogel et al., 2019). Very recent events with severe and unprecedented impacts attributed to
40 anthropogenic climate change indicate that thresholds to high risks may already have been crossed at recent
41 levels of global warming (ca. 1.1°C-1.2°C) including the Siberian fires and the 2019 Australian bushfires
42 that were linked to extreme heat and drought conditions (Van Oldenborgh et al., 2017) and extreme
43 precipitation linked to increased storm activity in the US (Van Oldenborgh et al., 2017). Severe and
44 unprecedented impacts occurred with current low levels of adaptation (16.2.3.4). The global-scale risk of
45 wildfire considerably degrading ecosystems and increasing illnesses and death of people has been assessed to
46 transition from undetectable to moderate over the range 0.6 to 0.9°C with high confidence (Chapter 2, Table
47 2.S.4, Figure 2.11).
48
49 In addition, long-term trends in various types of extremes are now detectable (WGI AR6 Chapter 11,
50 Seneviratne et al., 2021). This includes increases in hot extremes over most land regions (virtually certain),
51 increases in heavy precipitation at the global scale and over most regions with sufficient observations (high
52 confidence), and increases in agricultural and ecological droughts in some regions (medium confidence)
53 (WGI AR6 Chapter 11). There has also been overall a likely increase in the probability of compound events,
54 such as an increase in concurrent heatwaves and droughts (high confidence) (WGI AR6 Chapter 11). There
55 is medium confidence that weather conditions that promote wildfires (fire weather) have become more
56 probable in southern Europe, northern Eurasia, the US, and Australia over the last century (WGI AR6
57 Chapter 11; SRCCL Chapter 2, Jolly et al., 2015; Abatzoglou and Williams, 2016). Furthermore, food
Do Not Cite, Quote or Distribute 16-103 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 security and livelihoods are being affected by short-term food shortages caused by climate extremes (5.12.1;
2 Chapter 16, Food Security RKR) which have affected the productivity of all agricultural and fishery sectors
3 (high confidence). The frequency of sudden food production losses has increased since at least mid-20th
4 century on land and sea (medium evidence, high agreement). Droughts, floods, and marine heatwaves
5 contribute to reduced food availability and increased food prices, threatening food security, nutrition, and
6 livelihoods of millions (high confidence). Changes in sea surface temperatures drive simultaneous variation
7 in climate extremes increasing the risk of multi-breadbasket failures (Cai et al., 2014; Perry et al., 2017).
8 Droughts induced by the 2015-2016 El Niño, partially attributable to human influences (medium confidence),
9 caused acute food insecurity in various regions, including eastern and southern Africa and the dry corridor of
10 Central America (high confidence). Human-induced climate change warming also worsened the 2007
11 drought in southern Africa, causing food shortages, price spikes, and acute food insecurity in Lesotho
12 (Verschuur et al., 2021). In the fisheries and aquaculture sector, marine heat waves are estimated to have
13 doubled in frequency between 1982 and 2016, as well as increasing in intensity and length, with
14 consequences for fish mortality (Ch 5) (Smale et al., 2019; Laufkötter et al., 2020). In the northeast Pacific, a
15 recent 5-year warm period impacted the migration, distribution, and abundance of key fish resources (high
16 confidence). At 1ºC warming the number of people affected by six categories of extreme events was found to
17 have already increased by a factor of 2.3 relative to preindustrial (Lange et al., 2020).
18
19 The general picture is one of annual or more frequent occurrences of severe extremes with widespread
20 impacts (as also reflected in section 16.2), and of multiple extremes, meeting the criteria for the `severe and
21 widespread' nature of risks that is required for classification at a `high' level of risk. This is consistent with
22 AR5 Chapter 19 (Oppenheimer et al., 2014), and gives high confidence that the lower threshold for entering
23 high risks associated with extreme weather events is +1ºC, and that the best estimate is that this transition
24 already occurred now that global warming has reached its present-day level of ca. 1.2ºC (WMO, 2020),
25 slightly above the 1.09°C average conditions in the 2010s, i.e. 2011-2020 (IPCC WGI AR6 Chapter 2, Gulev
26 et al., 2021).
27
28 A range of literature projects further substantial increases in several extreme event types with a global
29 warming of +1.5ºC, notably hot extremes in most regions, heavy precipitation in several regions, and
30 drought in some regions (IPCC SR15; WGI AR6 , Chapter 11). In particular, heavy precipitation and
31 associated flooding are projected to intensify and be more frequent in most regions in Africa and Asia (high
32 confidence), North America (medium to high confidence depending on the region), and Europe (medium
33 confidence). Also, more frequent and/or severe agricultural and ecological droughts are projected in a few
34 regions in all continents except Asia, compared to 18501900 (medium confidence); increases in
35 meteorological droughts are also projected in a few regions (medium confidence). Increases at 1.5°C of
36 global warming are projected in marine heatwaves (Laufkötter et al., 2020) and the occurrence of fire
37 weather (IPCC, 2019a). Heat-related mortality is assessed to increase from moderate to high levels of risk
38 under about 1.5ºC warming under SSP3, a socioeconomic scenario with large challenges to adaptation (Ebi
39 et al., 2021) especially in urban centres (Chapter 6). An additional 350 million people living in urban areas
40 are estimated would be exposed to water scarcity from severe droughts at 1.5°C warming (Section 6.1;
41 Section 6.2.2; CCP2 Coastal Cities). In summary, there is high confidence that the best estimate for the
42 transition from moderate to high risk is 1.2°C of global warming, with 1°C as lower estimate and 1.5°C as
43 upper estimate. The latter would be set to 1.3°C for an assessment at medium confidence.
44
45 As in RFC1, one of the criteria for identification of very high risks is limits to adaptation. Though the
46 literature explicitly considering societal adaptation to extreme weather events is limited, there is evidence
47 that investments in hydro-meteorological information, early warning systems and anticipatory forecast-based
48 finance are a cost-effective way to prevent some of the most adverse effects of extreme events (Coughlan de
49 Perez et al., 2016; Fakhruddin and Schick, 2019; Merz et al., 2020). Despite a lack of systematic methods for
50 assessing general adaptation effectiveness, there is some evidence of risk reduction for particular places and
51 hazards, especially flood and heat vulnerability (16.3.2.4) including investment in flood protection, building
52 design and monitoring and forecasting, air conditioning, reduced social vulnerability, and improved
53 population health. One study finds declining global mortality and economic loss due to extreme weather
54 events over the past four decades Formetta and Feyen (2019) especially in low income countries. Using
55 SSP2 as a proxy for expanded adaptation, Ebi et al. (2021) assesses that the transition to high risk for heat-
56 related mortality increases to 1.8ºC (compared to 1.5ºC with less adaptation under SSP3). There is evidence
57 of adaptation avoiding heat-related mortality at low levels of global warming, using early warning and
Do Not Cite, Quote or Distribute 16-104 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 response systems and sustainable alterations of the thermal environment at the individual, building, urban,
2 and landscape levels (Jay et al., 2021). Despite the evidence that adaptation can reduce risks of heat stress,
3 the impact of projected climate change on temperature-related mortality is expected to be a net increase
4 under a wide range of climate change scenarios, even with adaptation (Ch 7, high confidence). Much of the
5 adaptation literature focuses on coping with long-term gradual climate change and largely does not take into
6 account the increased difficulty of adapting to climate extremes and general higher variability in climate that
7 is projected to occur in the future. However, expanding and more coordinated adaptation, including wider
8 implementation and multi-level coordination, has the potential to reduce the risks to crops from heatwaves at
9 intermediate (but not high) levels of warming.(IPCC AR5 Ch7, Ahmed et al., 2018; Ahmed et al., 2019,
10 Section 16.3.2.2; EEA, 2019; Raza et al., 2019; Tripathi and Sindhi, 2020).
11
12 The transition from high to very high risk for the RFC2 was not assessed in the AR5 or in SR15. Some new
13 evidence suggests, however, that very high risks associated with weather and climate extremes would be
14 reached at higher levels of global warming. In particular, changes in several hazards would be more
15 widespread and pronounced at 2°C compared to 1.5°C global warming, including increases in multiple and
16 concurrent extremes (IPCC WGI AR6 SPM; IPCC WGI AR6 Chapter 11, IPCC WGI AR6 Chapter 12). On
17 average over land, high temperature events that would have occurred once in 50 years in the absence of
18 anthropogenic climate change are projected to become 13.9 times more likely with 2ºC warming, and 39.2
19 times more likely with 4ºC warming (IPCC AR6 WGI SPM Figure 6, IPCC, 2021b) indicating a non-linear
20 increase with warming. Ch 2 has assessed that risk of wildfire transitions from moderate to high over the
21 range 1.5ºC to 2.5ºC warming (medium confidence, Table 2.S.4, Figure 2.11). The intensity of heavy
22 precipitation events increase overall by about 7% for each additional degree of global warming (IPCC AR6
23 WGI SPM), while their frequency increases non-linearly. Events that would have occurred once every 10
24 years in a climate without human influence are projected to become 1.7 times more likely with 2ºC warming,
25 and 2.7 times more likely with 4ºC warming (IPCC AR6 WGI SPM Figure 6). Several AR6 regions are
26 projected to be affected by increases in agricultural and ecological droughts at 2°C of global warming,
27 including W. North-America, C. North-America, N. Central-America, S. Central-America, Caribbean, N.
28 South-America, N.E. South-America, South-American-Monsoon, S.W. South-America, S. South-America,
29 West & Central-Europe, Mediterranean, W. Southern-Africa, E. Southern-Africa, Madagascar, E. Australia,
30 and S. Australia (IPCC WGI AR6, Chapter 11, Seneviratne et al., 2021). This is a substantially larger
31 number compared to projections at 1.5°C (IPCC WGI AR6, Chapter 11, Seneviratne et al., 2021). In these
32 drying regions, events that would have occurred once every 10 years in a climate without human influence
33 are projected to happen 2.4 times more frequently at 2°C of global warming (IPCC WGI AR6 SPM Figure
34 6). Urban land exposed to floods and droughts is very likely to have more than doubled between 2000 and
35 2030, and the risk of flooding accelerates after 2050 (Ch 4). At 2ºC of global warming, there are also
36 significant projected increases in fluvial flood frequency and resultant risks associated with higher
37 populations exposed to these flood risks (Alfieri et al., 2017; Dottori et al., 2018) projected.
38
39 Heat-related mortality is assessed to increase from high to very high by 3°C under SSP3, a socioeconomic
40 scenario with large challenges to adaptation (Ebi et al., 2021). SRCCL assessed that very high risks would be
41 reached in association with wildfire above 3ºC of global warming (IPCC, 2019a). Chapter 2 has assessed that
42 risk of fire weather itself transitions from high to very high over the range 3ºC to 4.5ºC warming (medium
43 confidence, Table 2.S.4, Figure 2.11). Matthews et al. (2017) show that at 1.5ºC of global warming, about
44 40% of all megacities would be affected at least 1 day per year with a heat index above 40.6ºC (i.e., with
45 40.6ºC `feels-like' temperatures, accounting for moisture effects). This number would reach about 65% of
46 megacities at 2.7ºC and close to 80% at 4ºC. In addition, there is evidence for a higher risk of concurrent
47 heat extremes at different locations with increasing global warming (Vogel et al., 2019), meaning that
48 several cities could be affected by deadly heatwaves simultaneously. Laufkötter et al. (2020) found that
49 marine heatwave events would become annual to decadal events under 3ºC of global warming, with
50 consequences for aquaculture (Chapter 5). Gaupp et al. (2019) conclude that risks of simultaneous crop
51 failure across worldwide breadbasket regions, due to changes in maximum temperatures in the crop-growth
52 relevant season or cumulative precipitation in relevant time frames, increase disproportionately between
53 1.5ºC and 2ºC of global warming. Populations exposed to extreme weather and climate events may consume
54 inadequate or insufficient food, leading to malnutrition and increasing the risk of disease (Ch 5, high
55 confidence). Hence, there is the potential for very high risks associated with changes in climate extremes for
56 food security in the low adaptation case, already above 2ºC of global warming. Finally, studies suggest that
57 regional thresholds for climate extremes could be reached at 2ºC of global warming, for instance in the
Do Not Cite, Quote or Distribute 16-105 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 Mediterranean (Guiot and Cramer, 2016). Samaniego et al. (2018) conclude that soil moisture droughts in
2 that region would become 23 times longer than at the end of the 20th century at 2ºC, and 34 times longer
3 (125 days long per year) at 3ºC of global warming. There is clear evidence of very high risk at 3ºC global
4 warming for wildfires, marine heatwaves, and heatwaves in megacities (the latter being set at 2.7ºC).
5
6 Based on the available evidence, we assess that there is medium confidence that the transition to very high
7 risk would happen at a median value 2ºC of global warming, considering the increased risk for breadbasket
8 failure and irreversible impacts associated with changes in extremes at this warming level (e.g. damages to
9 ecosystems, health impacts, severe coastal storms), but that due to the disproportionate increases in risk
10 between 1.5 and 2ºC this transition begins already at 1.8ºC. The higher range for this transition is set with
11 medium confidence at 2.5ºC in this low/no adaptation scenario, owing to the further projected non-linear
12 increases in risks associated with high temperature events above 2ºC (WGI AR6 Figure SPM.6,, IPCC,
13 2021b; Cross-Chapter Box12.1, Ranasinghe et al., 2021), and also the limits to adaptation associated with
14 dealing with a rapid escalation of extreme weather events globally during this century; extreme events are
15 particularly difficult to adapt to and thus more often exceed hard limits to adaptation, particularly in natural
16 ecosystem settings (Section 16.4).
17
18 16.6.3.3 Distribution of Impacts (RFC3)
19
20 RFC3 reflects how key risks are distributed unevenly across regions and different population groups, due to
21 the non-uniform spatial distributions of physical climate change hazards, exposure, and vulnerability across
22 regions. It addresses how risks disproportionately affect particularly vulnerable societies and socio-
23 ecological systems, including disadvantaged people and communities in countries at all levels of
24 development. AR5 concluded that low-latitude and less developed areas generally face greater risk than
25 higher latitude and more developed countries, including for food- and health-related risks. This conclusion
26 remains valid and is now supported by greater evidence across a range of sectors and geographic regions.
27
28 Note that the assessment here is largely based on the national and regional distribution of impacts, rather
29 than sub-national distribution or explicit consideration of vulnerable elements of society. Climate risks are
30 also strongly related to inequalities, often but not always intersecting with poverty (16.1), geographic
31 location, political and socio-cultural aspects. Thus, countries with high inequality tend to be more vulnerable,
32 and more exposed, to climate hazards (16.1). Whilst the literature assessed here tends to be insufficiently
33 granular to resolve local inequalities, it does confirm the AR5 finding that low-latitude and less developed
34 areas generally face greater risk.
35
36 AR6 continues to highlight the uneven regional distribution of projected climate change risks. Biodiversity
37 loss is projected to affect a greater number of regions with increasing warming, and to be highest in northern
38 South America, southern Africa, most of Australia, and northern high latitudes (Section 2.5.1.3, medium
39 confidence). Climate change is projected to increase the number of people at risk of hunger in mid-century,
40 concentrated in Sub-Saharan Africa, South Asia and Central America (Chapter 5, high confidence),
41 increasing undernutrition, stunting, and related childhood mortality particularly in Africa and Asia and
42 disproportionately affecting children and pregnant women (Chapter 7, high confidence) strongly mediated by
43 socio-economic factors (Section 7.2.4.4, 7.3.1, very high confidence). Strong geographical differences in
44 heat-related mortality are projected to emerge later this century, mainly driven by growth in regions with
45 tropical and subtropical climates (Section 7.3.1, very high confidence)
46
47 In AR5 and SR15, the transition from undetectable to moderate risk was located below what were at the time
48 `recent' temperatures of between 0.5 to 0.8°C above pre-industrial levels, with medium to high confidence,
49 based on evidence of distributional impacts on crop production and water resources. New literature has
50 continued to confirm this transition has already taken place including more recent observed impacts for
51 regions and groups within the food and water sectors, strongly linked to Representative Key Risks for
52 Health, Water and Food Security (Section 16.2; 16.5; 5.4.1, 5.5.1, 5.8.1 and 5.12; Chapter 7).
53 In AR6, moderate risks have already been assessed to have occurred in Africa for economic growth and
54 reduced inequality, biodiversity and ecosystems, mortality and morbidity due to heat extremes and infectious
55 disease, and food production in fisheries and crop production (Figure 9.6). In Europe moderate risks to heat
56 stress, mortality and morbidity have already been reached, as well as for water scarcity in some regions
57 (Figure 13.30, Figure 13.33). In Australasia, moderate risks are assessed as present already for heat related
Do Not Cite, Quote or Distribute 16-106 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 mortality risk as well as cascading effects on cities and settlements; and also very high risks already present
2 in coral reef systems, and high risks to kelp forests and alpine biodiversity (Figure 11.7). In North America,
3 moderate risks have already been reached for freshwater scarcity, water quality (Figure 14.4), agriculture,
4 forestry, tourism, transport, energy & mining and construction (Figure 14.10).
5
6 For this assessment, the transition to moderate risk was assessed to have occurred between 0.7°C and 1.0°C
7 of warming with high confidence, demonstrating that a moderate level of risk exists at present. The 0.2°C
8 increase in this temperature range as compared with AR5 reflects the fact that AR6 WGI has assessed that
9 the level of global warming reached by 1986-2005 was 0.52-0.82°C (as opposed to 0.55-0.67°C in previous
10 assessments), and also reflects the opportunity for observations to be have made of the observed
11 consequences of the additional rise in temperature that has taken place since the literature underpinning the
12 AR5 assessment was published.
13
14 In AR5, the transition from moderate to high risk was assessed to occur between 1.6°C and 2.6°C above the
15 pre-industrial levels with medium confidence. In SR15, new literature on projected risks allowed this range
16 to be narrowed to 1.52°C. There is now substantial literature providing robust evidence of larger regional
17 risks at 2°C warming than 1.5°C and in a range of systems, including crop production (with risks of
18 simultaneous crop failure) (Thiault et al.; Gaupp et al., 2019), aquaculture and fisheries (Cheung et al.,
19 2018b; Froehlich et al., 2018; Stewart-Sinclair et al., 2020), nutrition-related health (Springmann et al., 2016;
20 Lloyd et al., 2018; Sulser et al., 2021), and exposure to stressors such as drought, floods (Alfieri et al., 2017;
21 Hirabayashi et al., 2021) and extreme heat (Dosio et al., 2018; Harrington et al., 2018; Sun et al., 2019). One
22 study (Gaupp et al., 2019) found that the risk of simultaneous crop failure in maize is estimated to increase
23 from 6% to 40% at 1.5 °C relative to the historical baseline climate. In particular, further research on
24 projected regional yield declines of wheat and maize between 1.5ºC and 2ºC, especially in Africa, has
25 accrued Asseng et al. (2015), including in Ethiopia (Abera et al., 2018) with associated economic effects
26 (Wang et al., 2019). Optimum maize production areas in E Asia are projected to reduce in area by 38% for
27 global warming of 1.52.0°C (He et al., 2019). A study of Jamaica also estimated that warming of less than
28 1.5°C will have an overall negative impact on crop suitability and a general reduction in the range of crops,
29 but above 1.5°C, irreversible changes to Jamaica's agriculture sector were projected (Rhiney et al., 2018).
30
31 Projections of increasing flood risk associated with global warming of 1.5 and 2°C continue to highlight
32 regional disparities, with larger than average increases projected in Asia and Africa (Hirabayashi et al.,
33 2021), including in China, India and Bangladesh (Alfieri et al., 2017). Similarly, nearly 80% of the 8-80
34 million additional people projected to be at risk of hunger owing to climate change are located in Africa and
35 Asia (Springmann et al., 2016; Lloyd and Oreskes, 2018; Nelson et al., 2018). Schleussner et al. (2016b)
36 analysed hotspots of multi-sectoral risks with 1.5°C and especially 2°C warming, highlighted projected crop
37 yield reductions in West Africa, South-East Asia, as well as Central and northern South America; a reduction
38 in water availability in the Mediterranean; and widespread bleaching of tropical coral reefs.
39
40 High risks to crop production are assessed to occur in Africa ~1.5-2ºC warming (Figure 9.6), to agriculture
41 in North America for ~1.5ºC warming (Figure 14.10), and ~ 2.8ºC Europe (Figure 13.30). High risks of
42 mortality and morbidity due to heat extremes and infectious disease are assessed to be reached in Africa with
43 ~1.5ºC warming (Figure 9.6); heat stress, mortality and morbidity in Europe is assessed to reach a high level
44 of risk at ~2ºC (Figure 13.30). Heat related mortality risk transitions to a high level by ~1.5-2ºC warming in
45 Australasia while cascading effects on cities reach high risk with ~1.2ºC warming (Figure 11.7). Risks to
46 water scarcity, forestry, tourism and transportation in N America are projected to reach high levels with ~2ºC
47 warming (Figure 14.4, Figure 14.10).
48
49 Two complementary multi-sectoral analyses indicates that South Asia and Africa become hotspots of multi-
50 sectoral climate change risk, largely due to changes in water related indicators which also affect crop
51 production (Arnell et al., 2018; Byers et al., 2018). For instance, Byers et al. (2018) found that the doubling
52 in global exposure to multi-sector risks that accrues as warming increases from 1.5 to 2°C is concentrated in
53 Asian and African regions (especially East Africa), which together account for 85-95% of the global
54 exposure.
55 Considering this evidence, for this assessment, the temperature range for the transition from moderate to
56 high risk is located between 1.5°C to 2°C above pre-industrial levels, with high confidence in the lower
Do Not Cite, Quote or Distribute 16-107 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 bound of 1.5°C, but medium confidence in the upper bound of 2°C, because simulation studies do not
2 account for climate variability and therefore risks could be higher.
3
4 Very high risk implies limited ability to adapt. Adaptation potential not only differs across sectors and
5 regions, but also occurs on different timescales depending on the nature and implementation level of the
6 adaptation option under consideration and the system in which it is to be deployed. The costs of adaptation
7 actions that would be needed to offset projected climate change impacts for major crop production are
8 projected to rise once global warming reaches 1.5 °C (Iizumi et al., 2020). It has been estimated that the
9 number of additional people at risk of hunger with 2.0 °C global warming could be reduced from 40 million
10 to 30 million by raising the level of adaptation action (Baldos and Hertel, 2014) but beyond this level of
11 warming residual impacts are projected to escalate (Iizumi et al., 2020). Chapter 5 assessed the potential of
12 existing farm management practices to reduce yield losses, finding an average 8% loss reduction in mid-
13 century and 11% by end-century (Section 5.4.4.1), which is insufficient to offset the negative impacts from
14 climate change, particularly in currently warmer regions (5.4.3.2). The literature indicates that globally, crop
15 production may be sustained below 2.0 °C warming with adaptation, but negative impacts will prevail at 2.0
16 °C warming and above in currently warm regions (Section 5.4.4.1). Importantly, residual damage (that which
17 cannot be avoided despite adaptation) is projected to rise around 2.0 °C global warming (Iizumi et al., 2020).
18 Evidence of constraints and limits for food, fiber and other ecosystem products for the different regions is
19 evident for the various regions (16.4.3.1) indicating limited ability to adapt. Adaptation costs are also higher
20 relative to GDP in low-income countries, for example for the building of sea-dikes (Brown et al., 2021).
21
22 In previous reports, the transition from high to very high risk for the distribution of impacts was not assessed
23 due to limited available literature, but there is now sufficient evidence to do so. A range of literature
24 quantifies the increasing regional probability of drought as compared to the present day, with projected
25 increases in the area exposed to drought (Carrão et al., 2018; Pokhrel et al., 2021), as well as the duration
26 (Naumann et al., 2018) and frequency of droughts with higher warming levels. Naumann et al. (2018)
27 showed that, for drying areas, drought durations are projected to rise from 2 months/°C below 1.5 °C to 4.2
28 months/°C near 3°C warming. Most of Africa, Australia, southern Europe, southern and central United
29 States, Central America, the Caribbean, north-west China, and parts of Southern America are projected to
30 experience more frequent droughts. Adverse effects of climate change on food production are projected to
31 become much more severe (Section 5.4.3.2) when global temperatures rise more than 2ºC globally but there
32 are predicted to be much more negative impacts experienced sooner on food security in low- to mid-latitudes
33 (Richardson et al., 2018a) (Sections 5.4.1). For instance, climate change by 2050 is projected to increase the
34 number of people at risk of hunger by between 8 and 80 million with 23°C warming compared to no climate
35 change conditions (Baldos and Hertel, 2014; Hasegawa et al., 2018; Nelson et al., 2018; Janssens et al., 2020). In
36 addition to effects upon crop yield, agricultural labour productivity, and food access, and food-related health are
37 projected to be negatively impacted by 23°C warming (Springmann et al., 2016; de Lima et al., 2021).
38 Regionally, substantial regional disparity in risks to food production is projected to persist at these higher
39 levels of warming. Risks for heat-related morbidity and mortality, ozone-related mortality, malaria, dengue,
40 Lyme disease, and West Nile fever are projected to increase regionally and globally (Chapter 7) with
41 potential infestation areas for disease-carrying vectors in multiple geographic regions that could be five times
42 higher at 4ºC than at 2ºC (Liu-Helmersson et al., 2019).
43
44 Very high risks to crop production are assessed to occur in Africa above ~2.5ºC warming (Figure 9.6) and
45 below 4ºC in Europe (Figure 13.30). Very high risks of mortality and morbidity due to heat extremes and
46 infectious disease are assessed to occur in Africa with 2.5ºC warming (Figure 9.6); heat stress, mortality and
47 morbidity in Europe is assessed to reach a very high level of risk at ~3.2ºC (Figure 13.30). Heat related
48 mortality risk and cascading effects on cities both transitions to a very high level by ~2.5C warming in
49 Australasia (Figure 11.7). Risks to water scarcity in N America are projected to reach very high levels with
50 3.5C warming (Figure 14.4). Hence this assessment concludes with medium confidence that a transition from
51 high to very high risks, in terms of distribution of impacts, begins at 2ºC global warming, with a full
52 transition to very high risks completed by 3.5ºC. However, it should be noted that many studies upon which
53 this assessment has been based have not taken into account the impacts of extreme weather events and
54 oscillations in sea surface temperatures hence risks at a given level of global warming might be
55 underestimated in the literature.
56
Do Not Cite, Quote or Distribute 16-108 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 16.6.3.4 Global Aggregate Impacts (RFC4)
2
3 This RFC considers impacts to socio-ecological systems that can be aggregated globally into a single metric,
4 such as monetary damages, lives affected, species lost or ecosystem degradation at a global scale
5 (Oppenheimer et al., 2014; O'Neill et al., 2017). RFC4 shares underlying key risk components with other
6 RFCs (e.g., RFC1 and RFC2, see O'Neill et al., 2017) and thus draws on a similar literature as those
7 assessments; however, this RFC focuses on impacts that reach levels of concern at the global level and also
8 weighs the composite effect of risk elements ranging from economic to biodiversity.
9
10 In AR5 Section 19.6.3.5 (Oppenheimer et al., 2014), the transition from undetectable to moderate risk was
11 assessed between 1.6 and 2.6ºC above pre-industrial levels (i.e., 1ºC and 2ºC above the 1986-2005 level)
12 based on impacts to both Earth's biodiversity and the overall global economy with medium confidence. The
13 risk transition between moderate and high risk was set around 3.6ºC above pre-industrial levels (i.e., 3ºC
14 above the 1986-2005 level), based on literature finding extensive species vulnerability and biodiversity
15 damage with associated loss of ecosystem goods and services at 3.5ºC (Foden et al., 2013; Warren et al.,
16 2013). In SR15 Section 3.5.2.4 (Hoegh-Guldberg et al., 2018b), economic literature on potential socio-
17 economic threshold events as well as empirical studies of global economic damages, combined with new
18 evidence on biome shifts, extinction risk, species range loss (especially noting the integral role of insects in
19 ecosystem function), and ecosystem degradation, were assessed and the upper bound of the transition to
20 moderate risk was lowered to 1.5ºC warming above pre-industrial levels, and the transition from moderate
21 and high risk was lowered to between 1.5ºC and 2.5ºC (medium confidence). The boundary between high
22 risk and very high risk was not assessed in either of these reports because the temperature threshold was
23 beyond the scope of the assessment in the case of SR15 and due to the limited literature available for this
24 highest transition in AR5.
25
26 Since AR5, many new global estimates of the aggregate, economy-wide risks of climate change have been
27 produced, though, as was the case in AR5, these continue to exhibit a low level of agreement, including for
28 today's level of global warming, due primarily to differences in methods. Cross-Working Group Box
29 ECONOMIC in this chapter includes a more thorough discussion of advancements and limitations of global
30 economic impact estimates and methodologies, finding significant variation in estimates that increases with
31 warming, indicating higher risk in terms of economic costs at higher temperatures (high confidence). Climate
32 change has been found to exacerbate poverty through declines in agricultural productivity, changes in
33 agricultural prices and extreme weather events (Hertel and Lobell, 2014; Hallegatte and Rozenberg, 2017).
34 In terms of biodiversity risks, the literature indicates that losses in terrestrial and marine ecosystems increase
35 substantially between 1.5ºC and 2ºC of warming (Hoegh-Guldberg et al., 2018b). Since SR15, further
36 evidence of degradation of biodiversity and ecosystem services and ocean acidification at the global
37 aggregate level has continued to accrue due to climate change (see Chapter 2).
38
39 For this RFC, the transition from undetectable to moderate risk to global aggregate impacts is assessed with
40 medium confidence to occur between 1.0ºC (start of transition) and 1.5ºC (completion of transition) with a
41 median judgment of transition at 1.3ºC, based on evidence of a combination of economic consequences,
42 widespread impacts to climate-sensitive livelihoods, changes in biomes and loss of terrestrial and marine
43 biodiversity. The start of the transition from undetectable to moderate risk is located at recent temperatures
44 based on observed impacts to biodiversity (16.2.3.1). Experts noted aggregate impacts on biodiversity are
45 detectable, with damages that have had global significance (e.g., drought, pine bark beetles, coral reef
46 ecosystems). Consistent with the start of this transition at 1ºC, a similar elicitation conducted in Chapter 2
47 assessed that risks to biodiversity globally have already transitioned to a moderate level with 1ºC warming;
48 whilst risks of widespread tree mortality are already moderate with 0.9ºC warming and finds that moderate
49 risks of ecosystem structure change began with warming of 0.5ºC (Table 2.S.4, Figure 2.11). Human-
50 induced warming has slowed growth of agricultural productivity over the past 50 years in mid- and low-
51 latitudes (Chapter 5; Hurlbert et al., 2019). Although there is not yet strong evidence of attributable loss of
52 life and livelihoods at the global level (16.5.2.3.4, 16.5.2.3.5), experts found that regional evidence of such
53 observed impacts were still relevant to defining the beginning of the transition (e.g., Table SM16.22, Chapter
54 9). Informing the median value and upper bound of the transition to moderate risk, empirical studies and
55 scenario analyses have found that regions with high dependence on climate-sensitive livelihoods like
56 agriculture, fisheries and forestry would be severely impacted even at low levels of warming under
57 conditions of low adaptation (RKR-D, Lobell et al., 2011; Hoegh-Guldberg et al., 2018b).
Do Not Cite, Quote or Distribute 16-109 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1
2 The transition to high risk is assessed with medium confidence to occur between 1.5ºC (start of transition)
3 and 2.5ºC (completion of transition) with a median judgment of transition at 2.0ºC. Though economic
4 estimates exhibit wide variation and low agreement at warming levels above 1.5ºC, many estimates are
5 nonlinear with marginal economic impacts increasing with temperature (see Cross-Working Group Box
6 ECONOMIC in this Chapter). At 1.5ºC warming, most aggregate global impacts to Gross Domestic Product
7 are negative across different estimation methods, including bottom-up estimation (e.g., Takakura et al.,
8 2019), meta-analysis (e.g., Howard and Sterner, 2017) and empirical estimations (e.g., Pretis et al., 2018;
9 Kalkuhl and Wenz, 2020). At 2°C Watts et al. (2021) estimate a relative decrease in effective labour by 10%,
10 which would have profound economic consequences. Byers et al. (2018) found that global exposure to multi-
11 sector risks approximately doubles between 1.5°C and 2°C, whilst the percentage of the global population
12 exposed to flooding is projected to rise by 24% with 1.5°C warming and by 30% with 2.0°C warning
13 (Hirabayashi et al., 2021).
14
15 Section 16.5.2.3.4 (RKR-D, underlying key risk on poverty) reports that under medium warming pathways,
16 climate change risks to poverty would become severe if vulnerability is high and adaptation is low (limited
17 evidence, high agreement). At and beyond 1.5ºC, approximately 200 million people with livelihoods derived
18 from small-scale fisheries would face severe risk, given sensitivity to ocean warming, acidification, and coral
19 reef loss (Cheung et al., 2018a; Froehlich et al., 2018; Free et al., 2019). Warming between 1.5 and 2ºC
20 could expose 330396 million people to lower agricultural yields and associated livelihood impacts (Byers et
21 al., 2018; Hoegh-Guldberg et al., 2018a), due to a high dependency of climate-sensitive livelihoods to
22 agriculture globally (World Bank, 2020). Models project that climate change will increase the number of
23 people at risk of hunger in 2050 by 8-80 million people globally, with the range depending on the level of
24 warming (1.52.9ºC) and SSPs (Nelson et al., 2018; Mbow et al., 2019; Janssens et al., 2020). Higher
25 atmospheric concentrations of carbon dioxide reduce the nutritional quality of wheat, rice, and other major
26 crops, potentially affecting millions of people at a doubling of carbon dioxide relative to pre-industrial (very
27 high confidence) (Section 7.3.1). Global ocean animal biomass is projected to decrease on average by 5% per
28 1°C increase, hence a 2.5C level of warming is associated with ~13% decline in ocean animal biomass,
29 which would considerably reduce marine food provisioning, fisheries distribution and revenue value, with
30 further consequences for ecosystem functioning (Chapter 5, medium confidence).
31
32 Losses in terrestrial and marine biodiversity increase substantially beyond 1.5ºC of warming (Hoegh-
33 Guldberg et al., 2018b). Section 16.5.2.3.2 (RKR-B, risks to terrestrial and marine ecosystems) finds that
34 substantial biodiversity loss globally, abrupt local ecosystem mortality impacts, and ecological species
35 disruption are all projected at global warming levels below 3ºC, with insular systems and biodiversity
36 hotspots at risk below 2ºC (medium confidence). Insects play a critical role in providing vital ecosystem
37 services that underpin human systems, with major losses of their climatically determined geographic range at
38 2°C warming implying adverse effects on ecosystem functioning. Consistent with the transitions presented
39 here, a similar burning ember developed in Chapter 2 assessed a transition from moderate to high risks
40 globally for marine and terrestrial biodiversity (e.g., widespread death of trees, damages to ecosystems, and
41 reduced provision of ecosystem services, and structural change, including biome shifts) beginning between
42 1.0 and 2.0ºC warming (Table 2.S.4, Figure 2.11).
43
44 Though explicit treatment of adaptation is limited in the RFC4 impacts literature (i.e., studies that compare
45 risks for specific adaptation scenarios in terms of globally aggregated impacts with quantified findings),
46 there is evidence of the potential for investments in improved hydro-meteorological information and early
47 warning systems to avoid some of the most adverse social and economic impacts from extreme weather
48 events in both developed and developing countries, with benefits at a globally significant level (Hallegatte,
49 2012). Studies of adaptation in the agriculture sector (e.g., changing crop variety, timing of crop planting,
50 new types of irrigation, etc.) and infrastructure (e.g., coastal protection, hardening of critical infrastructure,
51 flood and climate resistant building materials and water storage) show large potential benefits in terms of
52 reduced impacts to lives and livelihoods (van Hooff et al., 2015; Mees, 2017). At higher warming levels,
53 however, potential adaptations to address biodiversity loss are expected to be limited due to the projected
54 rate and magnitude of change as well as the resources required (Hannah et al., 2020).
55
56 The transition to very high risks is assessed to occur between a range of 2.54.5ºC with medium confidence
57 over the range, and low confidence assessed over a narrowed `best estimate' range of 2.73.7ºC. The lower
Do Not Cite, Quote or Distribute 16-110 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 end of the range reflects the loss of an increasingly large fraction of biodiversity globally. Chapter 2 has
2 assessed a transition from high to very high risks globally for biodiversity (marine and terrestrial) completing
3 at ~2.5ºC warming, noting widespread death of trees, damages to ecosystems, and reduced provision of
4 ecosystem services over the temperature range 2.5ºC4.5ºC (Table 2.S.4, Figure 2.11); and similarly a
5 transition from high to very high risks of ecosystem structure change (including biome shifts) between 3ºC
6 and 5ºC warming (Table 2.S.4, Figure 2.11). A global study of 115,000 common species projects
7 climatically determined geographic range losses of over 50% in 49% of insects, 44% of plants and 26% of
8 vertebrates with global warming of 3.2ºC, implying an associated effect on provisional and regulating
9 ecosystem services that support human wellbeing, including pollination and detrivory (Warren et al., 2018a).
10 The risk of abrupt impacts on ecosystems as multiple species approach tolerance limits simultaneously is
11 projected to threaten up to 15% of ecological communities with 4ºC of warming (Trisos et al., 2020). Under
12 a 4ºC warming scenario, models project global annual damages associated with sea level rise of $31,000
13 billion per year in 2100 (Brown et al., 2021)
14
15 In terms of global economic impact, while an emerging economic literature is addressing many gaps and
16 critiques of previous damage estimates for high warming (e.g., Jensen and Traeger, 2014; Burke et al., 2015;
17 Lontzek et al., 2015; Moore and Diaz, 2015; Lemoine and Traeger, 2016; Moore et al., 2017a; Cai and
18 Lontzek; Takakura et al., 2019, discussed further in Cross-Working Group Box ECONOMIC; Carleton et al.,
19 2020; Méjean et al., 2020; Rode et al., 2021), there remains wide variation across disparate methodologies,
20 though the spread of estimates increases with warming in all methodologies, indicating higher risk in terms
21 of economic costs at higher temperatures (high confidence). Section 16.5.2.3.4 (RKR-D) finds that risks to
22 aggregate economic output would become severe at the global scale at high warming (~4.4ºC) and minimal
23 adaptation (medium confidence), defining severity as `the potential for persistent annual economic losses due
24 to climate change to match or exceed losses during the world's worst historical economic recessions'.
25 Furthermore, climate change impacts on income inequality could compound risks to living standards (high
26 confidence, 16.5.2.3.4). Chapter 4 finds that at 4°C, 4 billion people are projected to be exposed to physical
27 water scarcity (medium confidence).
28
29
30 [START CROSS-WORKING GROUP BOX ECONOMIC HERE]
31
32 Cross-Working Group Box ECONOMIC: Estimating Global Economic Impacts from Climate Change
33
34 Authors: Steven Rose (USA), Delavane Diaz (USA), Tamma Carleton (USA), Laurent Drouet (Italy), Celine Guivarch
35 (France), Aurélie Méjean (France), Franziska Piontek (Germany)
36
37 This Cross-Working Group Box assesses literature estimating the potential global aggregate economic costs
38 of climate change and the social cost of carbon (SCC), where the former are sometimes referred to as
39 estimates of global `climate damages' and the latter are estimates of the potential monetized impacts to
40 society of an additional metric ton of carbon dioxide emitted to the atmosphere. These measures include the
41 economic costs of climate change that could be felt in market sectors such as agriculture, energy services,
42 labour productivity, and coastal resources, as well as non-market impacts such as other types of human
43 health risks (including mortality effects) and ecosystems. Global economic impacts estimates can inform
44 decisions about global climate management strategy, while SCC estimates can inform globally incremental
45 emissions decisions. In practice, economic damage estimates have been used to explore economically
46 efficient (`economically optimal') global emissions pathways (e.g.', Nordhaus and Moffat, 2017), while
47 SCCs have been used to inform federal and state-level policy assessment in some countries (Greenstone et
48 al., 2013; Rose and Bistline, 2016), but the type of SCC and application matter (Rose, 2017). This literature
49 has been assessed in previous WGII reports (e.g.., Arent et al., 2014) and this box serves this need for this
50 report. The assessment in this box was performed jointly across WGII and WGIII, building on the foundation
51 of WGII AR6 Chapter 16's `Risk to living standards' assessment (Section 16.5.2.3.4), which includes
52 consideration of severe risks to global aggregate economic output, and WGIII AR6 Chapter 3's assessment
53 of the benefits of mitigation. It also informs Chapter 16's global aggregate impacts Reason for Concern and
54 supports Chapter 18's assessment of global emissions transitions, risk management, and climate-resilient
55 development. In keeping with the broad risk framing presented in Chapter 1 of this report, other lines of
56 evidence regarding climate risks, beyond monetary estimates, should be considered in decision-making,
57 including key risks and Reasons for Concern.
Do Not Cite, Quote or Distribute 16-111 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1
2 Methods for estimating global economic costs of climate impacts
3
4 There are several broad approaches to estimating climate damages, including biophysical process models,
5 structural economic models, statistical methods (also called empirical or econometric) and hybrid
6 approaches, with each methodology having strengths and weaknesses. Process models simulate physical,
7 natural science, and/or engineering processes and their response to climate variables, that are then monetized
8 (e.g., Anthoff and Tol, 2014; Sieg et al., 2019; Narita et al., 2020). Process approaches have the advantage of
9 being explicit and interpretable, though they can be computationally intensive; may omit relevant impact
10 channels, interactions, and market dynamics affecting valuation; and, often lack a rigorous empirical basis
11 for calibration (Fisher-Vanden et al.). Structural economic modelling represents climate impacts on inputs,
12 production, household consumption, aggregate investment, and markets for economic sectors and regional
13 economies (e.g., Reilly et al., 2007; Roson and Van der Mensbrugghe, 2012; Anthoff and Tol, 2014; Dellink
14 et al., 2019; Takakura et al., 2019), often using computable general equilibrium (CGE) frameworks.
15 Structural models can evaluate how market and non-market impacts might enter and transmit through
16 economies, and adaptation responses within input and output markets, consumer and investment choices, and
17 inter-regional trade (e.g., Darwin and Tol, 2001; Dellink et al., 2019; Takakura et al., 2019). Statistical
18 methods estimate economic impacts in a given sector (e.g., Auffhammer, 2018) or in aggregate (e.g., Dell et
19 al., 2014; Burke et al., 2015; Hsiang et al., 2017; Pretis et al., 2018; Kahn et al., 2019), inferred from
20 observed changes in economic factors, weather, and climate, with responses and net results constrained by
21 available data. Since AR5, hybrid approaches have taken different forms to integrate process, statistical
22 and/or structural methods, and represent a potentially promising means of leveraging the strengths of
23 different approaches (e.g., Moore and Diaz, 2015; and Hsiang et al., 2017; Moore et al., 2017a; Ricke et al.,
24 2018; Yumashev et al., 2019; Chen et al., 2020b). There is also a small literature that uses expert elicitation
25 to gather subjective assessments of climate risks and potential economic impacts (Nordhaus, 1994; IPCC,
26 2019a; Pindyck, 2019).
27
28 In addition to differences in methods, there are also differences in scope geographic, sectoral, and
29 temporal. Global estimates are frequently based on an aggregation of independent sector and/or regional
30 modelling and estimates; however, there are examples of estimates from global modelling that simulate
31 multiple types of climate impacts and their potential interactions within a single, coherent framework (e.g.,
32 Roson and Van der Mensbrugghe, 2012; Dellink et al., 2019; Takakura et al., 2019). Differences in scope
33 also represent strengths and weaknesses between the methodologies, with narrower scope allowing for more
34 detailed assessment, but missing potential interactions with the scope not covered (e.g., other geographic
35 areas, sectors, markets, or periods of time).
36
37 Comprehensive economic estimates are challenging to produce for many reasons, including complex
38 interactions among physical, natural, and social systems; pervasive climate, socio-economic, and system
39 response uncertainties; and the heterogeneous nature of climate impacts that vary across space and time.
40 Critiques and commentaries of global estimation methods (Pindyck, 2013; Stern, 2013; van den Bergh and
41 Botzen, 2015; Cropper et al., 2017; Diaz and Moore, 2017; Pindyck, 2017; Rose et al., 2017; Stoerk et al.,
42 2018; DeFries et al., 2019; Pezzey, 2019; Calel et al., 2020; Warner et al., 2020; EPRI, 2021; Grubb et al.,
43 2021; Newell et al., 2021) include, among other things, concerns about statistical methods estimating
44 weather but not climate relationships, making out-of-sample extrapolations, and model specification
45 uncertainty, concerns about the observational grounding of structural modelling, overall concerns about the
46 lack of adaptation consideration, as well as representation and evaluation of potential large-scale singular
47 events such as ice sheet destabilisation or biodiversity destruction, some questioning the ability to generate
48 robust estimates (i.e., estimates insensitive to reasonable alternative inputs and specifications), and general
49 concerns about methodological details, transparency, and justification.
50
51 Additional methodological challenges to address (see, for instance, EPRI, 2021; Piontek et al., 2021) include
52 how to capture and represent uncertainty and variability in potential damage responses for a given climate
53 and societal condition, combine estimates from different methods and sources (including aggregating
54 independent sectoral and regional results), assess sensitivity and evaluate robustness of estimates (including
55 sensitivity to model specification), capture interactions and spillovers between regions and sectors, estimate
56 societal welfare implications (versus GDP changes) of market and non-market impacts, consider
57 distributional effects, represent micro and macro adaptation processes (and adaptation costs), specify
Do Not Cite, Quote or Distribute 16-112 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 nongradual damages and non-linearities, and improve understanding of potential long-run economic growth
2 effects. Note that, the treatment of time preference, risk aversion, and equity considerations have important
3 welfare implications for the aggregation of both potential economic impacts and climate change mitigation
4 costs.
5
6 In addition to updated and new methods and estimates, newer literature has explored nongradual damages,
7 such as climatic and socioeconomic tipping points (Lontzek et al., 2015; Méjean et al., 2020), potential
8 damage to economic growth (e.g., Burke et al., 2015; Moore and Diaz, 2015), valuing uncertainty in
9 potential damages (Jensen and Traeger, 2014; Lemoine and Traeger, 2016; Cai and Lontzek), and
10 representing adaptation (Takakura et al., 2019; Carleton et al., 2020; Rode et al., 2021). Going forward, to
11 help advance science and decisions, a key research priority is to understand and evaluate methodological
12 strengths and weaknesses in damage estimation, and reconcile the differences affecting comparability in such
13 a way that it informs use of the different lines of evidence. This will require greater transparency and
14 assessment of details and assumptions in individual methods, communication and evaluation of alternatives
15 for specifying or calibrating climate damage functional representations with respect to climate and non-
16 climate drivers and potential non-linearities, including evaluating data sufficiency for levels within and
17 beyond observations and for characterizing physical system dynamics, and evaluating the sensitivity of
18 results to model specification and input parameter choices (Cropper et al., 2017). Improving the robustness
19 of economic impact estimates is an active area of research. Below we describe the latest estimates.
20
21 Global estimates of the economic costs of climate impacts
22
23 Since AR5, many new estimates of the global economic costs of climate change have been produced. Figure
24 Cross-Working Group Box ECONOMIC.1 shows a wide spread of estimates, with growing variance at
25 higher levels of warming, both within and across methodology types (i.e., statistical, structural, or meta-
26 analysis). Meta-analysis is used here to refer to studies that treat other studies' estimates as data points in an
27 attempt to derive a synthesized functional form.
28
29 Global aggregate economic impact estimates (Figure Cross-Working Group Box ECONOMIC.1) are
30 generally found to increase with global average temperature change, as well as vary by other drivers, such as
31 income and population and the composition of the economy. Most estimates are nonlinear with higher
32 marginal economic impacts at higher temperature, although some recover declining marginal economic
33 impacts and functional forms cannot be determined for all studies. The drivers of non-linearity found in
34 economic impact estimates, and the differences in non-linearity across estimates (e.g., convex versus
35 concave, degree of curvature), are not well understood, with methodology construction, assumptions, and
36 data all potential factors. Relative to AR5, there have been more estimates and greater variation in estimates,
37 including some recent estimates significantly higher than the range reported in AR5. For most of the studies
38 shown in Figure Cross-Working Group Box ECONOMIC.1, the visible variation within a study represents
39 alternative socioeconomic projections and climate modelling, not economic impacts response uncertainty for
40 a given socioeconomic and climate condition. Response uncertainty could be significant as indicated by
41 some of the results shown in the figure (e.g., Burke et al., 2015; Rose et al., 2017), but methodological
42 differences in how uncertainty is characterized (model specification, errors, and confidence intervals versus
43 distributions of results) limits comparability and assessment. Note that modeling factors between global
44 temperature change and the economic impact calculation, such as regional temperature pattern assumptions
45 or assumed sea level rise dynamics, can also impact calculated estimates (e.g., Warren et al, 2021 PAGE09
46 estimates versus those in Rose et al, 2017, Chen et al, 2020 PAGE-ICE estimates versus Burke et al, 2015).
47
48 From Figure Cross-Working Group Box ECONOMIC.1, we find a large span of damage estimates, even
49 without considering uncertainty/confidence in damage responses, including for today's level of warming
50 (about 1°C). There is also evidence that some regions benefit from low levels of warming, leading to net
51 benefits globally at these temperatures. The size of the span of estimates grows with global warming level,
52 with variation across statistical estimates larger than variation in structural estimates. The structural and meta
53 analyses estimates appear to be in closer agreement, but that outcome is contingent on the meta analyses data
54 considerations and approach. Meta analyses to date have not assessed the alternative methods and dealt with
55 the lack of comparability between methods.
56
57
Do Not Cite, Quote or Distribute 16-113 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 Figure Cross-Working Group Box ECONOMIC.1: Global aggregate economic impact estimates by
2 global warming level (annual % global GDP loss relative to GDP without additional climate change). Top
3 row panels present estimates by methodology type: (a) statistical modeling, (b) structural modeling, and (c)
4 meta analyses, with all estimates from a paper in the same colour and estimates from methodologies other
5 than that highlighted by the panel in grey for reference. Second row left panel (d) presents AR5 estimates.
6 Second row right panel (e) presents all estimates in one figure, with the same colors as panels (a-d) using
7 outlined dots for the statistical modelling estimates, solid dots for structural modelling estimates, and
8 triangles for meta analysis estimates. In all panels, lines represent functions, with dashed and dotted lines 5th
9 and 95th percentile functions from structural modelling. To avoid duplication, estimates from papers using
10 the economic impacts estimates or model formulations already represented in the figure are not included
11 (e.g., Diaz and Moore, 2017; Chen et al., 2020b; Glanemann et al., 2020; Warren et al., 2021). The exception
12 is Burke et al. (2018), with the different estimates shown representing variation across climate scenarios for a
13 given aggregate economic impacts specification from Burke et al. (2015) the `pooled, short run' statistical
14 specification. Results shown for the latter are estimates with the author's different statistical model
15 specifications (and a fixed climate scenario, SSP5). From top to bottom, the Burke et al. (2015) estimates are
16 for the `pooled, long run,' `differentiated, long run,' `pooled, short run' (authors' base case), and
17 `differentiated, short run' statistical specifications. For Howard and Sterner (2017), the authors' preferred
Do Not Cite, Quote or Distribute 16-114 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 function is shown. Overall, estimates shown in the figure can correspond to different future years, reflecting
2 different socioeconomic conditions and climate pathways to a global warming level. Global average
3 temperature change bars relative to the period 1850-1900 are shown below the economic cost estimates to
4 provide context to potential future warming. Shown are the WGI AR6 assessed best estimates and 90%
5 intervals for the illustrative emissions scenarios considered for the near term 2021-2040, mid-term 2041-
6 2060, and long term 20812100.
7
8
9 Differences in methodology type and scope complicate comparison, assessment, and synthesis (Cropper et
10 al., 2017; Diaz and Moore, 2017; EPRI, 2021; Piontek et al., 2021). In particular, structural economic
11 modelling and empirical aggregate output modelling are fundamentally different, which has been identified
12 as an issue affecting the comparability of results (Cropper et al., 2017). The different methodologies affect
13 outcomes, with global aggregate estimates based on statistical methodologies typically higher than those
14 from structural modelling (Figure Cross-Working Group Box ECONOMIC.1). This is, in part, due to the
15 relationships in observational data captured by statistical modelling, assumed persistence of impacts in
16 statistical modelling, broader adaptation responses in structural modelling, and differences in the
17 representation of future societies and how they might evolve, respond, and interact. Within statistical
18 modelling, results are also found to be very sensitive to the statistical model specification (e.g., Burke et al.,
19 2015; Newell et al., 2021). Within structural modelling, differences in representations of biophysical changes
20 and economic structural dynamics contribute to differences across structural estimates (e.g., Rose et al.,
21 2017).
22
23 The wide range of estimates, and the lack of comparability between methodologies, does not allow for
24 identification of a robust range of estimates with confidence (high confidence). Evaluating and reconciling
25 differences in methodologies is a research priority for facilitating use of the different lines of evidence (high
26 confidence). However, the existence of higher estimates than AR5 indicate that global aggregate economic
27 impacts could be higher than previously estimated (low confidence due to the lack of comparability across
28 methodologies and robustness of estimates).
29
30 While Figure Cross-Working Group Box ECONOMIC.1 summarizes global aggregate estimates, the
31 literature exhibits significant heterogeneity in regional economic impacts that are also sensitive to
32 methodology, model specification, and societal assumptions (with, for instance, larger estimates due to the
33 assumed size of society, but offsetting adaptive capacity improvements and adaptation responses). Regional
34 results illustrate the potential for overall net benefits in more temperate regions at lower levels of warming
35 with potential lower energy demand and comparative advantages in agricultural markets; however, at higher
36 levels of warming net losses are estimated. In addition, economic impacts for poorer households and poorer
37 countries represent a smaller share in aggregate quantifications expressed in GDP terms than their influence
38 on well-being or welfare (Byers et al., 2018; Hallegatte et al., 2020).
39
40 Social cost of carbon methods and estimates
41
42 The global economic impact estimates discussed in the previous section serve as a key input into the
43 calculation of the value of potential net damages caused by a marginal ton of carbon dioxide emissions, or
44 the SCC. To compute an SCC, damage estimates are commonly combined in a multi-century modelling
45 framework with socioeconomic and emissions projections, a physical model of the climate, including a sea-
46 level rise component, and assumptions about the discount rate, with current frameworks having highly
47 stylized representations of these components. Though we do not present quantitative estimates here, due to
48 the challenge of comparability, for economic impacts methodologies (as discussed above) as well as other
49 SCC estimation elements, large variations in SCC estimates are found in the literature assessed due to,
50 among other things, differences in modelling component representations, input and parameter assumptions,
51 considerations of uncertainty, and discounting, inflation, and emissions year (e.g., Tol, 2009; Tol, 2018;
52 Pezzey, 2019; Iese et al., 2021). There are also different `variants' of SCC estimates that differ conceptually,
53 and in magnitude, depending on the reference condition for evaluating the impact of a marginal metric ton--
54 is it being evaluated relative to a no-climate-policy baseline, an economically efficient pathway that weighs
55 the benefits and costs of emissions mitigation, or a pathway based on a particular climate policy or goal such
56 as 2°C or a concentration target (Rose et al., 2017)? The variant of SCC has implications for its applicability
57 to different policy contexts (Rose and Bistline, 2016).
Do Not Cite, Quote or Distribute 16-115 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1
2 In addition to the economic impacts methodological challenges discussed above with respect to aggregate
3 economic impact estimates, the additional components needed for SCC calculations give rise to a new set of
4 technical issues and critiques, including incorporation of uncertainties in the components beyond climate
5 damages, links between components, and discounting (van den Bergh and Botzen, 2015; Cropper et al.,
6 2017; Diaz and Moore, 2017; Pindyck, 2017; Rose et al., 2017; EPRI, 2021). For component-specific
7 discussions and assessment, see Cropper et al. (2017), Rose et al. (2017), and EPRI (2021).
8
9 Substantial progress has been made in recent years to better reflect complexities in the global economy, the
10 climate system, and their interaction. For example, recent studies have explored damages to natural capital
11 (Bastien-Olvera and Moore, 2021), the influence of imperfect substitutability between environmental
12 services and market goods (Sterner and Persson, 2008; Weitzman, 2012; Drupp and Hänsel, 2021), the
13 implications of heterogeneous climate change impacts across income groups (Dennig et al., 2015; EPRI,
14 2021; Errickson et al., 2021), the potential for persistent climate impacts to economic growth instead of
15 effects on levels of economic output (Dietz and Stern, 2015; Moore and Diaz, 2015; Ricke et al., 2018;
16 Kikstra et al., 2021; Newell et al., 2021), valuing the risks of climate tipping points (Cai and Lontzek, 2019;
17 Rising et al., 2020), valuing uncertainty under risk aversion (Jensen and Traeger, 2014; Lemoine and
18 Traeger, 2016), and modelling a distinction between intertemporal inequality aversion and risk aversion in
19 the social welfare utility function (Crost and Traeger, 2013; Jensen and Traeger, 2014; Daniel et al., 2015).
20 These new studies have, in general, raised estimates of the SCC (Crost and Traeger, 2013; Jensen and
21 Traeger, 2014; Gerlagh and Michielsen, 2015; Moore and Diaz, 2015; Faulwasser et al., 2018; Guivarch and
22 Pottier, 2018; Budolfson et al., 2019; Cai and Lontzek, 2019; Dietz and Venmans, 2019; Kalkuhl and Wenz,
23 2020), in some cases by an order of magnitude (Ricke et al., 2018). However, challenges persist in terms of
24 moving from conceptual to practical application, such as pinning down parameter specifications, modelling
25 specific mechanisms for impacts, and more fully representing adaptation.
26
27 Despite these scientific advances, SCC estimates vary widely in the literature. Technical issues with past and
28 current modelling (e.g.', Pezzey, 2019; Pindyck, 2019; EPRI, 2021) and the challenge of comparability
29 across methodologies imply that many estimates are not robust (high confidence). Also, as a result, the issue
30 of directional bias of past estimates remains unsettled. Better representation of uncertainty in methods can
31 improve robustness, while detailed methodology assessment and comparison will help define the relative
32 biases of methods (high confidence).
33
34 Application to decision-making
35
36 The literature has also assessed the application of aggregate economic impact cost and SCC estimates (Rose
37 and Bistline, 2016; Rose et al., 2017; Kaufman et al., 2020) and identified conceptual and technical issues
38 that need to be considered when using results to inform policy decisions. These issues include: accounting
39 for endogenous marginal benefits and socioeconomic conditions in evaluating policies with non-incremental
40 global emissions implications; consistency in assumptions and treatment of uncertainty across benefit and
41 cost calculations; fully accounting for the streams of both mitigation costs and benefits over time; avoiding
42 inefficiently valuing or pricing emissions more than once across policies and jurisdictions; and accounting
43 for emissions leakage to capture net climate implications. Furthermore, concerns about the robustness of
44 estimates have led some to recommend considering alternatives, such as using marginal mitigation cost
45 estimates based on modelling of policy goals instead of the SCC (e.g., Rose, 2012; Pezzey, 2019; Kaufman
46 et al., 2020), although this comes with its own set of assumptions and technical challenges.
47
48 [END CROSS-WORKING GROUP BOX ECONOMIC HERE]
49
50
51 16.6.3.5 Large-scale Singular Events (RFC5)
52
53 This RFC, large-scale singular events (sometimes called tipping points or critical thresholds), considers
54 abrupt, drastic, and sometimes irreversible changes in physical, ecological, or social systems in response to
55 smooth variations in driving forces (accompanied by natural variability) (Oppenheimer et al., 2014; O'Neill
56 et al., 2017). SR15 Section 3.5.2.5 presented four examples, including the cryosphere (West Antarctic ice
57 sheet, Greenland ice sheet), thermohaline circulation (slowdown of the Atlantic Meridional Overturning
Do Not Cite, Quote or Distribute 16-116 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 Circulation), the El NiñoSouthern Oscillation (ENSO) as a global mode of climate variability, and the role
2 of the Southern Ocean in the global carbon cycle (Hoegh-Guldberg et al., 2018b). Whilst most of the
3 literature assessed here focuses on the resultant changes to climate-related hazards such as sea level rise, in
4 this assessment evidence about the implications of accelerated sea level rise for human and natural systems is
5 also considered. If sea level rise is accelerated by ice sheet melt, the associated impacts are projected to
6 occur decades earlier than otherwise, directly affecting coastal systems including cities and settlements by
7 the sea (CCP2) and wetlands (Chapter 2). The associated disruption to ports is projected to severely
8 compromise global supply chains and maritime trade with local-global geo-political and economic
9 consequences. In order to compensate for this acceleration, adaptation would need to occur much faster and
10 at a much greater scale than otherwise, or indeed than has previously been observed (CCP2). The costs of
11 accommodating port growth and adapting to sea level rise amount to USD22-768 billion before 2050
12 globally (medium evidence, high agreement) (see Section 2.1; Section 2.2; Cross-Chapter Box SLR in
13 Chapter 3).
14
15 In AR5 Section 19.6.3.6 (Oppenheimer et al., 2014), the boundary between undetectable and moderate risk is
16 set at levels between 0.6 and 1.6ºC above pre-industrial levels (i.e., 0ºC and 1ºC above the 1986-2005 level)
17 with high confidence, based on emerging early warning signals of regime shifts in Arctic and warm water
18 coral reef systems. The risk transition boundary between moderate and high risk was set between 1.6 and
19 3.6ºC above pre-industrial levels (i.e., 1ºC and 3ºC above the 1986-2005 level), with medium confidence
20 based on projections of ice sheet loss, with faster increase between 1ºC and 2ºC than between 2ºC and 3ºC.
21 The literature available at the time did not allow AR5 to assess the boundary between high and very high
22 risk.
23
24 In SR15 Section 3.5.2.5 (Hoegh-Guldberg et al., 2018b), new assessments of the potential collapse of the
25 West Antarctic ice sheet (WAIS) initiated by marine ice sheet instability (MISI) resulted in lowering the
26 upper end of the transition from undetectable and moderate risk from 1.6ºC to 1ºC warming above pre-
27 industrial levels, and lowering the upper end of the transition from moderate to high risk to 2.5ºC. Although
28 SR15 did not produce embers beyond 2.5ºC, authors reported that the transition to very high risk was
29 assessed at lying above 5ºC in light of growing literature on ice sheet contributions to sea level rise.
30
31 AR6 provides new evidence that relates to the location of the transition from undetectable to moderate risk.
32 At the time of SR15, observations were suggesting that MISI might already be taking place in some parts of
33 the WAIS while AR5 supported assessment of an additional MISI contribution to sea-level rise of several
34 additional tenths of a metre over the next two centuries. Since SR15, new observations (WGI AR6 Section
35 9.4.2.1, Fox-Kemper et al., 2021) support the assessment of enhanced grounding line retreat and subsequent
36 mass loss through basal melt in various parts of Antarctica, and year 2100 sea-level projections for the
37 RCP8.5 scenario have increased by 10-12 cm owing to ice dynamics. However, the onset of MISI is driven
38 by ocean warming in specific locations (ice cavities beneath floating ice shelves) and the relation between
39 these ocean temperatures and global mean temperature is indirect and ambiguous. In addition, MISI implies
40 a self-sustaining instability in the absence of further forcing. Because forcing is still increasing, it cannot be
41 unambiguously assessed whether MISI is driving the observed retreat of grounding lines in the WAIS, or
42 whether this retreat is a purely forced response (and would stop if the warming stops), or is just a
43 manifestation of natural variability in upwelling of warmer waters on the Antarctic continental shelves and,
44 as a result is just a temporary effect. Consistent with SROCC, AR6 states with medium confidence that
45 sustained mass losses of several major glaciers in the Amundsen Sea Embayment (ASE) are compatible with
46 the onset of MISI, but that whether unstable WAIS retreat already has begun or is imminent remains a
47 critical uncertainty.
48
49 Whether associated with MISI or not, WGI AR6 (Fox-Kemper et al., 2021) now assesses with very high
50 confidence that mass loss from both the Antarctic (whether associated with MISI or not) and Greenland Ice
51 Sheets, is more than seven times higher over the period 2010-2016 than over the period 1992-1999 for
52 Greenland and four times higher for the same time-intervals for Antarctica. Given their multi-century
53 commitments to global sea level rise this reinforces the assessment of estimating the boundary between
54 undetectable and moderate risks for ice sheets to lie between 0.7ºC (the level of global warming in the 1990s
55 when melting began to accelerate) and 1ºC (as in SR15), with a median at 0.9ºC.
56
Do Not Cite, Quote or Distribute 16-117 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 In the Amazon forest, increases in tree mortality and a decline in the carbon sink are already reported
2 (Brienen et al., 2015; Hubau et al., 2020) and old-growth Amazon rainforest may have become a net carbon
3 source for the period 2010-2019 (Qin et al., 2021). Estimates which include land-use emissions indicate the
4 region may have become a net carbon source (Gatti et al., 2021). Fire activity is an important driver and both
5 bigger fires (Lizundia-Loiola et al., 2020) and longer fire season (Jolly et al., 2015) have been reported in
6 South America, although this is strongly linked to land-use and land-use change as well as climate (Kelley et
7 al., 2021), and indeed land use change may be a stronger driver of potential loss of the Amazon forest than
8 climate change. The risk of climate-change related loss of the Amazon forest is assessed already above
9 `undetectable' but has only emerged over the last few years, when global warming had reached 1ºC, and is
10 linked to land-use as well as GSAT levels. Chapter 2 has assessed ecosystem carbon loss from tipping points
11 in tropical forest and loss of Arctic permafrost, and finds a transition to moderate risk over the range 0.6 to
12 0.9C (medium confidence). Specifically, WGII AR6 Table 2.S.4 finds that `Primary tropical forest comprised
13 a net source of carbon to the atmosphere, 2001-2019 (emissions 0.6 Gt y-1, net 0.1 Gt y-1) (Harris et al.,
14 2021). Anthropogenic climate change has thawed Arctic permafrost (Guo et al., 2020), carbon emissions 1.7
15 ± 0.8 Gt y-1, 2003-2017 (Natali et al., 2019)`. This also supports the upper limit for this transition lying at
16 1ºC.
17
18 The potential global loss of an entire ecosystem type, coral reefs, is also considered a large-scale singular
19 event. In the 1990's when global warming was around 0.7ºC large scale coral reef bleaching also became
20 apparent (16.2.3.1), also supporting the lower boundary for this transition in respect of coral reefs.
21
22 Overall, given the above evidence on ice sheets, Amazon forest, and coral reefs, the transition from
23 undetectable to moderate risk is therefore assessed to occur between 0.7ºC and 1ºC warming with a median
24 of 0.9ºC with high confidence.
25
26 The transition from moderate to high risk is informed by an assessment of risks at higher levels of warming
27 than present. Nearly all climate models do show warmer temperatures around Antarctica in conjunction with
28 rising global mean temperature and all ice sheet models do show sustained mass loss from the WAIS after
29 temperature increase halts (thus implying MISI takes place) at various levels between 1.5ºC and 5ºC, and an
30 increasing fraction of ice sheet models shows additional sustained mass loss from the East Antarctic Ice
31 Sheet (EAIS) for peak warming between 2ºC and 4ºC, and all ice sheet models show mass loss for peak
32 warming higher than 4ºC. Therefore, we assess an increasing link between MISI, WAIS collapse and
33 Antarctic mass loss, for increasing temperature levels (high confidence).
34
35 There is high confidence in the existence of threshold behaviour of the Greenland Ice Sheet in a warmer
36 climate (WGI AR6 Ch 9, Fox-Kemper et al., 2021), however there is low agreement on the nature of the
37 thresholds and the associated tipping points. Similarly the likelihood for accelerated and irreversible mass
38 loss from Antarctica increases with increasing temperatures but thresholds cannot yet be unambiguously
39 identified. By the year 2100, sea-level projections (AR6 WG1 SPM Fig SPM 8) now range from 0.57 m
40 (0.37-0.85) for the SSP1-1.9 scenario to 1.35 m (1.02-1.89) for the SS5-8.5 scenario and become 1.99 m for
41 the latter scenario (1.02-4.83) in case of low-likelihood, high-impact outcomes resulting from ice sheet
42 instability, for which there is limited evidence. It should be noted that inclusion of such low-likelihood, high-
43 impact outcomes dominated by not-well understood processes affecting ice dynamics on the large icecaps of
44 Greenland, and in particular Antarctica, would also enhance the sea-level projections for other scenarios, but
45 to a lesser extent for increasingly weaker forcing. No quantitative assessment of their effect in other
46 scenarios than SSP5-8.5 yet exists as such simulations with ice-sheet models have not been carried out, or
47 only in a very limited amount.
48
49 It should be noted that ice sheets may take many centuries to respond, implying that risk levels increase over
50 time for the same warming level. Therefore we base judgments about risk transitions related to ice sheets
51 primarily on their implications for 2000-year commitments to sea level rise from sustained mass loss from
52 both ice sheets as projected by various ice sheet models, reaching 2.3-3.1 m at 1.5°C peak warming and 2-6
53 m at 2.0°C peak warming (WGI AR6 TS, Box TS.4 Figure 1, (Arias et al., 2021)). This is an important
54 feature of the approach to this RFC (i.e., it is not primarily focused on implications for the next 100-200
55 years). In addition, since the AR5, there is new evidence about the Last Interglacial (LIG), when global mean
56 temperature was about 0.5-1.5°C above the pre-industrial era. AR6 assesses that it is virtually certain that
57 sea-level was higher than today at that time, likely by 510 m (medium confidence) (B.5.4 WGI AR6
Do Not Cite, Quote or Distribute 16-118 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 SPM,(IPCC, 2021a)). Mid-Pliocene temperatures of 2.5°C (about 3 million years ago when global
2 temperatures were 2.5°C4°C higher) also provide evidence as an upper limit for the transition to high risk
3 associated with long-term equilibrium sea-level rise of 5-25 m (WGI AR6 SPM B.5.4). In 2300 projected
4 sea-level rise in an RCP8.5 or SSP5-8.5 scenario (consistent with a peak warming range of 4ºC-6ºC, varies
5 between 1.7-6.8 and 2.2-5.9m respectively (WGI AR6 TS Box TS.4, Arias et al., 2021)), and when
6 accounting for Marine Ice Cliff Instability taking place on Antarctica these numbers may increase to a range
7 of 9.5-16.2 m (WGI AR6 TS Box TS.4, Arias et al., 2021) ).
8
9 CMIP6 climate models project drying in the Amazon especially in June-July-August, irrespective of future
10 forcing scenario, but which increases with GSAT/higher scenarios (Lee et al., 2021). For higher GSAT
11 levels Burton et al. (2021) explore different forcing scenarios and found, regardless of scenario, burned area
12 increases markedly with GSAT. New understanding of the role of vegetation stomata will act to exacerbate
13 this drying (Richardson et al., 2018b). A transition to high risk of savannization for the Amazon alone was
14 assessed to lie between 1.5 and 3ºC with a median value at 2.0ºC. A mean temperature increase of 2ºC could
15 reduce Arctic permafrost area ~15% by 2100 (Comyn-Platt et al., 2018). Chapter 2 has assessed ecosystem
16 carbon loss from tipping points in tropical forest and loss of Arctic permafrost, and finds a transition from
17 moderate to high risk over the range 1.5 to 3ºC with a median of 2ºC (medium confidence, Table 2.S.4,
18 Figure 2.11). Its assessment of the transition from high to very high risk is located over the range 3ºC - 5ºC
19 (low confidence, Table 2.S.4, Figure 2.11) based on the potential for Amazon forest dieback between 4-5ºC
20 temperature increase above the pre-industrial period (Salazar and Nobre, 2010).
21
22 One of the criteria for locating a transition to very high risk is a limited ability to adapt. In natural systems
23 limiting warming to 1.5°C rather than 2°C would enhance the ability of coastal wetlands to adapt naturally to
24 sea level rise, since natural sedimentation rates more likely keep up with sea level rise (SR15, Hoegh-
25 Guldberg 2018). In human systems, there is medium confidence that technical limits will be reached for hard
26 protection to SLR beyond 2100 under high emissions scenarios, with limits associated with socio-economic
27 and governance issues reached before 2100 (CCP2).
28
29 We therefore estimate the boundary between moderate and high risk to lie between 1.5ºC and 2.5ºC, with a
30 median at 2.0ºC, with medium confidence based on projections for melting ice sheets and drying in the
31 Amazon. We also estimate the boundary between high and very high risk to lie between 2.5ºC and 4ºC, but
32 with low confidence due to uncertainties in the projections of sea level rise at higher levels of warming and
33 differences between levels of warming at which very high risks were assessed in different systems.
34
35 16.6.4 Summary
36
37 The updated Reasons for Concern (RFC) show that transitions between levels of risk are now assessed to
38 occur at lower levels of global warming than in previous assessments (high confidence), levels of confidence
39 in assigning transitions have generally increased, evidence on the potential for adaptation to adequately
40 address risks at different warming levels remains limited, and transitions from high to very high levels of risk
41 have been assessed for all five RFCs, compared to just two RFCs in AR5, together showing how literature
42 published since AR5 is informing us on our future climate risks.
43
44 In particular, risks to unique and threatened systems (RFC1) are now assessed to be already at a high
45 level today, as compared with a moderate level in previous assessments, and transition to a very high
46 level is assessed to occur beginning at 1.2ºC, passing through a median value of 1.5°C, and
47 completing the transition at 2.0ºC warming (high confidence).
48
49 Risks associated with extreme weather events (RFC2) are assessed to have begun to transition to a
50 high level already when global warming reached 1ºC, with that transition projected to complete for a
51 warming of 1.5ºC (high confidence). Newly in AR6, a transition between high and very high levels
52 of risk was assessed to lie at 2.0ºC warming for RFC2 (range 1.8- 2.5ºC).
53
54 For risks associated with the distribution of impacts (RFC3), there is now high confidence that a
55 transition to moderate risk has already occurred, and the transition to high risk is now projected to
56 occur between 1.52.0ºC warming with medium confidence. Furthermore, a transition from high to
Do Not Cite, Quote or Distribute 16-119 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 very high risk is provided for the first time in this AR6 assessment, between 2.03.5ºC warming
2 (medium confidence).
3
4 Global aggregate impacts (RFC4) are assessed to have begun to transition to a moderate level
5 already when global warming reached 1ºC, and are projected to transition to a high level with
6 warming of 1.5 - 2.5ºC (median 2ºC) with medium confidence. An assessment of a transition to very
7 high risk is provided for the first time in AR6, over the range 2.5 to 4.5ºC with low confidence.
8
9 Risks associated with large-scale singular events are assessed to have already completed
10 transitioning to moderate with 1ºC warming (high confidence), with a transition to high risk between
11 1.52.5ºC [median 2ºC] (medium confidence). An assessment of a transition to very high risk is
12 provided for the first time in AR6, over the range 2.54.5ºC with low confidence.
13
14 In summary, risks to unique and threatened systems (RFC1) are higher at recent and projected levels of
15 warming than assessed previously (very high confidence); risks associated with extreme weather events
16 (RFC2) are assessed comparably to AR5 and SR15 at recent and low levels of warming, but notably much
17 higher at projected warming above 1.8°C (medium confidence); risks associated with distribution of impacts
18 (RFC3) and global aggregate impacts (RFC4) are similar to SR15 and higher than AR5 above 2°C (medium
19 confidence); and those associated with large-scale singular events (RFC5) are similar to SR15 and higher at
20 both recent and projected warming than AR5 (medium confidence).
21
22 Limiting global warming to 1.5ºC would ensure risk levels remain moderate for RFC3, RFC4 and RFC5
23 (medium confidence) but risk for RFC2 would have transitioned to a high risk at 1.5ºC and RFC1 would be
24 well into the transition to very high risk (high confidence). Remaining below 2ºC warming (but above 1.5ºC)
25 would imply that risk for RFC3 through 5 would be transitioning to high, and risk for RFC1 and RFC2
26 would be transitioning to very high (high confidence). By 2.5ºC warming, RFC1 will be in very high risk
27 (high confidence) and all other RFCs will have begun their transitions to very high risk (medium confidence
28 for RFC2 and RFC3, low confidence for RFC4 and RFC5). These highest levels of risk are associated with
29 an irreversible component, such that some impacts would persist even were global temperatures to
30 subsequently decline in an `overshooting' scenario.
31
32 Lack of evidence on the potential for adaptation to adequately reduce risk is a critical gap in our ability to
33 assess global risk transitions at the RFC level, but not only. In some cases, such as RFC1, the widespread
34 nature and rapid speed of the escalating risks, in combination with limited ability to adapt means that
35 transitions to high risk may occur despite medium or even high levels of adaptation. Risks that are largely
36 natural and not widely mediated by human vulnerability, are thus less likely to have risk transitions that shift
37 under higher societal adaptation. Risk transitions that are mediated through human systems, such as
38 distribution impacts, for example, are more likely to shift in response to adaptation as impacts are strongly
39 mediated through vulnerability within human systems, but such a shift is difficult to quantify given
40 knowledge gaps in the literature (Section 16.3). However, in some circumstances, expanded global
41 adaptation could slow some of these transitions (low confidence); in the case of RFC2, RFC3 and RFC4, the
42 literature suggests that coordinated global adaptation could increase the global temperature at which risks
43 transition from moderate to high, for example the prevention of mortality associated with heat stress within
44 RFC2.
45
46 A higher level of adaptation, applied globally and effectively, could have larger benefits for several RFC,
47 either postponing the onset of a high level of risk until a higher level of warming is reached (and allowing
48 time for mitigation efforts) or allowing a system to survive a temporary overshoot of a lower temperature
49 threshold. Adaptations are likely to have significant potential to reduce risks (Magnan et al., 2021) in
50 particular for risks mediated through human systems. However, there is limited evidence available to assess
51 the extent to which current or potential adaptations are or would be adequate in reducing climate risks at
52 different levels of warming, and adaptation implications for risk transitions will be highly localized.
53 Pathways and opportunities for risk management and adaptation actions with transformational potential are
54 discussed in Chapter 17, together with enabling factors, governance frameworks, financing, success factors,
55 and monitoring and evaluation discussed in Chapter 18, supporting sustainable system transitions and
56 leading to options for climate resilient development pathways.
57
Do Not Cite, Quote or Distribute 16-120 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1
2 [START FAQ16.1 HERE]
3
4 FAQ16.1: What are key risks in relation to climate change?
5
6 A few clusters of key risks can be identified which have the potential to become particularly severe and pose
7 significant challenges for adaptation worldwide. These risks, therefore, deserve special attention. They
8 include risks to important resources such as food and water, risks to critical infrastructures, economies,
9 health and peace, as well as risks to threatened ecosystems and coastal areas.
10
11 The IPCC defines key risks related to climate change as potentially severe risks that are relevant to the
12 primary goal of the United Nations Framework Convention on Climate Change treaty to avoid `dangerous
13 human interference with the climate system', and whatever the scale considered (global to local). What
14 constitutes `dangerous' or `severe' risks is partly a value judgment and can therefore vary widely across
15 people, communities, or countries. However, the severity of risks also depends on criteria like the magnitude,
16 irreversibility, timing, likelihood of the impacts they describe, as well as the adaptive capacity of the affected
17 systems (species or societies). The Working Group II authors use these criteria in various ways to identify
18 those risks that could become especially large in the future due to the interaction of physical changes to the
19 climate system with vulnerable populations and ecosystems exposed to them. For example, some natural
20 systems may be at risk of collapsing, as is the case for warm water coral reefs by mid-century, even if global
21 warming is limited to +1.5°C. For human systems, severe risks can include increasing restriction of water
22 resources that are already being observed; mortality or economic damages that are large compared to
23 historical crises; or impacts on coastal systems from sea level rise and storms that could make some locations
24 uninhabitable.
25
26 More than 130 key risks across sectors and regions have been identified by the chapters of this report, which
27 have then been clustered into a set of 8 overarching risks, called representative key risks, which can occur
28 from global to local scales but are of potential significance for a wide diversity of regions and systems
29 globally. As shown in figure FAQ16.1, the representative key risks include risks to (1) low-lying coastal
30 areas, (2) terrestrial and marine ecosystems, (3) critical infrastructures and networks, (4) living standards, (5)
31 human health, (6) food security, (7) water security and (8) peace and mobility.
32
33 These representative key risks are expected to increase in the coming decades and will depend strongly not
34 only on how much climate change occurs, but also on how the exposure and vulnerability of society changes,
35 as well as on the extent to which adaptation efforts will be effective enough to substantially reduce the
36 magnitude of severe risks. The report finds that risks are highest when high warming combines with
37 development pathways with continued high levels of poverty and inequality, poor health systems, lack of
38 capacity to invest in infrastructure, and other characteristics making societies highly vulnerable. Some
39 regions already have high levels of exposure and vulnerability, such as in many developing countries as well
40 as communities in small islands, Arctic areas and high mountains; in these regions, even low levels of
41 warming will contribute to severe risks in the coming decades. Some risks in industrialized countries could
42 also become severe over the course of this century, for example if climate change affects critical
43 infrastructure such as transport hubs, power plants, or financial centres. In some cases such as coral reef
44 environments and areas already severely affected by intense extreme events (e.g. recent typhoons or
45 wildfires), for example, climate risks are already considered severe.
46
47
Do Not Cite, Quote or Distribute 16-121 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1
2 Figure FAQ16.1.1: Presentation of the 8 representative key risks assessed in this report (and their underlying main key
3 risks).
4
5 [END FAQ16.1 HERE]
6
7
8 [START FAQ16.2 HERE]
9
10 FAQ16.2: How does adaptation help to manage key risks and what are its limits?
11
12 Adaptation helps to manage key risks by reducing vulnerability or exposure to climate hazards. However,
13 constraining factors make it harder to plan or implement adaptation and result in adaptation limits beyond
14 which risks cannot be prevented. Limits to adaptation are already being experienced, for instance by coastal
15 communities, small-scale farmers and some natural systems.
16
17 Adaptation-related responses are actions that are taken with the intention of managing risks by reducing
18 vulnerability or exposure to climate hazards. While mitigation responses aim to reduce greenhouse gas
19 emissions and slow warming, adaptations respond to the impacts and risks that are unavoidable, either due to
20 past emissions or failure to reduce emissions. However, while these responses intend to reduce risks, it is
21 difficult to determine precise levels of risk reduction that can be attributed to adaptation. Changing levels of
22 risk as well as other actions --such as economic development -- make it challenging to definitively connect
23 specific levels of risk reduction with adaptation. Although it is not feasible to assess the adequacy of
24 adaptation for risk reduction at global or regional levels, evidence from specific localized adaptation projects
25 do show that adaptation-related responses reduce risk. Moreover, many adaptation measures offer near-term
26 co-benefits related to mitigation and to sustainable development, including enhancing food security and
27 reducing poverty.
28
29 Adaptation responses can occur in natural systems without the intervention of humans, such as species
30 shifting their range, time of breeding, or migration behaviour. Humans can also assist adaptation in natural
31 systems through, for example, conservation activities such as species regeneration projects or protecting
32 ecosystem services. Other adaptation-related responses by humans aim to reduce risk by decreasing
33 vulnerability and/or exposure of people to climate hazards. This includes infrastructural projects (e.g.
Do Not Cite, Quote or Distribute 16-122 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 upgrading water systems to improve flood control), technological innovation (e.g. early warning systems for
2 extreme events), behavioural change (e.g. shift to new crop types or livelihood strategies), cultural shifts
3 (e.g. changing perspectives on urban greenspace, or increased recognition of Indigendous Knowledge and
4 Local Knowledge), and institutional governance (e.g. adaptation planning, funding, and legislation).
5
6 While adaptation is important to reduce risk, adaptation cannot prevent all climate impacts from occurring.
7 Adaptation has soft and hard limits, points at which adaptive actions are unable to prevent risks. Soft limits
8 can change over time as additional adaptation options become available, while hard limits will not change as
9 there are no additional adaptive actions that are possible. Soft limits occur largely due to constraints-- factors
10 that make it harder to plan and implement adaptation, such as lack of financial resources or insufficient
11 human capacity. Across regions and sectors, the most challenging constraints to adaptation are financial and
12 those related to governance, institutions and policy measures. Limited funding and ineffective governance
13 structures make it difficult to plan and implement adaptation-related responses which can lead to insufficient
14 adaptation to prevent risks. Small-scale farmers and coastal communities are already facing soft limits to
15 adaptation as measures that they have put in place are not enough to prevent loss. If constraints that are
16 limiting adaptation are addressed, then additional adaptation can take place and these soft limits can be
17 overcome. Evidence on limits to adaptation is largely focused on terrestrial and aquatic species and
18 ecosystems, coastal communities, water security, agricultural production, and human health and heat.
19
20 Adaptation is critical for responding to unavoidable climate risks. Greater warming will mean more and
21 more severe impacts requiring a high level of adaptation which may face greater constraints and reach soft
22 and hard limits. At high levels of warming, it may not be possible to adapt to some severe impacts.
23
24
25 [END FAQ16.2 HERE]
26
27
28 [START FAQ16.3 HERE]
29
30 FAQ16.3: How do climate scientists differentiate between impacts of climate change and changes in
31 natural or human systems that occur for other reasons?
32
33 We can already observe many impacts of climate change today. The large body of climatic impact data and
34 research confirms this. To decide whether an observed change in a natural or human system is at least
35 partly an impact of climate change we systematically compare the observed situation to a theoretical
36 situation without observed levels of climate change. This is detection and attribution research.
37
38 Global mean temperature has already risen by more than 1°C and that also means that the impacts of climate
39 change become more visible. Many natural and human systems are sensitive to weather conditions. Crop
40 yields, river floods and associated damages, ecosystems such as coral reefs, or the extent of wildfires are
41 affected by temperatures and precipitation changes. Other factors also come into play. So for example, crop
42 yields around the world have increased over the last decades because of increasing fertilizer input, improved
43 management and varieties. How do we detect the effect of climate change itself on these systems, when the
44 other factors are excluded? This question is central for impact attribution. `Impact of climate change' is
45 defined as the difference between the observed state of the system (e.g., level of crop yields, damage induced
46 by a river flood, coral bleaching) and the state of the system assuming the same observed levels of non-
47 climate related drivers (e.g. fertilizer input, land use patterns, or settlement structures) but no climate change.
48 So:
49 `Impact of climate change' is defined as the difference between the observed state of the system and the state
50 of the system assuming the same observed levels of non-climate related drivers but no climate change. For
51 example, we can compare the level of crop yields, damage induced by a river flood, and coral bleaching with
52 differences in fertilizer input, land use patterns, or settlement structures, without climate change and with
53 climate change occurring.
54
55 While this definition is quite clear, there certainly is the problem that in real life, we do not have a `no
56 climate change world' to compare with. We use model simulations where the influence of climate change
57 can be eliminated to estimate what might have happened without climate change. In a situation where the
Do Not Cite, Quote or Distribute 16-123 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 influence of other non-climate related drivers is known to be minor (e.g., in very remote locations) the non-
2 climate change situation can also be approximated by observation from an early period where climate change
3 was still minor. Often a combination of different approaches increases our confidence in the quantification of
4 the impact of climate change.
5
6 Impacts of climate change have been identified in a wide range of natural, human, and managed systems. For
7 example, climate change is the major driver of observed widespread shifts in the timing of events in the
8 annual cycle of marine and terrestrial species, the extent of areas burned by wildfires is increased by climate
9 change in certain regions, it has increased heat-related mortality and had an impact on the expansion of
10 vector-borne diseases.
11 In some other cases research has made considerable progress in identifying the sensitivity of certain
12 processes to weather conditions without yet attributing observed changes to long-term climate change. Two
13 examples of weather sensitivity without attribution are observed crop price fluctuations and waterborne
14 diseases.
15
16 Finally it is important to note that `attribution to climate change' does not necessarily mean `attribution to
17 anthropogenic climate change'. Instead, according to the IPCC definition, climate change means any long
18 term change in the climate system no matter where it comes from.
19
20 [END FAQ16.3 HERE]
21
22
23 [START FAQ16.4 HERE]
24
25 FAQ16.4: What adaptation-related responses to climate change have already been observed, and do
26 they help reduce climate risk?
27
28 Adaptation-related responses are the actions taken with the intention of managing risks by reducing
29 vulnerability or exposure to climate hazards. Responses are increasing and expanding across global regions
30 and sectors, although there is still a lot of opportunity for improvement. Examining the adequacy and
31 effectiveness of the responses is important to guide, planning, implementation and expansion.
32
33 The most frequently reported adaptation-related responses are behavioural changes made by individuals and
34 households in response to drought, flooding, and rainfall variability in Africa and Asia. Governments are
35 increasingly undertaking planning, and implementing policy and legislation, including for example new
36 zoning regulations and building codes, coordination mechanisms, disaster and emergency planning, or
37 extension services to support farmer uptake of drought tolerant crops. Local governments are particularly
38 active in adaptation-related responses, particularly in protecting infrastructure and services, such as water
39 and sanitation. Across all regions, adaptation-related responses are strongly linked to food security, with
40 poverty alleviation a key strategy in the Global South.
41
42 Overall, however, the extent of adaptation-related responses globally is low. On average, responses tend to
43 be local, incremental, fragmented, and consistent with business-as-usual practices. There are no global
44 regions or sectors where the overall adaptation-related response has been rapid, widespread, substantial, and
45 has overcome or challenged key barriers. The extent of adaptation thus remains low globally, with
46 significant potential for increased scope, depth, speed, and the challenging of adaptation limits. Examples of
47 low extent adaptations include shifts by subsistence farmers in crop variety or timing, household flood
48 barriers to protect houses and gardens, and harvesting of water for home and farm use. In contrast, high
49 extent adaptation means that responses are widespread, coordinated, involve major shifts from normal
50 practices, are rapid, and challenge existing constraints to adaptation. Examples of high extent adaptations
51 include planned relocation of populations away from increasingly flood-prone areas, and widely
52 implemented social support to communities to prevent migration or displacement due to climate hazards.
53
54 Increasing the extent of adaptation-related responses will require more widespread implementation and
55 coordination, more novel and radical shifts from business-as-usual practices, more rapid transitions, and
56 challenging or surmounting limits -- key barriers -- to adaptation. This might include, for example, best-
57 practice programmes implemented in a few communities being expanded to a larger region or country,
Do Not Cite, Quote or Distribute 16-124 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 accelerated implementation of behaviours or regulatory frameworks, coordination mechanisms to support
2 deep structural reform within and across governments, and strategic planning that challenges fundamental
3 norms and underlying constraints to change.
4
5 We have very little information on whether existing adaptation-related responses that have already been
6 implemented are reducing climate risks. There is evidence that risks due to extreme heat and flooding have
7 declined, though it is not clear if these are due to specific adaptation-related responses or general and
8 incremental socio-economic development. It is difficult to assess the effectiveness of adaptation-related
9 responses, and even more difficult to know whether responses are adequate to adapt to rising climate risk.
10 These remain unknown but important questions in guiding implementation and expansion of adaptation-
11 related responses.
12
13 [END FAQ16.4 HERE]
14
15
16 [START FAQ16.5 HERE]
17
18 FAQ16.5: How does climate risk vary with temperature?
19
20 Climate risk is a complex issue and communicating and it is fraught with difficulties. Risk generally
21 increases with global warming, though it depends on a combination of many factors such as exposure,
22 vulnerability and response. To present scientific findings succinctly, a risk variation diagram can help
23 visualize the relationship between warming level and risk. The diagram can be useful in communicating the
24 change in risk with warming for different types of risk across sectors and regions, as well as for five
25 categories of global aggregate risk called `Reasons for Concern'.
26
27 A picture speaks a thousand words. The use of images to share ideas and information to convey scientific
28 understanding is an inclusive approach for communicating complex ideas. A risk variation diagram is a
29 simple way to present the risk levels that have been evaluated for any particular system. These diagrams take
30 the form of bar charts where each bar represents a different category of risk. The traffic light colour system is
31 used as a basis for doing the risks, making it universally understandable. These diagrams are known
32 colloquially as `burning ember' diagrams, and have been a cornerstone of IPCC assessments since the Third
33 Assessment Report, and further developed and updated in subsequent reports. The fact that the diagrams are
34 designed to be simple, intuitive, and easily understood with the caption alone, has contributed to their
35 longstanding effectiveness. Here, in Figure FAQ16.5.1 below, we provide a simplified figure of this
36 chapter's burning embers for five categories of global aggregate risk, called Reasons for Concern (RFC),
37 which collectively synthesize how global risk changes with temperature. The diagram shows the levels of
38 concern that scientists have about the consequences of climate change (for a specified risk category and
39 scope), and how this relates to the level of temperature rise.
40
41
42
43 Figure FAQ16.5.1: Simplified presentation of the five Reasons for Concern burning ember diagrams as assessed in this
44 report (adapted from Figure 16.15). The colours indicate the level of risk accrual with global warming for a low
Do Not Cite, Quote or Distribute 16-125 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 adaptation scenario. RFC1 Unique and threatened systems: ecological and human systems that have restricted
2 geographic ranges constrained by climate related conditions and have high endemism or other distinctive properties.
3 Examples include coral reefs, the Arctic and its indigenous people, mountain glaciers and biodiversity hotspots. RFC2
4 Extreme weather events: risks/impacts to human health, livelihoods, assets and ecosystems from extreme weather
5 events such as heatwaves, heavy rain, drought and associated wildfires, and coastal flooding. RFC3 Distribution of
6 impacts: risks/impacts that disproportionately affect particular groups due to uneven distribution of physical climate
7 change hazards, exposure or vulnerability. RFC4 Global aggregate impacts: impacts to socio-ecological systems that
8 can be aggregated globally into a single metric, such as monetary damages, lives affected, species lost or ecosystem
9 degradation at a global scale. RFC5 Large-scale singular events: relatively large, abrupt and sometimes irreversible
10 changes in systems caused by global warming, such as ice sheet disintegration or thermohaline circulation slowing.
11
12
13 In this diagram, the risk variation bars or embers are shown with temperature on the y-axis, and the base of
14 the ember corresponds to a baseline temperature. Typically this baseline temperature is that before global
15 warming started (i.e., average temperatures for the pre-industrial period of 1850 to 1900). This area of the
16 ember appears white, which indicates no to negligible impacts due to climate change. Moving up the ember
17 bar, changing colours show the increase in risk as the earth warms globally in terms of degrees Celsius
18 yellow for moderate risk, red for high risk, and purple for very high risk. Definitions of the risk levels are
19 presented in Figure FAQ16.5.1 The risk transitions are informed by the latest literature and scientific
20 evidence, and developed through consultation and development of consensus among experts. The bars depict
21 an averaged assessment across the world which has the disadvantage of hiding regional variation. For
22 example, some locations or regions could face high risk even when the global risk level is moderate.
23
24 When the embers for different risk categories are placed next to each other, it is possible to compare risk
25 levels at different levels of global warming. For example, at 1ºC warming all embers appear yellow or white,
26 so it is possible to say that keeping global warming below that particular temperature would help ensure risks
27 remain moderate for all five categories of concern assessed. In contrast, at 2ºC warming, risk levels have
28 transitioned to high for all categories assessed, and even reach a very high level of risk in the case of unique
29 and threatened systems.
30
31 [END FAQ16.5 HERE]
32
33
34 [START FAQ16.6 HERE]
35
36 FAQ16.6: What is the role of extreme weather events in the risks we face from climate change?
37
38 Climate change has often been perceived as a slow and gradual process but by now it is abundantly clear
39 that many of its impacts arise through shocks, such as extreme weather events. Many places are facing more
40 frequent and intense extremes, and also more surprises. The impact of such shocks is shaped by exposure
41 and vulnerability, where we live, and how we are prepared for and able to cope with shocks and surprises.
42
43 The rising risk of extreme events is one of the major reasons for concern about climate change. It is clear that
44 this risk has already increased today. Many recent disasters already have a fingerprint of climate change.
45
46 There are large differences in such risks from country to country, place to place, and person to person. This
47 is of course partly due to differences in hazards such as heatwaves, floods, droughts, storms, storm surges,
48 etc., and the way those hazards are influenced by climate change. However, an even more important aspect is
49 people's exposure and vulnerability: do these hazards occur in places where people live and work, and how
50 badly do they affect people's lives and livelihoods? Some groups are especially vulnerable, for instance
51 elderly in the case of heatwaves, or people with disabilities in the case of floods. In general, poor and
52 marginalised people tend to be much more affected than rich people, partly because they have less reserves
53 and support systems that help them to prepare for, cope with and recover from a shock. On the other hand,
54 absolute economic losses are generally higher in richer places, simply because more assets are at risk there.
55
56 Many problems caused by extreme weather do not just appear because of one weather extreme, but due to a
57 combination of several events. For instance, dryness may increase the risk of a subsequent heatwave. But the
58 increased risk may also cascade through human systems, for instance when several consecutive disasters
Do Not Cite, Quote or Distribute 16-126 Total pages: 173
FINAL DRAFT Chapter 16 IPCC WGII Sixth Assessment Report
1 erode people's savings, or when a heatwave reduces the ability of power plants to produce electricity, which
2 subsequently affects availability of electricity to turn on air conditioning to cope with the heat. Many shocks
3 also have impacts beyond the place where they occur, for instance when a failed harvest affects food prices
4 elsewhere. Climate risks can also be aggravated by other shocks, such as in the case of COVID-19, which
5 not only had a direct health impact, but also affected livelihoods around the world and left many people
6 much more vulnerable to weather extremes.
7
8 Understanding the risks we face can help in planning for the future. This may be a combination of short-term
9 preparation, such as early warning systems, and longer-term strategies to reduce vulnerability, for instance
10 through urban planning, as well as reducing greenhouse gases to avoid longer-term increases in risk. Many
11 interventions to increase people's resilience are effective in the face of a range of shocks. For instance, social
12 safety nets can help mitigate the impact of a drought on farmers' livelihoods, but also of the economic
13 impacts of COVID-19.
14
15 Climate-related shocks are threats to society, but they can also offer opportunities for learning and change.
16 Recent disasters can motivate action during a short window of opportunity when awareness of the risks is
17 higher and policy attention is focused on solutions to adapt and reduce risk. However, those windows tend to
18 be short, and attention is often directed at the event that was recently experienced, rather than resilience in
19 the face of a wider range of risks.
20
21 [END FAQ 16.6 HERE]
22
23
24
Do Not Cite, Quote or Distribute 16-127 Total pages: 173