Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 Table of Contents
2
3 Chapter 3: Mitigation Pathways Compatible with Long-Term Goals ........................................... 3-1
4 Executive summary.......................................................................................................................... 3-4
5 3.1 Introduction ........................................................................................................................ 3-10
6 3.1.1 Assessment of mitigation pathways and their compatibility with long-term goals ... 3-10
7 3.1.2 Linkages to other Chapters in the Report ................................................................... 3-10
8 3.1.3 Complementary use of large scenario ensembles and a limited set of Illustrative
9 Mitigation Pathways (IMPs) ...................................................................................................... 3-11
10 3.2 What are mitigation pathways compatible with long-term goals? ..................................... 3-12
11 3.2.1 Scenarios and emission pathways .............................................................................. 3-12
12 3.2.2 The utility of Integrated Assessment Models............................................................. 3-13
13 3.2.3 The scenario literature and scenario databases .......................................................... 3-15
14 3.2.4 The AR6 scenario database ........................................................................................ 3-16
15 3.2.5 Illustrative Mitigation Pathways ................................................................................ 3-19
16 3.3 Emission pathways, including socio-economic, carbon budget and climate responses
17 uncertainties ................................................................................................................................... 3-24
18 3.3.1 Socio-economic drivers of emissions scenarios ......................................................... 3-24
19 3.3.2 Emission pathways and temperature outcomes.......................................................... 3-26
20 Cross-Chapter Box 3: Understanding net zero CO2 and net zero GHG emissions .................... 3-37
21 3.3.3 Climate impacts on mitigation potential .................................................................... 3-49
22 3.4 Integrating sectoral analysis into systems transformations ................................................ 3-50
23 3.4.1 Cross-sector linkages ................................................................................................. 3-51
24 3.4.2 Energy supply ............................................................................................................ 3-57
25 3.4.3 Buildings .................................................................................................................... 3-59
26 3.4.4 Transport .................................................................................................................... 3-61
27 3.4.5 Industry ...................................................................................................................... 3-63
28 3.4.6 Agriculture, Forestry and Other Land Use (AFOLU) ................................................ 3-64
29 3.4.7 Other Carbon Dioxide Removal Options ................................................................... 3-67
30 3.5 Interaction between near-, medium- and long-term action in mitigation pathways ........... 3-68
31 3.5.1 Relationship between long-term climate goals and near- to medium-term emissions
32 reductions ................................................................................................................................... 3-69
33 3.5.2 Implications of near-term emission levels for keeping long-term climate goals within
34 reach 3-72
35 3.6 Economics of long-term mitigation and development pathways, including mitigation costs
36 and benefits .................................................................................................................................... 3-82
37 3.6.1 Economy-wide implications of mitigation ................................................................. 3-83
38 3.6.2 Economic benefits of avoiding climate changes impacts........................................... 3-91
39 3.6.3 Aggregate economic implication of mitigation co-benefits and trade-offs................ 3-95
40 3.6.4 Structural change, employment and distributional issues along mitigation pathways ... 3-
41 95
Do Not Cite, Quote or Distribute 3-2 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 3.7 Sustainable development, mitigation and avoided impacts ............................................... 3-97
2 3.7.1 Synthesis findings on mitigation and sustainable development ................................. 3-97
3 3.7.2 Food ......................................................................................................................... 3-102
4 3.7.3 Water ........................................................................................................................ 3-103
5 3.7.4 Energy ...................................................................................................................... 3-104
6 3.7.5 Health ....................................................................................................................... 3-105
7 3.7.6 Biodiversity (land and water) ................................................................................... 3-107
8 3.7.7 Cities and infrastructure ........................................................................................... 3-108
9 3.8 Feasibility of socio/techno/economic transitions ............................................................. 3-109
10 3.8.1 Feasibility frameworks for the low carbon transition and scenarios ........................ 3-109
11 3.8.2 Feasibility appraisal of low carbon scenarios .......................................................... 3-110
12 3.8.3 Feasibility in the light of socio-technical transitions ............................................... 3-114
13 3.8.4 Enabling factors ....................................................................................................... 3-115
14 3.9 Methods of assessment and gaps in knowledge and data ................................................ 3-116
15 3.9.1 AR6 mitigation pathways......................................................................................... 3-116
16 3.9.2 Models assessed in this chapter ............................................................................... 3-116
17 Frequently Asked Questions (FAQs) ........................................................................................... 3-117
18 References .................................................................................................................................... 3-119
19
20
21
Do Not Cite, Quote or Distribute 3-3 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 Executive summary
2 Chapter 3 assesses the emissions pathways literature in order to identify their key
3 characteristics (both in commonalities and differences) and to understand how societal choices
4 may steer the system into a particular direction (high confidence). More than 2000 quantitative
5 emissions pathways were submitted to the IPCC’s Sixth Assessment Report (AR6) database, out of
6 which 1202 scenarios included sufficient information for assessing the associated warming consistent
7 with WGI. Five Illustrative Mitigation Pathways (IMPs) were selected, each emphasizing a different
8 scenario element as its defining feature: heavy reliance on renewables (IMP-Ren), strong emphasis on
9 energy demand reductions (IMP-LD), extensive use of CDR in the energy and the industry sectors to
10 achieve net negative emissions (IMP-Neg), mitigation in the context of broader sustainable
11 development (IMP-SP), and the implications of a less rapid and gradual strengthening of near-term
12 mitigation actions (IMP-GS).{3.2, 3.3}
13
14 Pathways consistent with the implementation and extrapolation of countries’ current policies
15 see GHG emissions reaching 52-60 GtCO2-eq yr-1 by 2030 and to 46-67 GtCO2-eq yr-1 by 2050,
16 leading to a median global warming of 2.4°C to 3.5°C by 2100 (medium confidence). These
17 pathways consider policies at the time that they were developed. The Shared Socioeconomic
18 Pathways (SSPs) permit a more systematic assessment of future GHG emissions and their
19 uncertainties than was possible in AR5. The main emissions drivers include growth in population,
20 reaching 8.5-9.7 billion by 2050, and an increase in global GDP of 2.7-4.1% per year between 2015
21 and 2050. Final energy demand in the absence of any new climate policies is projected to grow to
22 around 480 to 750 EJ yr-1 in 2050 (compared to around 390 EJ in 2015). (medium confidence) The
23 highest emissions scenarios in the literature result in global warming of >5°C by 2100, based on
24 assumptions of rapid economic growth and pervasive climate policy failures. (high confidence). {3.3}
25
26 Many pathways in the literature show how to likely limit global warming compared to
27 preindustrial times to 2°C with no overshoot or to 1.5°C with limited overshoot. The likelihood
28 of limiting warming to 1.5C with no or limited overshoot has dropped in AR6 compared to
29 SR1.5 because global GHG emissions have risen since the time SR1.5 was published, leading to
30 higher near-term emissions (2030) and higher cumulative CO2 emissions until the time of net
31 zero (medium confidence). Only a small number of published pathways limit global warming to
32 1.5°C without overshoot over the course of the 21st century. {3.3, Annex III.II.3}
33
34 Cost-effective mitigation pathways assuming immediate actions to likely limit warming to 2°C
35 are associated with net global GHG emissions of 30-49 GtCO2-eq yr-1 by 2030 and 13-27 GtCO2-
36 eq yr-1 by 2050 (medium confidence). This corresponds to reductions, relative to 2019 levels, of 12-
37 46% by 2030 and 52-77% by 2050. Pathways that limit global warming to below 1.5°C with no or
38 limited overshoot require a further acceleration in the pace of the transformation, with net GHG
39 emissions typically around 21-36 GtCO2-eq yr-1 by 2030 and 1-15 GtCO2-eq yr-1 by 2050; thus
40 reductions of 38–63% by 2030 and 75-98% by 2050 relative to 2019 levels. {3.3}
41 Pathways following current NDCs1 until 2030 reach annual emissions of 47-57 GtCO2-eq by
42 2030, thereby making it impossible to limit warming to 1.5°C with no or limited overshoot and
43 strongly increasing the challenge to likely limit warming to 2°C (high confidence). A high
FOOTNOTE 1 Current NDCs refers to the most recent nationally determined contributions submitted to the
UNFCCC as well as those publicly announced with sufficient detail on targets, but not yet submitted, up to 11
October 2021, and reflected in studies published up to 11 October 2021.
Do Not Cite, Quote or Distribute 3-4 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 overshoot of 1.5°C increases the risks from climate impacts and increases the dependence on large
2 scale carbon dioxide removal from the atmosphere. A future consistent with current NDCs implies
3 higher fossil fuel deployment and lower reliance on low carbon alternatives until 2030, compared to
4 mitigation pathways with immediate action to likely limit warming to 2°C and below. To likely limit
5 warming to 2°C after following the NDCs to 2030, the pace of global GHG emission reductions
6 would need to accelerate quite rapidly from 2030 onward: to an average of 1.3-2.1 GtCO2-eq per year
7 between 2030 and 2050, which is similar to global CO2 emission reductions in 2020 due to the
8 COVID-19 pandemic, and around 70% faster than in immediate action pathways likely limiting
9 warming to 2°C. Accelerating emission reductions after following an NDC pathway to 2030 would be
10 particularly challenging because of the continued build-up of fossil fuel infrastructure that would be
11 expected to take place between now and 2030. {3.5, 4.2}
12 Pathways accelerating actions compared to current NDCs that reduce annual GHG emissions to
13 47 (38-51) GtCO2-eq by 2030, or 3-9 GtCO2-eq below projected emissions from fully
14 implementing current NDCs reduce the mitigation challenge for likely limiting warming to 2°C
15 after 2030. (medium confidence). The accelerated action pathways are characterized by a global, but
16 regionally differentiated, roll-out of regulatory and pricing policies. Compared to NDCs, they see less
17 fossil fuels and more low-carbon fuels until 2030, and narrow, but do not close the gap to pathways
18 assuming immediate global action using all available least-cost abatement options. All delayed or
19 accelerated action pathways likely limiting warming to below 2°C converge to a global mitigation
20 regime at some point after 2030 by putting a significant value on reducing carbon and other GHG
21 emissions in all sectors and regions. {3.5}
22 Mitigation pathways limiting warming to 1.5°C with no or limited overshoot reach 50%
23 reductions of CO2 in the 2030s, relative to 2019, then reduce emissions further to reach net zero
24 CO2 emissions in the 2050s. Pathways likely limiting warming to 2°C reach 50% reductions in
25 the 2040s and net zero CO2 by 2070s (medium confidence). {3.3, Cross-Chapter Box 3 in Chapter
26 3}
27
28 Peak warming in mitigation pathways is determined by the cumulative net CO 2 emissions until
29 the time of net zero CO2 and the warming contribution of other GHGs and climate forcers at
30 that time (high confidence). Cumulative net CO2 emissions from 2020 to the time of net zero CO2
31 are 510 (330-710) GtCO2 in pathways that limit warming to 1.5°C with no or limited overshoot and
32 890 (640-1160) GtCO2 in pathways likely limiting warming to 2.0°C. These estimates are consistent
33 with the assessment of remaining carbon budgets by WGI after adjusting for differences in peak
34 warming levels. {3.3, Box 3.4}
35 Rapid reductions in non-CO2 GHGs, particularly methane, would lower the level of peak
36 warming (high confidence). Residual non-CO2 emissions at the time of reaching net zero CO2 range
37 between 4-11 GtCO2-eq yr-1 in pathways likely limiting warming to 2.0°C or below. Methane (CH4)
38 is reduced by around 20% (1-46%) in 2030 and almost 50% (26-64%) in 2050, relative to 2019.
39 Methane emission reductions in pathways limiting warming to 1.5°C with no or limited overshoot are
40 substantially higher by 2030, 33% (19-57%), but only moderately so by 2050, 50% (33-69%).
41 Methane emissions reductions are thus attainable at relatively lower GHG prices but are at the same
42 time limited in scope in most 1.5-2°C pathways. Deeper methane emissions reductions by 2050 could
43 further constrain the peak warming. N2O emissions are reduced too, but similar to CH4, emission
44 reductions saturate for more stringent climate goals. In the mitigation pathways, the emissions of
45 cooling aerosols are reduced due to reduced use of fossil fuels. The overall impact on non-CO2-related
46 warming combines these factors. {3.3}
Do Not Cite, Quote or Distribute 3-5 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 Net zero GHG emissions imply net negative CO2 emissions at a level compensating residual non-
2 CO2 emissions. Only 30% of the pathways likely limiting warming to 2°C or below reach net
3 zero GHG emissions in the 21st century. (high confidence). In those pathways reaching net zero
4 GHGs, it is achieved around 10-20 years later than for net zero CO2. (medium confidence). The
5 reported quantity of residual non-CO2 emissions depends on accounting: the choice of GHG metric.
6 Reaching and sustaining global net zero GHG emissions, measured in terms of GWP-100, results in a
7 gradual decline of temperature. (high confidence) {3.3, Cross-Chapter Box 3 in Chapter 3, Cross-
8 Chapter Box 2 in Chapter 2}
9
10 Pathways likely limiting warming to 2°C and below exhibit substantial reductions in emissions
11 from all sectors (high confidence). Projected CO2 emissions reductions between 2019 and 2050 in
12 1.5°C pathways with no or limited overshoot are around 77% (31-96%) for energy demand, 115% for
13 energy supply (90 to 167%), and 148% for AFOLU (94 to 387%). In pathways likely limiting
14 warming to 2°C, projected CO2 emissions are reduced between 2019 and 2050 by around 49% for
15 energy demand, 97% for energy supply, and 136% for AFOLU. (medium confidence){3.4}
16
17 Delaying or sacrificing emissions reductions in one sector or region involves compensating
18 reductions in other sectors or regions if warming is to be limited (high confidence). Mitigation
19 pathways show differences in the timing of decarbonization and when net zero CO2 emissions are
20 achieved across sectors and regions. At the time of global net zero CO2 emissions, emissions in some
21 sectors and regions are positive while others are negative; the ordering depends on the mitigation
22 options available, the cost of those options, and the policies implemented. In cost-effective mitigation
23 pathways, the energy supply sector typically reaches net zero CO2 before the economy as a whole,
24 while the demand sectors reach net zero CO2 later, if ever (high confidence). {3.4}
25
26 Pathways likely limiting warming to 2°C and below involve substantial reductions in fossil fuel
27 consumption and a near elimination of the use of coal without CCS (high confidence). These
28 pathways show an increase in low carbon energy, with 88% (69-97%) of primary energy coming from
29 these sources by 2100. {3.4}
30
31 Stringent emissions reductions at the level required for 2°C and below are achieved through
32 increased direct electrification of buildings, transport, and industry, resulting in increased
33 electricity generation in all pathways (high confidence). Nearly all electricity in pathways likely
34 limiting warming to 2℃ or below is from low or no carbon technologies, with different shares of
35 nuclear, biomass, non-biomass renewables, and fossil CCS across pathways. {3.4}
36
37 The measures required to likely limit warming to 2°C or below can result in large scale
38 transformation of the land surface (high confidence). Pathways likely limiting warming to 2°C or
39 below are projected to reach net zero CO2 emissions in the AFOLU sector between 2020s and 2070,
40 with an increase of forest cover of about 322 million ha (-67 to 890 million ha) in 2050 in pathways
41 limiting warming to 1.5°C with no or limited overshoot. Cropland area to supply biomass for
42 bioenergy (including BECCS) is around 199 (56-482) million ha in 2100 in pathways limiting
43 warming to 1.5°C with no or limited overshoot. The use of bioenergy can lead to either increased or
44 reduced emissions, depending on the scale of deployment, conversion technology, fuel displaced, and
45 how/where the biomass is produced (high confidence). {3.4}
46
47 Anthropogenic land CO2 emissions and removals in IAM pathways cannot be directly compared
48 with those reported in national GHG inventories (high confidence). Methodologies enabling a
49 more like-for-like comparison between models’ and countries’ approaches would support more
50 accurate assessment of the collective progress achieved under the Paris Agreement. {3.4, 7.2.2.5}
Do Not Cite, Quote or Distribute 3-6 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1
2 Pathways that likely limiting warming to 2°C or below involve some amount of CDR to
3 compensate for residual GHG emissions remaining after substantial direct emissions reductions
4 in all sectors and regions (high confidence). CDR deployment in pathways serves multiple
5 purposes: accelerating the pace of emissions reductions, offsetting residual emissions, and creating the
6 option for net negative CO2 emissions in case temperature reductions need to be achieved in the long
7 term (high confidence). CDR options in the pathways are mostly limited to BECCS, afforestation and
8 DACCS. CDR through some measures in AFOLU can be maintained for decades but not in the very
9 long term because these sinks will ultimately saturate (high confidence). {3.4}
10
11 Mitigation pathways show reductions in energy demand relative to reference scenarios, through
12 a diverse set of demand-side interventions (high confidence). Bottom-up and non-IAM studies
13 show significant potential for demand-side mitigation. A stronger emphasis on demand-side
14 mitigation implies less dependence on CDR and, consequently, reduced pressure on land and
15 biodiversity. {3.4, 3.7}
16
17 Limiting warming requires shifting energy investments away from fossil-fuels and towards low-
18 carbon technologies (high confidence). The bulk of investments are needed in medium- and low-
19 income regions. Investment needs in the electricity sector are on average 2.3 trillion USD2015 yr -1
20 over 2023-2052 for pathways limiting temperature to 1.5°C with no or limited overshoot, and 1.7
21 trillion USD2015 yr-1 for pathways likely limiting warming to 2°C. {3.6.1}
22
23 Pathways likely avoiding overshoot of 2°C warming require more rapid near-term
24 transformations and are associated with higher up-front transition costs, but meanwhile bring
25 long-term gains for the economy as well as earlier benefits in avoided climate change impacts
26 (high confidence). This conclusion is independent of the discount rate applied, though the modelled
27 cost-optimal balance of mitigation action over time does depend on the discount rate. Lower discount
28 rates favour earlier mitigation, reducing reliance on CDR and temperature overshoot. {3.6.1, 3.8}
29
30 Mitigation pathways likely limiting warming to 2°C entail losses in global GDP with respect to
31 reference scenarios of between 1.3% and 2.7% in 2050; and in pathways limiting warming to
32 1.5°C with no or limited overshoot, losses are between 2.6% and 4.2%. Yet, these estimates do
33 not account for the economic benefits of avoided climate change impacts (medium confidence).
34 In mitigation pathways likely limiting warming to 2°C, marginal abatement costs of carbon are about
35 90 (60-120) USD2015/tCO2 in 2030 and about 210 (140-340) USD2015/tCO2 in 2050; in pathways
36 that limit warming to 1.5°C with no or limited overshoot, they are about 220 (170-290)
37 USD2015/tCO2 in 2030 and about 630 (430-990) USD2015/tCO2 in 20502. {3.6.1}
38
39 The global benefits of pathways likely limiting warming to 2°C outweigh global mitigation costs
40 over the 21st century, if aggregated economic impacts of climate change are at the moderate to
41 high end of the assessed range, and a weight consistent with economic theory is given to
42 economic impacts over the long-term. This holds true even without accounting for benefits in
43 other sustainable development dimensions or non-market damages from climate change
44 (medium confidence). The aggregate global economic repercussions of mitigation pathways include
45 the macroeconomic impacts of investments in low-carbon solutions and structural changes away from
46 emitting activities, co-benefits and adverse side effects of mitigation, (avoided) climate change
47 impacts, and (reduced) adaptation costs. Existing quantifications of global aggregate economic
FOOTNOTE2 Numbers in parenthesis represent he interquartile range of the scenario samples.
Do Not Cite, Quote or Distribute 3-7 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 impacts show a strong dependence on socioeconomic development conditions, as these shape
2 exposure and vulnerability and adaptation opportunities and responses. (Avoided) impacts for poorer
3 households and poorer countries represent a smaller share in aggregate economic quantifications
4 expressed in GDP or monetary terms, whereas their well-being and welfare effects are comparatively
5 larger. When aggregate economic benefits from avoided climate change impacts are accounted for,
6 mitigation is a welfare-enhancing strategy. (high confidence) {3.6.2}
7
8 The economic benefits on human health from air quality improvement arising from mitigation
9 action can be of the same order of magnitude as mitigation costs, and potentially even larger
10 (medium confidence). {3.6.3}
11
12 Differences between aggregate employment in mitigation pathways compared to reference
13 scenarios are relatively small, although there may be substantial reallocations across sectors,
14 with job creation in some sectors and job losses in others. The net employment effect (and its sign)
15 depends on scenario assumptions, modelling framework, and modelled policy design. Mitigation has
16 implications for employment through multiple channels, each of which impacts geographies, sectors
17 and skill categories differently. (medium confidence) {3.6.4}
18
19 The economic repercussions of mitigation vary widely across regions and households, depending
20 on policy design and level of international cooperation (high confidence). Delayed global
21 cooperation increases policy costs across regions, especially in those that are relatively carbon
22 intensive at present (high confidence). Pathways with uniform carbon values show higher mitigation
23 costs in more carbon-intensive regions, in fossil-fuels exporting regions and in poorer regions (high
24 confidence). Aggregate quantifications expressed in GDP or monetary terms undervalue the economic
25 effects on households in poorer countries; the actual effects on welfare and well-being are
26 comparatively larger (high confidence). Mitigation at the speed and scale required to likely limit
27 warming to 2°C or below implies deep economic and structural changes, thereby raising multiple
28 types of distributional concerns across regions, income classes and sectors (high confidence). {3.6.1,
29 3.6.4}
30
31 The timing of mitigation actions and their effectiveness will have significant consequences for
32 broader sustainable development outcomes in the longer term (high confidence). Ambitious
33 mitigation can be considered a precondition for achieving the Sustainable Development Goals,
34 especially for vulnerable populations and ecosystems with little capacity to adapt to climate impacts.
35 Dimensions with anticipated co-benefits include health, especially regarding air pollution, clean
36 energy access and water availability. Dimensions with potential trade-offs include food, employment,
37 water stress, and biodiversity, which come under pressure from large-scale CDR deployment, energy
38 affordability/access, and mineral resource extraction. (high confidence) {3.7}
39
40 Many of the potential trade-offs of mitigation measures for other sustainable development
41 outcomes depend on policy design and can thus be compensated or avoided with additional
42 policies and investments or through policies that integrate mitigation with other SDGs (high
43 confidence). Targeted SDG policies and investments, for example in the areas of healthy nutrition,
44 sustainable consumption and production, and international collaboration, can support climate change
45 mitigation policies and resolve or alleviate trade-offs. Trade-offs can be addressed by complementary
46 policies and investments, as well as through the design of cross-sectoral policies integrating
47 mitigation with the Sustainable Development Goals of health, nutrition, sustainable consumption and
48 production, equity and biodiversity. {3.7}
49
Do Not Cite, Quote or Distribute 3-8 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 Decent living standards, which encompass many SDG dimensions, are achievable at lower
2 energy use than previously thought (high confidence). Mitigation strategies that focus on lower
3 demands for energy and land-based resources exhibit reduced trade-offs and negative consequences
4 for sustainable development relative to pathways involving either high emissions and climate impacts
5 or those with high consumption and emissions that are ultimately compensated by large quantities of
6 BECCS. {3.7}
7
8 Different mitigation pathways are associated with different feasibility challenges, though
9 appropriate enabling conditions can reduce these challenges. Feasibility challenges are transient
10 and concentrated in the next two to three decades (high confidence). They are multi-dimensional,
11 context-dependent and malleable to policy, technological and societal trends. {3.8}
12
13 Mitigation pathways are associated with significant institutional and economic feasibility
14 challenges rather than technological and geophysical. The rapid pace of technological
15 development and deployment in mitigation pathways is not incompatible with historical records.
16 Institutional capacity is rather a key limiting factor for a successful transition. Emerging economies
17 appear to have highest feasibility challenges in the short to medium term. {3.8}
18
19 Pathways relying on a broad portfolio of mitigation strategies are more robust and resilient
20 (high confidence). Portfolios of technological solutions reduce the feasibility risks associated with the
21 low carbon transition. {3.8}
22
23
Do Not Cite, Quote or Distribute 3-9 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1
2 3.1 Introduction
3 3.1.1 Assessment of mitigation pathways and their compatibility with long-term goals
4 Chapter 3 takes a long-term perspective on climate change mitigation pathways. Its focus is on the
5 implications of long-term targets for the required short- and medium-term system changes and
6 associated greenhouse gas (GHG) emissions. This focus dictates a more global view and on issues
7 related to path-dependency and up-scaling of mitigation options necessary to achieve different
8 emissions trajectories, including particularly deep mitigation pathways that require rapid and
9 fundamental changes.
10
11 Stabilizing global average temperature change requires to reduce CO2 emissions to net zero. Thus, a
12 central cross-cutting topic within the Chapter is the timing of reaching net zero CO2 emissions and
13 how a “balance between anthropogenic emissions by sources and removals by sinks” could be
14 achieved across time and space. This includes particularly the increasing body of literature since the
15 IPCC Special Report on Global Warming of 1.5C (SR1.5) which focuses on net zero CO2 emissions
16 pathways that avoid temperature overshoot and hence do not rely on net negative CO2 emissions. The
17 chapter conducts a systematic assessment of the associated economic costs as well as the benefits of
18 mitigation for other societal objectives, such as the Sustainable Development Goals (SDGs). In
19 addition, the Chapter builds on SR1.5 and introduces a new conceptual framing for the assessment of
20 possible social, economic, technical, political, and geophysical “feasibility” concerns of alternative
21 pathways, including the enabling conditions that would need to fall into place so that stringent climate
22 goals become attainable.
23
24 The structure of the Chapter is as follows: Section 3.2 introduces different types of mitigation
25 pathways as well as the available modelling. Section 3.3 explores different emissions trajectories
26 given socio-economic uncertainties and consistent with different long-term climate outcomes. A
27 central element in this section is the systematic categorization of the scenario space according to key
28 characteristics of the mitigation pathways (including e.g., global average temperature change, socio-
29 economic development, technology assumptions, etc.). In addition, the section introduces selected
30 Illustrative Mitigation Pathways (IMPs) that are used across the whole report. Section 3.4 conducts a
31 sectoral analysis of the mitigation pathways, assessing the pace and direction of systems changes
32 across sectors. Among others, this section aims at the integration of the sectoral information across
33 WGIII AR6 chapters through a comparative assessment of the sectoral dynamics in economy-wide
34 systems models compared to the insights from bottom-up sectoral models (from Chapters 6-11).
35 Section 3.5 focuses on the required timing of mitigation actions, and implication of near-term choices
36 for the attainability of a range of long-term climate goals. After having explored the underlying
37 systems transitions and the required timing of the mitigation actions, Section 3.6 assesses the
38 economic implications, mitigation costs and benefits; and Section 3.7 assesses related co-benefits,
39 synergies, and possible trade-offs for sustainable development and other societal (non-climate)
40 objectives. Section 3.8 assumes a central role in the Chapter and introduces a multi-dimensional
41 feasibility metric that permits the evaluation of mitigation pathways across a range of feasibility
42 concerns. Finally, methods of the assessment and knowledge gaps are discussed in Section 3.9,
43 followed by Frequently Asked Questions.
44
45 3.1.2 Linkages to other Chapters in the Report
46 Chapter 3 is linked to many other chapters in the report. The most important connections exist with
47 Chapter 4 on mitigation and development pathways in the near to mid-term, with the sectoral chapters
48 (Chapters 6-11), with the chapters dealing with cross-cutting issues (e.g., feasibility), and finally also
49 with WGI and WGII AR6.
Do Not Cite, Quote or Distribute 3-10 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1
2 Within the overall framing of the AR6 report, Chapter 3 and Chapter 4 provide important
3 complementary views of the required systems transitions across different temporal and spatial scales.
4 While Chapter 3 focuses on the questions concerning the implications of the long-term objectives for
5 the medium-to-near-term transformations, Chapter 4 comes from the other direction, and focuses on
6 current near-term trends and policies (such as the Nationally Determined Contributions - NDCs) and
7 their consequences with regards to GHG emissions. The latter chapter naturally focuses thus much
8 more on the regional and national dimensions, and the heterogeneity of current and planned policies.
9 Bringing together the information from these two chapters enables the assessment of whether current
10 and planned actions are consistent with the required systems changes for the long-term objectives of
11 the Paris Agreement.
12
13 Important other linkages comprise the collaboration with the “sectoral” Chapters 6-11 to provide an
14 integrated cross-sectoral perspective. This information (including information also from the sectoral
15 chapters) is taken up ultimately also by Chapter 5 on demand/services and Chapter 12 for a further
16 assessment of sectoral potential and costs.
17
18 Linkages to other chapters exist also on the topic of feasibility, which are informed by the policy, the
19 sectoral and the demand chapters, the technology and finance chapters, as well as Chapter 4 on
20 national circumstances.
21
22 Close collaboration with WGI permitted the use of AR6-calibrated emulators, which assure full
23 consistency across the different working groups. Linkages to WGII concern the assessment of macro-
24 economic benefits of avoided impacts that are put into the context of mitigation costs as well as co-
25 benefits and trade-offs for sustainable development.
26
27 3.1.3 Complementary use of large scenario ensembles and a limited set of Illustrative
28 Mitigation Pathways (IMPs)
29 The assessment of mitigation pathways explores a wide scenario space from the literature within
30 which seven Illustrative Pathways (IPs) are explored. The overall process is indicated in Figure 3.5a.
31
32 For a comprehensive assessment, a large ensemble of scenarios is collected and made available
33 through an interactive AR6 scenario database. The collected information is shared across the chapters
34 of AR6 and includes more than 3000 different pathways from a diverse set of studies. After an initial
35 screening and quality control, scenarios were further vetted to assess if they sufficiently represented
36 historical trends (Annex III). Subsequently, the climate consequences of each scenario were assessed
37 using the climate emulator (leading to further classification). The assessment in Chapter 3 is however
38 not limited to the scenarios from the database, and wherever necessary other literature sources are also
39 assessed in order to bring together multiple lines of evidence.
40
41 In parallel, based on the overall AR6 assessment seven illustrative pathways (IP) were defined
42 representing critical mitigation strategies discussed in the assessment. The seven pathways are
43 composed of two sets: (i) one set of five Illustrative Mitigation Pathways (IMPs) and (ii) one set of
44 two reference pathways illustrative for high emissions. The IMPs are on the one hand representative
45 of the scenario space but help in addition to communicate archetypes of distinctly different systems
46 transformations and related policy choices. Subsequently, seven scenarios were selected from the full
47 database that fitted these storylines of each IP best. For these scenarios are more strict vetting criteria
48 were applied. The selection was done by first applying specific filters based on the storyline followed
49 by a final selection (see Box 3.1 and Figure 3.5 a).
50
Do Not Cite, Quote or Distribute 3-11 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1
2 START BOX HERE
3 Box 3.1 Illustrative Mitigation Pathways
4 The literature shows a wide range of possible emissions trajectories, depicting developments in the
5 absence of new climate policies or showing pathways consistent with the Paris Agreement. From the
6 literature, a set of five Illustrative Mitigation Pathways (IMPs) was selected to denote implications of
7 choices on socio-economic development and climate policies, and the associated transformations of
8 the main GHG emitting sectors (see Figure 3.5b). The IMPs include a set of transformative pathways
9 that illustrates how choices may lead to distinctly different transformations that may keep temperature
10 increase to below 2°C or 1.5°C. These pathways illustrate the implications of a focus on renewable
11 energy such as solar and wind; reduced energy demand; extensive use of CDR in the energy and the
12 industry sectors to achieve net negative emissions and reliance on other supply-side measures;
13 strategies that avoid net-negative carbon emissions, and gradual strengthening. In addition, one IMP
14 explores how climate policies consistent with keeping temperature to 1.5˚C can be combined with a
15 broader shift towards sustainable development. These IMPs are used in various chapters, exploring for
16 instance their implications for different sectors, regions, and innovation characteristics (see Figure
17 3.5b).
18
19 END BOX HERE
20 3.2 What are mitigation pathways compatible with long-term goals?
21 3.2.1 Scenarios and emission pathways
22 Scenarios and emission pathways are used to explore possible long-term trajectories, the effectiveness
23 of possible mitigation strategies, and to help understand key uncertainties about the future. A scenario
24 is an integrated description of a possible future of the human–environment system (Clarke et al.
25 2014), and could be a qualitative narrative, quantitative projection, or both. Scenarios typically
26 capture interactions and processes driving changes in key driving forces such as population, GDP,
27 technology, lifestyles, and policy, and the consequences on energy use, land use, and emissions.
28 Scenarios are not predictions or forecasts. An emission pathway is a modelled trajectory of
29 anthropogenic emissions (Rogelj et al. 2018a) and, therefore, a part of a scenario.
30 There is no unique or preferred method to develop scenarios, and future pathways can be developed
31 from diverse methods, depending on user needs and research questions (Turnheim et al. 2015;
32 Trutnevyte et al. 2019a; Hirt et al. 2020). The most comprehensive scenarios in the literature are
33 qualitative narratives that are translated into quantitative pathways using models (Clarke et al. 2014;
34 Rogelj et al. 2018a). Schematic or illustrative pathways can also be used to communicate specific
35 features of more complex scenarios (Allen et al. 2018). Simplified models can be used to explain the
36 mechanisms operating in more complex models (e.g., Emmerling et al. (2019)). Ultimately, a diversity
37 of scenario and modelling approaches can lead to more robust findings (Gambhir et al. 2019; Schinko
38 et al. 2017).
39 3.2.1.1 Reference scenarios
40 It is common to define a reference scenario (also called a baseline scenario). Depending on the
41 research question, a reference scenario could be defined in different ways (Grant et al. 2020): 1) a
42 hypothetical world with no climate policies or climate impacts (Kriegler et al. 2014b), 2) assuming
43 current policies or pledged policies are implemented (Roelfsema et al. 2020), or 3) a mitigation
44 scenario to compare sensitivity with other mitigation scenarios (Kriegler et al. 2014a; Sognnaes et al.
45 2021).
Do Not Cite, Quote or Distribute 3-12 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 No-climate-policy reference scenarios have often been to compare with mitigation scenarios (Clarke
2 et al. 2014). A no-climate-policy scenario assumes that no future climate policies are implemented,
3 beyond what is in the model calibration, effectively implying that the carbon price is zero. No-
4 climate-policy reference scenarios have a broad range depending on socioeconomic assumptions and
5 model characteristics, and consequently are important when assessing mitigation costs (Riahi et al.
6 2017; Rogelj et al. 2018b). As countries move forward with climate policies of varying stringency,
7 no-climate-policy baselines are becoming increasingly hypothetical (Hausfather and Peters 2020).
8 Studies clearly show current policies are having an effect, particularly when combined with the
9 declining costs of low carbon technologies (IEA 2020a; UNEP 2020; Roelfsema et al. 2020; Sognnaes
10 et al. 2021), and, consequently, realised trajectories begin to differ from earlier no-climate-policy
11 scenarios (Burgess et al. 2020). High-end emission scenarios, such as RCP8.5 and SSP5-8.5, are
12 becoming less likely with climate policy and technology change (see Box 3.3), but high-end
13 concentration and warming levels may still be reached with the inclusion of strong carbon or climate
14 feedbacks (Pedersen et al. 2020; Hausfather and Peters 2020).
15 3.2.1.2 Mitigation scenarios
16 Mitigation scenarios explore different strategies to meet climate goals and are typically derived from
17 reference scenarios by adding climate or other policies. Mitigation pathways are often developed to
18 meet a predefined level of climate change, often referred to as a backcast. There are relatively few
19 IAMs that include an endogenous climate model or emulator due to the added computational
20 complexity, though exceptions do exist. In practice, models implement climate constraints by either
21 iterating carbon price assumptions (Strefler et al. 2021b) or by adopting an associated carbon budget
22 (Riahi et al. 2021). In both cases, other GHGs are typically controlled by CO2-equivalent pricing. A
23 large part of the AR5 literature has focused on forcing pathways towards a target at the end of the
24 century (van Vuuren et al. 2007, 2011; Clarke et al. 2009; Blanford et al. 2014; Riahi et al. 2017),
25 featuring a temporary overshoot of the warming and forcing levels (Geden and Löschel 2017). In
26 comparison, many recent studies explore mitigation strategies that limit overshoot (Johansson et al.
27 2020; Riahi et al. 2021). An increasing number of IAM studies also explore climate pathways that
28 limit adverse side-effects with respect to other societal objectives, such as food security (van Vuuren
29 et al. 2019; Riahi et al. 2021) or larger sets of sustainability objectives (Soergel et al. 2021a).
30
31 3.2.2 The utility of Integrated Assessment Models
32 Integrated Assessment Models (IAMs) are critical for understanding the implications of long-term
33 climate objectives for the required near-term transition. For doing so, an integrated systems
34 perspective including the representation of all sectors and GHGs is necessary. IAMs are used to
35 explore the response of complex systems in a formal and consistent framework. They cover broad
36 range of modelling frameworks (Keppo et al. 2021). Given the complexity of the systems under
37 investigation, IAMs necessarily make simplifying assumptions and therefore results need to be
38 interpreted in the context of these assumptions. IAMs can range from economic models that consider
39 only carbon dioxide emissions through to detailed process-based representations of the global energy
40 system, covering separate regions and sectors (such as energy, transport, and land use), all GHG
41 emissions and air pollutants, interactions with land and water, and a reduced representation of the
42 climate system. IAMs are generally driven by economics and can have a variety of characteristics
43 such as partial-, general, or non-equilibrium, myopic or perfect foresight, be based on optimization or
44 simulation, have exogenous or endogenous technological change, amongst many other characteristics.
45 IAMs take as input socioeconomic and technical variables and parameters to represent various
46 systems. There is no unique way to integrate this knowledge into a model, and due to their
47 complexity, various simplifications and omissions are made for tractability. IAMs therefore have
48 various advantages and disadvantages which need to be weighed up when interpreting IAM outcomes.
49 Annex III contains an overview of the different types of models and their key characteristics.
Do Not Cite, Quote or Distribute 3-13 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 Most IAMs are necessarily broad as they capture long-term dynamics. IAMs are strong in showing
2 the key characteristics of emission pathways and are most suited to questions related to short- versus
3 long-term trade-offs, key interactions with non-climate objectives, long-term energy and land-use
4 characteristics, and implications of different overarching technological and policy choices (Rogelj et
5 al. 2018a; Clarke et al. 2014). While some IAMs have an high level of regional and sectoral detail, for
6 questions that require higher levels of granularity (e.g., local policy implementation) specific region
7 and sector models may be better suited. Utility of the IAM pathways increases when the quantitative
8 results are contextualized through qualitative narratives or other additional types of knowledge to
9 provide deeper insights (Geels et al. 2016a; Weyant 2017; Gambhir et al. 2019).
10 IAMs have a long history in addressing environmental problems, particularly in the IPCC assessment
11 process (van Beek et al. 2020). Many policy discussions have been guided by IAM-based
12 quantifications, such as the required emission reduction rates, net zero years, or technology
13 deployment rates required to meet certain climate outcomes. This has led to the discussion whether
14 IAM scenarios have become performative, meaning that they act upon, transform or bring into being
15 the scenarios they describe (Beck and Mahony 2017, 2018). Transparency of underlying data and
16 methods is critical for scenario users to understand what drives different scenario results (Robertson
17 2020). A number of community activities have thus focused on the provision of transparent and
18 publicly accessible databases of both input and output data (Riahi et al. 2012; Huppmann et al. 2018;
19 Krey et al. 2019; Daioglou et al. 2020) as well as the provision of open-source code, and increased
20 documentation (Annex III). Transparency is needed to reveal conditionality of results on specific
21 choices in terms of assumptions (e.g., discount rates) and model architecture. More detailed
22 explanations of underlying model dynamics would be critical to increase the understanding of what
23 drives results (Bistline et al. 2020; Butnar et al. 2020; Robertson 2020).
24 Mitigation scenarios developed for a long-term climate constraint typically focus on cost-effective
25 mitigation action towards a long-term climate goal. Results from IAM as well as sectoral models
26 depend on model structure (Mercure et al. 2019), economic assumptions (Emmerling et al. 2019),
27 technology assumptions (Pye et al. 2018), climate/emissions target formulation (Johansson et al.
28 2020), and the extent to which pre-existing market distortions are considered (Guivarch et al. 2011).
29 The vast majority of IAM pathways do not consider climate impacts (Schultes et al. 2021). Equity
30 hinges upon ethical and normative choices. As most IAM pathways follow the cost-effectiveness
31 approach, they do not make any additional equity assumptions. Notable exceptions include (Tavoni et
32 al. 2015; Pan et al. 2017; van den Berg et al. 2020; Bauer et al. 2020). Regional IAM results need thus
33 to be assessed with care, considering that emissions reductions are happening where it is most cost-
34 effective, which needs to be separated from the fact who is ultimately paying for the mitigation costs.
35 Cost-effective pathways can provide a useful benchmark, but may not reflect real-world developments
36 (Trutnevyte 2016; Calvin et al. 2014a). Different modelling frameworks may lead to different
37 outcomes (Mercure et al. 2019). Recent studies have shown that other desirable outcomes can evolve
38 with only minor deviations from cost-effective pathways (Neumann and Brown 2021; Bauer et al.
39 2020). IAM and sectoral models represent social, political, and institutional factors only in a
40 rudimentary way. This assessment is thus relying on new methods for the ex-post assessment of
41 feasibility concerns (Jewell and Cherp 2020; Brutschin et al. 2021). A literature is emerging that
42 recognises and reflects on the diversity and strengths/weaknesses of model-based scenario analysis
43 (Keppo et al. 2021).
44 The climate constraint implementation can have a meaningful impact on model results. The literature
45 so far included many temperature overshoot scenarios with heavy reliance on long-term CDR and net
46 negative CO2 emissions to bring back temperatures after the peak (Johansson et al. 2020; Rogelj et al.
47 2019b). New approaches have been developed to avoid temperature overshoot. The new generation of
48 scenarios show that CDR is important beyond its ability to reduce temperature, but is essential also for
Do Not Cite, Quote or Distribute 3-14 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 offsetting residual emissions to reach a net zero CO2 emissions (Rogelj et al. 2019b; Johansson et al.
2 2020; Riahi et al. 2021; Strefler et al. 2021b).
3 Many factors influence the deployment of technologies in the IAMs. Since AR5, there has been
4 fervent debate on the large-scale deployment of Bioenergy with Carbon Capture and Storage
5 (BECCS) in scenarios (Geden 2015; Fuss et al. 2014; Smith et al. 2016; Anderson and Peters 2016;
6 van Vuuren et al. 2017; Galik 2020; Köberle 2019). Hence, many recent studies explore mitigation
7 pathways with limited BECCS deployment (Grubler et al. 2018; Soergel et al. 2021a; van Vuuren et
8 al. 2019; Riahi et al. 2021). While some have argued that technology diffusion in IAMs occurs too
9 rapidly (Gambhir et al. 2019), others argued that most models prefer large-scale solutions resulting in
10 a relatively slow phase-out of fossil fuels (Carton 2019). While IAMs are particularly strong on
11 supply-side representation, demand-side measures still lag in detail of representation despite progress
12 since AR5 (Grubler et al. 2018; van den Berg et al. 2019; Lovins et al. 2019; O’Neill et al. 2020b;
13 Hickel et al. 2021; Keyßer and Lenzen 2021). The discount rate has a significant impact on the
14 balance between near-term and long-term mitigation. Lower discount rates <4% (than used in IAMs)
15 may lead to more near-term emissions reductions – depending on the stringency of the target
16 (Emmerling et al. 2019; Riahi et al. 2021). Models often use simplified policy assumptions (O’Neill et
17 al. 2020b) which can affect the deployment of technologies (Sognnaes et al. 2021). Uncertainty in
18 technologies can lead to more or less short-term mitigation (Grant et al. 2021; Bednar et al. 2021).
19 There is also a recognition to put more emphasis on what drives the results of different IAMs
20 (Gambhir et al. 2019) and suggestions to focus more on what is driving differences in result across
21 IAMs (Nikas et al. 2021). As noted by Weyant (2017) (p.131), “IAMs can provide very useful
22 information, but this information needs to be carefully interpreted and integrated with other
23 quantitative and qualitative inputs in the decision-making process.”
24 3.2.3 The scenario literature and scenario databases
25 IPCC reports have often used voluntary submissions to a scenario database in its assessments. The
26 database is an ensemble of opportunity, as there is not a well-designed statistical sampling of the
27 hypothetical model or scenario space: the literature is unlikely to cover all possible models and
28 scenarios, and not all scenarios in the literature are submitted to the database. Model inter-
29 comparisons are often the core of scenario databases assessed by the IPCC (Cointe et al. 2019; Nikas
30 et al. 2021). Single model studies may allow more detailed sensitivity analyses or address specific
31 research questions. The scenarios that are organised within the scientific community are more likely
32 to enter the assessment process via the scenario database (Cointe et al. 2019), while scenarios from
33 different communities, in the emerging literature, or not structurally consistent with the database may
34 be overlooked. Scenarios in the grey literature may not be assessed even though they may have
35 greater weight in a policy context.
36 One notable development since IPCC AR5 is the Shared Socioeconomic Pathways (SSPs),
37 conceptually outlined in (Moss et al. 2010) and subsequently developed to support integrated climate
38 research across the IPCC Working Groups (O’Neill et al. 2014). Initially, a set of SSP narratives were
39 developed, describing worlds with different challenges to mitigation and adaptation (O’Neill et al.
40 2017a): SSP1 (sustainability), SSP2 (middle of the road), SSP3 (regional rivalry), SSP4 (inequality)
41 and SSP5 (rapid growth). The SSPs have now been quantified in terms of energy, land-use change,
42 and emission pathways (Riahi et al. 2017), for both no-climate-policy reference scenarios and
43 mitigation scenarios that follow similar radiative forcing pathways as the RCPs assessed in AR5 WGI.
44 Since then the SSPs have been successfully applied in 1000s of studies (O’Neill et al. 2020b)
45 including some critiques on the use and application of the SSP framework (Rosen 2021; Pielke and
46 Ritchie 2021). A selection of the quantified SSPs are used prominently in IPCC AR6 WGI as they
47 were the basis for most climate modelling since AR5 (O’Neill et al. 2016). Since 2014, when the first
48 set of SSP data was made available, there has been a divergence between scenario and historic trends
Do Not Cite, Quote or Distribute 3-15 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 (Burgess et al. 2020). As a result, the SSPs require updating (O’Neill et al. 2020b). Most of the
2 scenarios in the AR6 database are SSP-based and consider various updates compared to the first
3 release (Riahi et al. 2017).
4 3.2.4 The AR6 scenario database
5 To facilitate this assessment, a large ensemble of scenarios has been collected and made available
6 through an interactive WGIII AR6 scenario database. The collection of the scenario outputs is
7 coordinated by Chapter 3 and expands upon the IPCC SR1.5 scenario explorer (Huppmann et al.
8 2018; Rogelj et al. 2018a). A complementary database for national pathways has been established by
9 Chapter 4. Annex III contains full details on how the scenario database was compiled.
10 The AR6 scenario database contains 3131 scenarios (see Figure 3.5a). After an initial screening and
11 quality control, scenarios were further vetted to assess if they sufficiently represented historical trends
12 (Annex III). Of the initial 2266 scenarios with global scope, 1686 scenarios passed the vetting process
13 and are assessed in this Chapter. The scenarios that did not pass the vetting are still available in the
14 database. The vetted scenarios were from over 50 different model families, or over 100 when
15 considering all versions of the same family (Figure 3.1). The scenarios originated from over 15
16 different model intercomparison projects, with very few scenarios originating from individual studies
17 (Figure 3.2). Because of the uneven distribution of scenarios from different models and projects,
18 uncorrected statistics from the database can be misleading.
19
20 Figure 3.1 Scenario counts from each model family defined as all versions under the same model’s name.
21
Do Not Cite, Quote or Distribute 3-16 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1
2 Figure 3.2 Scenario counts from each named project.
3
4 Each scenario with sufficient data is given a temperature classification using climate model emulators.
5 Three emulators were used in the assessment: FAIR (Smith et al. 2018), CICERO-SCM (Skeie et al.
6 2021), MAGICC (Meinshausen et al. 2020). Only the results of MAGICC are shown in this chapter as
7 it adequately covers the range of outcomes. The emulators are calibrated against the behaviour of
8 complex climate models and observation data, consistent with the outcomes of AR6 WGI AR6 (WGI
9 cross-chapter Box 7.1). The climate assessment is a three-step process of harmonization, infilling and
10 a probabilistic climate model emulator run (Annex III.2.5.1). Warming projections until the year 2100
11 were derived for 1574 scenarios, of which 1202 passed vetting, with the remaining scenarios having
12 insufficient information (Table 3.1 and Figure 3.3). For scenarios that limit warming to 2°C or below,
13 the SR15 classification was adopted in AR6, with more disaggregation provided for higher warming
14 levels (Table 3.1). These choices can be compared with the selection of common global warming
15 levels (GWL) of 1.5°C, 2°C, 3°C and 4°C to classify climate change impacts in the WGII assessment.
16
17 Table 3.1 Classification of emissions scenarios into warming levels using MAGICC
Description Subset WGI SSP WGIII Scenarios
IP
C1: Below 1.5°C with no or <1.5°C peak warming with ≥33% chance and < 1.5°C SSP1-1.9 SP, LD, 97
limited overshoot end of century warming with >50% chance -, Ren
C2: Below 1.5°C with high <1.5°C peak warming with <33% chance and < 1.5°C 133
overshoot end of century warming with >50% chance
C3: Likely below 2°C <2°C peak warming with >67% chance SSP2-2.6 GS, Neg 311
C4: Below 2oC <2°C peak warming with >50% chance 159
C5: Below 2.5°C <2.5°C peak warming with >50% chance 212
Do Not Cite, Quote or Distribute 3-17 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
C6: Below 3°C <3°C peak warming with >50% chance SSP2-4.5 Mod-Act 97
C7: Below 4°C <4°C peak warming with >50% chance SSP3-7.0 Cur-Pol 164
C8: Above 4°C >4°C peak warming with ≥50% chance SSP5-8.5 29
1
2
3 Figure 3.3 Of the 1686 scenarios that passed vetting, 1202 had sufficient data available to be classified
4 according to temperature, with an uneven distribution across warming levels.
5 In addition to the temperature classification, each scenario is assigned to one of the following policy
6 categories: (P0) diagnostic scenarios – 100 of 1686 vetted scenarios; (P1) scenarios with no globally
7 coordinated policy and either (P1a) no climate mitigation efforts – 119, (P1b) current national
8 mitigation efforts – 59, (P1c) Nationally Determined Contributions (NDCs) – 110, or (P1d) other non-
9 standard assumptions – 104; (P2) globally coordinated climate policies with immediate (i.e. before
10 2030) action – 73, (P2a) without any transfer of emission permits – 435, (P2b) with transfers – 70; or
11 (P2c) with additional policy assumptions – 55; (P3) globally coordinated climate policies with
12 delayed (i.e. from 2030 onwards or after 2030) action, preceded by (P3a) no mitigation commitment
13 or current national policies – 7, (P3b) NDCs – 376, (P3c) NDCs and additional policies – 18 and
14 (P3d) other non-standard regional assumptions – 0; (P4) Cost-Benefit Analysis (CBA) – 2. The policy
15 categories were identified using text pattern matching on the scenario metadata and calibrated on the
16 best-known scenarios from model intercomparisons, with further validation against the related
17 literature, reported emission and carbon price trajectories, and exchanges with modellers. If the
18 information available is enough to qualify a policy category number but not sufficient for a
19 subcategory, then only the number is retained (e.g., P2 instead of P2a/b/c). A suffix added after P0
20 further qualifies a diagnostic scenario as one of the other policy categories.To demonstrate the
21 diversity of the scenarios, the vetted scenarios were classified into different categories along the
22 dimensions of population, GDP, energy, and cumulative emissions (Figure 3.4). The number of
23 scenarios in each category provides some insight into the current literature, but this does not indicate a
24 higher probability of that category occurring in reality. For population, the majority of scenarios are
25 consistent with the SSP2 ‘middle of the road’ category, with very few scenarios exploring the outer
Do Not Cite, Quote or Distribute 3-18 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 extremes. GDP has a slightly larger variation, but overall, most scenarios are around the SSP2
2 socioeconomic assumptions. The level of CCS and CDR is expected to change depending on the
3 extent of mitigation, but there remains extensive use of both CDR and CCS in scenarios. CDR is
4 dominated by bioenergy with CCS (BECCS) and sequestration on land, with relatively few scenarios
5 using Direct Air Capture with Carbon Storage (DACCS) and even less with Enhanced Weathering
6 and other technologies (not shown). In terms of energy consumption, final energy has a much smaller
7 range than primary energy as conversion losses are not included in final energy. Both mitigation and
8 reference scenarios are shown, so there is a broad spread in different energy carriers represented in the
9 database. Bioenergy has a number of scenarios at around 100EJ, representing a constraint used in
10 many model intercomparisons.
11
12
13
14 Figure 3.4 Histograms for key categories in the AR6 scenario database. Only scenarios that passed vetting
15 are shown. For population and GDP, the SSP input data are also shown. The grey shading represents the
16 0-100% range (light grey), 25-75% range (dark grey), and the median is a black line. The figures with
17 white areas are outside of the scenario range, but the axis limits are retained to allow comparability with
18 other categories. Each subfigure potentially has different x- and y-axis limits. Each figure also potentially
19 contains different numbers of scenarios, depending on what was submitted to the database.
20 Source: AR6 scenarios database.
21 3.2.5 Illustrative Mitigation Pathways
22 Successive IPCC Ars have used scenarios to illustrate key characteristics of possible climate (policy)
23 futures. In IPCC AR5 four RCPs made the basis of climate modelling in WGI and WGII, with WGIII
24 assessing over 1,000 scenarios spanning those RCPs (Clarke et al. 2014). Of the over 400 hundred
25 scenarios assessed in SR15, four scenarios were selected to highlight the trade-off between short-term
26 emission reductions and long-term deployment of BECCS (Rogelj et al. 2018a), referred to as
27 ‘Illustrative Pathways’. AR6 WGI and WGII rely on the scenarios selected for CMIP6, called
28 ScenarioMIP (O’Neill et al. 2016), to assess warming levels. IPCC AR6 WGIII uses in addition to the
29 full set of scenarios also selected Illustrative Mitigation Pathways (IMPs).
30 In WGIII, IMPs were selected to denote the implications of different societal choices for the
31 development of future emissions and associated transformations of main GHG emitting sectors (see
32 Box 3.1 and Figure 3.5a). The most important function of the IMPs is to illustrate key themes that
Do Not Cite, Quote or Distribute 3-19 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 form a common thread in the report, both with a storyline and a quantitative illustration. The storyline
2 describes the key characteristics that define an IMP. The quantitative versions of the IMPs provide
3 numerical values that are internally consistent and comparable across chapters of the report. The
4 quantitative IMPs have been selected from the AR6 scenario database. No assessment of the
5 likelihood of each IMP has been made.
6 The selected scenarios (IPs) are divided into two sets (Figure 3.5 and Figure 3.6): two reference
7 pathways illustrative of high emissions and five Illustrative Mitigation Pathways (IMPs). The
8 narratives are explained in full in Annex III. The two reference pathways explore the consequences of
9 current policies and pledges: Current Policies (CurPol) and Moderate Action (ModAct). The CurPol
10 pathway explores the consequences of continuing along the path of implemented climate policies in
11 2020 and only a gradual strengthening after that. The scenario illustrates the outcomes of many
12 scenarios in the literature that project the outcomes of current policies. The ModAct pathway explores
13 the impact of implementing the Nationally Determined Contributions as formulated in 2020 and some
14 further strengthening after that. In line with current literature, these two reference pathways lead to an
15 increase in global mean temperature of more than 2°C (see Section 3.3).
16 The Illustrative Mitigation Pathways properly explore different pathways consistent with meeting the
17 long-term temperature goals of the Paris Agreement. They represent five different pathways that
18 emerge from the overall assessment. The IMPs consist of pathways with: gradual strengthening of
19 current policies (GS), extensive use of net negative emissions (Neg), renewables (Ren), low demand
20 (LD), and shifting pathways (SP). Each of these pathways can be implemented with different levels
21 of ambition. In the IMP framework, GS is consistent with staying likely below 2°C (C3), Neg shows a
22 strategy that also stays likely below 2°C level but returns to nearly 1.5°C by the end of the century
23 (hence indicated as C2*). The other variants that can limit warming to 1.5°C (C1) were selected. In
24 addition to these IMPs, sensitivity cases that explore alternative warming levels (C3) for Neg and Ren
25 are assessed (Neg-2.0 and Ren-2.0).
Do Not Cite, Quote or Distribute 3-20 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 Panel A
2
3 Panel B
4
5 Figure 3.5 Panel A: Process for creating the AR6 scenario database and selecting the illustrative
6 (mitigation) pathways. The compiled scenarios in the AR6 scenarios database were vetted for consistency
7 with historical statistics and subsequently a temperature classification was added using climate model
8 emulators. The illustrative (mitigation) pathways were selected from the full set of pathways based on
9 storylines of critical mitigation strategies that emerged from the assessment. Panel B: An overview of the
10 Illustrative Pathways selected for use in IPCC AR6 WGIII, consisting of pathways illustrative of higher
11 emissions, Current Policies (CurPol) and Moderate Action (ModAct), and Illustrative Mitigation
12 Pathways (IMPs): gradual strengthening of current policies (GS), extensive use of net negative emissions
13 (Neg), renewables (Ren), low demand (LD), and shifting pathways (SP).
Do Not Cite, Quote or Distribute 3-21 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1
2
3 Figure 3.6 Overview of the net CO2 emissions and Kyoto GHG emissions for each IMP
4 The IMPs are selected to have different mitigation strategies, which can be illustrated looking at the
5 energy system and emission pathways (Figure 3.7 and Figure 3.8). The mitigation strategies show the
6 different options in emission reduction (Figure 3.7). Each panel shows the key characteristics leading
7 to total GHG emissions, consisting of residual (gross) emissions (fossil CO2 emissions, CO2 emissions
8 from industrial processes, and non-CO2 emissions) and removals (net land-use change, bioenergy with
9 carbon capture and storage, and direct air carbon capture and storage), in addition to avoided
10 emissions through the use of carbon capture and storage on fossil fuels. The Neg and GS scenarios
11 were shown to illustrate scenarios with a significant role of CDR. The energy supply (Figure 3.8)
12 shows the phase-out of fossil fuels in the LD, Ren and SP cases, but a less substantial decrease in the
13 Neg case. The GS case needs to make up its slow start by 1) rapid reductions mid-century and 2)
14 massive reliance on net negative emissions by the end of the century. The CurPol and ModAct cases
15 both result in relatively high emissions, showing a slight increase and stabilization compared to
16 current emissions, respectively.
Do Not Cite, Quote or Distribute 3-22 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1
2 Figure 3.7 The residual fossil fuel and industry emissions, net land-use change, CDR, and non-CO2
3 emissions (using AR6 GWP100) for each of the seven illustrative pathways. Fossil CCS is also shown,
4 though this does not lead to emissions to the atmosphere.
5
6 Figure 3.8 The energy system in each of the illustrative pathways.
7
Do Not Cite, Quote or Distribute 3-23 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 3.3 Emission pathways, including socio-economic, carbon budget and
2 climate responses uncertainties
3 3.3.1 Socio-economic drivers of emissions scenarios
4 Greenhouse gas emissions mainly originate from the use and transformation of energy, agriculture,
5 land use (change) and industrial activities. The future development of these sources is influenced by
6 trends in socio-economic development, including population, economic activity, technology, politics,
7 lifestyles, and climate policy. Trends for these factors are not independent, and scenarios provide a
8 consistent outlook for these factors together (see Section 3.2). Marangoni et al. (2017) show that in
9 projections, assumptions influencing energy intensity (e.g., structural change, lifestyle and efficiency)
10 and economic growth are the most important determinants of future CO2 emissions from energy
11 combustion. Other critical factors include technology assumptions, preferences, resource assumptions
12 and policy (van Vuuren et al. 2008). As many of the factors are represented differently in specific
13 models, the model itself is also an important factor – providing a reason for the importance of model
14 diversity (Sognnaes et al. 2021). For land use, Stehfest et al. (2019) show that assumptions on
15 population growth are more dominant given that variations in per capita consumption of food are
16 smaller than for energy. Here, we only provide a brief overview of some key drivers. We focus first
17 on so-called reference scenarios (without stringent climate policy) and look at mitigation scenarios in
18 detail later. We use the SSPs to discuss trends in more detail. The SSPs were published in 2017, and
19 by now, some elements will have to be updated (O’Neill et al. 2020b). Still, the ranges represent the
20 full literature relatively well.
21
22 Historically, population and GDP have been growing over time. Scenario studies agree that further
23 global population growth is likely up to 2050, leading to a range of possible outcomes of around 8.5-
24 11 billion people (see Figure 3.9). After 2050, projections show a much wider range. If fertility drops
25 below replacement levels, a decline in the global population is possible (as illustrated by SSP1 and
26 SSP5). This typically includes scenarios with rapid development and investment in education.
27 However, median projections mostly show a stabilisation of the world population (e.g., SSP2), while
28 high-end projections show a continued growth (e.g., SSP3). The UN Population Prospects include
29 considerably higher values for both the medium projection and the high end of the range than the SSP
30 scenarios (KC and Lutz 2017; UN 2019). The most recent median UN projection reaches almost 11
31 billion people in 2100. The key differences are in Africa and China: here, the population projections
32 are strongly influenced by the rate of fertility change (faster drop in SSPs). Underlying, the UN
33 approach is more based on current demographic trends while the SSPs assume a broader range of
34 factors (including education) driving future fertility.
35
36 Economic growth is even more uncertain than the population projections (Figure 3.9). The average
37 growth rate of GDP was about 2.8% per year (constant USD) in the 1990-2019 period (The World
38 Bank 2021). In 2020, the Covid-19 crisis resulted in a considerable drop in GDP (estimated around 4-
39 5%) (IMF 2021). After a recovery period, most economic projections assume growth rates to
40 converge back to previous projections, although at a lower level (IMF 2021; OECD 2021) (see also
41 Box 3.2). In the long-term, assumptions on future growth relate to political stability, the role of the
42 progress of the technology frontier and the degree to which countries can catch up (Johansson et al.
43 2013). The SSP scenarios cover an extensive range, with low per capita growth in SSP3 and SSP4
44 (mostly in developing countries) and rapid growth in SSP1 and SSP5. At the same, however, also
45 scenarios outside the range have some plausibility – including the option of economic decline (Kallis
46 et al. 2012) or much faster economic development (Christensen et al. 2018). The OECD long-term
47 projection is at the global level reasonably consistent with SSP2. Equally important economic
48 parameters include income distribution (inequity) and the type of growth (structural change, i.e.,
49 services vs manufacturing industries). Some projections (like SSP1) show a considerable convergence
50 of income levels within and across countries, while in other projections, this does not occur (e.g.,
51 SSP3). Most scenarios reflect the suggested inverse relationship between the assumed growth rate for
52 income and population growth (Figure 3.9e). SSP1 and SSP5 represent examples of scenarios with
53 relatively low population increase and relatively high-income increase over the century. SSP3
Do Not Cite, Quote or Distribute 3-24 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 represents an example of the opposite – while SSP2 and SSP4 are placed more in the middle. Nearly
2 all scenarios assessed here do not account for climate impacts on growth (mostly for methodological
3 reasons). As discussed in Section 3.5 these impacts can be considerable. An emerging area of
4 literature emphasises the possibility of stabilisation (or even decline) of income levels in developed
5 countries, arguing that such a trend would be preferred or even needed for environmental reasons
6 (Anderson and Larkin 2013; Kallis et al. 2020; Hickel and Kallis 2020; Keyßer and Lenzen 2021;
7 Hickel et al. 2021) (see also Chapter 5). Such scenarios are not common among IAM outcomes, that
8 are more commonly based on the idea that decarbonisation can be combined with economic growth
9 by a combination of technology, lifestyle and structural economic changes. Still, such scenarios could
10 result in a dramatic reduction of energy and resource consumption (see further).
11
12 Figure 3.9 Trends in key scenarios characteristics and driving forces as included in the SSP scenarios
13 (showing 5-95th percentiles of the reference scenarios as included in the database in grey shading).
14 Reference (dotted lines) refers to UN low, medium and high population scenario (UN 2019), OECD Long-
15 term economic growth scenario (OECD 2021), the scenarios from IEA’s World Energy Outlook (IEA
16 2019), and the scenarios in the FAO assessment (FAO 2018)
17 Scenarios show a range of possible energy projections. In the absence of climate policy, most
18 scenarios project the final energy demand to continue to grow to around 650-800 EJ yr-1 in 2100
19 (based on the scenario database). Some projections show a very high energy demand up to 1000 EJ yr-
1
20 (comparable to SSP5). The scenario of the IEA lies within the SSP range but near the SSP1
21 projection. However, it should be noted that the IEA scenario includes current policies (most
22 reference scenarios do not) and many scenarios published before 2021 did not account for the Covid-
23 19 crisis. Several researchers discuss the possibility of decoupling material and energy demand from
Do Not Cite, Quote or Distribute 3-25 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 economic growth in the literature, mainly in developed countries (Kemp-Benedict 2018) (decoupling
2 here refers to either a much slower increase in demand or even a decrease). In the scenario literature,
3 this is reflected by scenarios with very low demand for final energy based on increased energy
4 efficiency and less energy-intensive lifestyles (e.g., SSP1 and the LED scenario) (Grubler et al. 2018;
5 van Vuuren et al. 2018). While these studies show the feasibility of such pathways, their energy
6 efficiency improvement rates are considerably above the historic range of around 2% (Haberl et al.
7 2020).
8
9 (Vrontisi et al. 2018; Roelfsema et al. 2020; Sognnaes et al. 2021)(IEA 2021a; Höhne et al.
10 2021)(Jeffery et al. 2018; Gütschow et al. 2018)(Giarola et al. 2021)(Sognnaes et al. 2021)(Höhne et
11 al. 2021)(Höhne et al. 2021)These scenarios also show clear differences in food consumption and the
12 amount of land used for agriculture. Food demand in terms of per capita caloric intake is projected to
13 increase in most scenarios. However, it should be noted that there are large differences in dietary
14 composition across the scenarios (from more meat-intensive in scenarios like SSP5 to a decrease in
15 meat consumptions in other scenarios such as SSP1). Land use projections also depend on assumed
16 changes in yield and the population scenarios. Typically, changes in land use are less drastic than
17 some other parameters (in fact, the 5-95th database range is almost stable). Agriculture land is
18 projected to increase in SSP3, SSP2, and SSP4 – more-or-less stable in SSP5 and is projected to
19 decline in SSP1.
20
21 3.3.2 Emission pathways and temperature outcomes
22
23 3.3.2.1 Overall mitigation profiles and temperature consequences
24
25 Figure 3.10 shows the GHG and CO2 emission trajectories for different temperature categories as
26 defined in Section 3.2 (the temperature levels are calculated using simple climate models, consistent
27 with the outcomes of the recent WGI assessment, Cross-Chapter Box 7.1). It should be noted that
28 most scenarios currently in the literature do not account for an impact of Covid-19 (see Box 3.2). The
29 higher categories (C6 and C7) mostly included scenarios with no or modest climate policy. Because
30 of the progression of climate policy, it is becoming more common that reference scenarios incorporate
31 implemented climate policies. Modelling studies typically implement current or pledged policies up
32 until 2030 (Vrontisi et al. 2018; Roelfsema et al. 2020; Sognnaes et al. 2021) with some studies
33 focusing also on the policy development in the long term (IEA 2021a; Höhne et al. 2021). (Jeffery et
34 al. 2018; Gütschow et al. 2018)Based on the assessment by Chapter 4, reference pathways consistent
35 with the implementation and extrapolation of current policies are associated with increased GHG
36 emissions from 59 (53-65) GtCO2-eq yr-1 in 2019 to 52-60 GtCO2-eq yr-1 by 2030 and to 46-67
37 GtCO2-eq yr-1 by 2050 (see Figure 3.6). Pathways with these near-term emissions characteristics, lead
38 to a median global warming of 2.4°C to 3.5°C by 2100 (see also further in this section). These
39 pathways consider policies at the time that they were developed. A recent model comparison that
40 harmonised socioeconomic, technological, and policy assumptions (Giarola et al. 2021) found a 2.2-
41 2.9°C median temperature rise in 2100 for current and stated policies, with the results sensitive to the
42 model used and the method of implementing policies (Sognnaes et al. 2021). Scenario inference and
43 construction methods using similar policy assumptions leads to a median range of 2.9-3.2°C in 2100
44 for current policies and 2.4-2.9°C in 2100 for 2030 pledges (Höhne et al. 2021). The median spread of
45 1°C across these studies (2.2-3.2°C) indicates the deep uncertainties involved with modelling
46 temperature outcomes of 2030 policies through to 2100. (Höhne et al. 2021)
47
48 The lower categories include increasingly stringent assumed climate policies. For all scenario
49 categories, except the highest category, emissions peak in the 21st century. For the lowest categories,
50 the emissions peak is mostly before 2030. In fact, for scenarios in the category that avoids
51 temperature overshoot for the 1.5oC scenario (C1 category), GHG emissions are reduced already to
52 almost zero around the middle of the century. Typically, CO2 emissions reach net zero about 10-20
53 years before total GHG emissions reach net zero. The main reason is that scenarios reduce non-CO2
54 greenhouse gas emissions less than CO2 due to a limited mitigation potential (see 3.3.2.2). The figure
55 also shows that many scenarios in the literature with a temperature outcome below 2 °C show net
Do Not Cite, Quote or Distribute 3-26 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 negative emissions. There are, however, also exceptions in which more immediate emission
2 reductions limits the need for CDR. The IMPs illustrate alternative pathways to reach the C1-C3
3 temperature levels.
4
5 START BOX HERE
6
7 Box 3.2 Impact of Covid-19 on long-term emissions
8
9 The reduction in CO2 emissions of the Covid-19 pandemic in 2020 was estimated to be about 6% (see
10 section 4.2.2.4 and Table S4.2) lower than 2019 levels (BP 2021, Crippa et al. 2021, IEA 2021, Le
11 Quéré et al. 2021; Friedlingstein et al. 2020; Forster et al. 2020; Liu et al. 2020c). Near-real-time
12 monitoring estimates show a rebound in emissions levels, meaning 2021 emissions levels are
13 expected to be higher than 2020 (Le Quéré et al. 2021). The longer-term effects are uncertain but so
14 far do not indicate a clear structural change for climate policy related to the pandemic. The increase in
15 renewable shares in 2020 could stimulate a further transition, but slow economic growth can also slow
16 down (renewable) energy investments. Also, lifestyle changes during the crisis can still develop in
17 different directions (working from home, but maybe also living further away from work). Without a
18 major intervention, most long-term scenarios project that emission will start to follow a similar
19 pathway as earlier projections (although at a reduced level) (IEA 2020b; Kikstra et al. 2021a;
20 Rochedo et al. 2021). If emissions reductions are limited to only a short time, the adjustment of
21 pathways will lead to negligible outcomes in the order of 0.01K (Forster et al. 2020; Jones et al.
22 2021). At the same time, however, the large amount of investments pledged in the recovery packages
23 could provide a unique opportunity to determine the long-term development of infrastructure, energy
24 systems and land use (Hepburn et al. 2020; Pianta et al. 2021; Andrijevic et al. 2020b). Near-term
25 alternative recovery pathways have been shown to have the potential to influence carbon price
26 pathways, and energy investments and electrification requirements under stringent mitigation targets
27 (Kikstra et al. 2021a; Rochedo et al. 2021; Bertram et al. 2021; Pollitt et al. 2021; Shan et al. 2021).
28 Most studies suggest a noticeable reduction in 2030 emissions. However, much further reductions
29 would be needed to reach the emission levels consistent with mitigation scenarios that would likely
30 stay below 2oC or lower (see Chapter 4). At the moment, the share of investments in greenhouse gas
31 reduction is relatively small in most recovery packages, and no structural shifts for climate policies
32 are observed linked to the pandemic. Finally, most of the scenarios analysed in this Chapter do not
33 include the 2020 emissions reduction related to the Covid-19 pandemic. The effect of the pandemic
34 on the pathways will likely be very small. The assessment of climate mitigation pathways in this
35 chapter should be interpreted as being almost exclusively based on the assumption of a fast recovery
36 with limited persistent effects on emissions or structural changes.
37
38 END BOX HERE
39
Do Not Cite, Quote or Distribute 3-27 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1
2 Figure 3.10 Total emission profiles in the scenarios based on climate category for GHGs (AR6 GWP-100)
3 and CO2. The IMPs are also indicated
4
5 START BOX HERE
6
7 Box 3.3 The likelihood of high-end emission scenarios
8
9 At the time the Representative Concentration Pathways (RCPs) were published, they included 3
10 scenarios that could represent emission developments in the absence of climate policy: RCP4.5, RCP6
11 and RCP8.5, described as, respectively, low, medium and high-end scenarios in the absence of strong
12 climate policy (van Vuuren et al. 2011). RCP8.5 was described as representative of the top 5%
13 scenarios in the literature. The SSPs-based set of scenarios covered the RCP forcing levels adding a
14 new low scenario (at 1.9 W/m2). Hausfather and Peters (2020) pointed out that since 2011, the rapid
15 development of renewable energy technologies and emerging climate policy have made it
16 considerably less likely that emissions could end up as high as RCP8.5. Still, emission trends in
17 developing countries track RCP8.5 Pedersen et al. (2020), and high land-use emissions could imply
18 that emissions would continue to do so in the future, even at the global scale (Schwalm et al. 2020).
19 Other factors resulting in high emissions include higher population or economic growth as included in
20 the SSPs (see subsection 3.3.1) or rapid development of new energy services. Climate projections of
21 RCP8.5 can also result from strong feedbacks of climate change on (natural) emission sources and
22 high climate sensitivity (see WGI Chapter 7), and therefore their median climate impacts might also
23 materialise while following a lower emission path (e.g., Hausfather and Betts (2020)). The discussion
24 also relates to a more fundamental discussion on assigning likelihoods to scenarios, which is
25 extremely difficult given the deep uncertainty and direct relationship with human choice. However, it
26 would help to appreciate certain projections (e.g., Ho et al. (2019)). All-in-all, this means that high-
27 end scenarios have become considerably less likely since AR5 but cannot be ruled out. It is important
28 to realize that RCP8.5 and SSP5-8.5 do not represent a typical ‘business-as-usual projection but are
29 only useful as high-end, high-risk scenarios. Reference emission scenarios (without additional climate
30 policy) typically end up in C5-C7 categories included in this assessment.
31
Do Not Cite, Quote or Distribute 3-28 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 END BOX HERE
2
3 Figure 3.11: Global mean temperature outcome of the ensemble of scenarios included in the climate
4 categories C1-C7 (based on RCM calibrated to the WGI assessment, both in terms of future and historic
5 warming). The left panel shows the ranges of scenario uncertainty (shaded area) with the P50 RCM
6 probability (line). The right panel shows the P5 to P95 range of RCM climate uncertainty (C1-C7 is
7 explained in Table 3.1) and the P50 (line) and P66 (dashed line).
8
9 Figure 3.11 shows the possible consequences of the different scenario categories for global mean
10 temperature calculated using a reduced complexity model calibrated to the IPCC WGI assessment
11 (see Annex III and WGI report). For the C5-C7 categories (containing most of the reference and
12 current policy scenarios), the global mean temperature is expected to increase throughout the century
13 (and further increase will happen after 2100 for C6 and C7). While warming would likely be in the
14 range from 2.2-3.8 °C – warming above 5°C cannot be excluded. The highest emissions scenarios in
15 the literature combine assumptions about rapid long-term economic growth and pervasive climate
16 policy failures, leading to a reversal of some recent trends (see box 3.3). For the categories C1-C4, a
17 peak in global mean temperature is reached mid-century for most scenarios in the database, followed
18 by a small (C3/C4) or more considerable decline (C1/C2). There is a clear distinction between the
19 scenarios with no or limited overshoot (typically <0.1 °C, C1) compared to those with high overshoot
20 (C2): in emissions, the C1 category is characterised by steep early reductions and a relatively small
21 contribution of net negative emissions (like LD and Ren) (Figure 3.10). In addition to the temperature
22 caused by the range of scenarios in each category (main panel), also climate uncertainties contribute
23 to a range of temperature outcomes (including uncertainties regarding the carbon cycle, climate
24 sensitivity, and the rate of change, see WGI). The bars on the right of Figure 3.11 show the
25 uncertainty range for each category (combining scenario and uncertainty). While the C1 category
26 more likely than not leads to warming below 1.5 °C by the end of the century, even with such a
27 scenario, warming above 2oC cannot be excluded (95th percentile). The uncertainty range for the
28 highest emission categories (C7) implies that these scenarios could lead to a warming above 6oC.
29
30 3.3.2.2 The role of carbon dioxide and other greenhouse gases
31 The trajectory of future CO2 emissions plays a critical role in mitigation, given CO2 long-term impact
32 and dominance in total greenhouse gas forcing. As shown in Figure 3.12, CO2 dominates total
33 greenhouse gas emissions in the high emissions scenarios but is also reduced most, going from
34 scenarios in the highest to lower categories. In C4 and below, most scenarios exhibit net negative CO2
35 emissions in the second half of the century compensating for some of the residual emissions of non-
36 CO2 gases as well as reducing overall warming from an intermediate peak. Still, early emission
Do Not Cite, Quote or Distribute 3-29 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 reductions and further reductions in non-CO2 emissions can also lead to scenarios without net
2 negative emissions in 2100, even in C1 and C3 (shown for the 85-95th percentile). In C1, avoidance of
3 significant overshoot implies that immediate gross reductions are more relevant than long-term net
4 negative emissions (explaining the lower number than in C2) but carbon dioxide removal is still
5 playing a role in compensating for remaining positive emissions in hard-to-abate sectors.
6
7 Figure 3.12 Upper left: The role of CO2 and other greenhouse gases. Emission in CO 2-eq in 2100 (using
8 AR6 GWP-100) (other = halogenated gases) and upper right: Cumulative CO 2 emissions in the 2020-2100
9 period. The lower left and right panel show the development of CH4 and N2O emissions over time.
10 Energy emissions include the contribution of BECCS. For both energy and AFOLU sectors, the positive
11 and negative values represent the cumulated annual balances. In both panels, the three bars per scenario
12 category represent the lowest 5-15th percentile, the average value and the highest 5-15th percentile. These
13 illustrate the range of scenarios in each category. The definition of C1-C7 can be found in Table 3.1
14 CH4 and N2O emissions are also reduced from C7 to C1, but this mostly occurs between C7 and C5.
15 The main reason is the characteristics of abatement potential: technical measures can significantly
16 reduce CH4 and N2O emissions at relatively low costs to about 50% of the current levels (e.g., by
17 reducing CH4 leaks from fossil fuel production and transport, reducing landfill emissions gazing, land
18 management and introducing measure related to manure management, see also Chapter 7 and 11).
19 However, technical potential estimates becomes exhausted even if the stringency of mitigation is
20 increased (Harmsen et al. 2019a,b; Höglund-Isaksson et al. 2020). Therefore, further reduction may
21 come from changes in activity levels, such as switching to a less meat-intensive diet reducing
22 livestock (Stehfest et al. 2009; Willett et al. 2019; Ivanova et al. 2020) (see also Chapter 7). Other
23 non-CO2 GHG emissions (halogenated gases) are reduced to low levels for scenarios below 2.5oC.
24
Do Not Cite, Quote or Distribute 3-30 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 Short-lived climate forcers (SLCFs) also play an important role in climate change, certainly for short-
2 term changes (see Figure SPM.2, WGI) (Shindell et al. 2012). These forcers consist of 1) substances
3 contributing to warming, such as methane, black carbon and tropospheric ozone and 2) substances
4 contributing to cooling (other aerosols, such as related to sulphur emissions). Most SLCFs are also air
5 pollutants, and reducing their emissions provides additional co-benefits (Shindell et al. 2017a,b;
6 Hanaoka and Masui 2020). In the case of the first group, emission reduction thus leads to both air
7 pollution and climate benefits. For the second, group there is a possible trade-off (Shindell and Smith
8 2019; Lund et al. 2020). As aerosol emissions are mostly associated with fossil fuel combustion, the
9 benefits of reducing CO2 could, in the short term, be reduced as a result of lower aerosol cooling.
10 There has been an active discussion on the exact climate contribution of SLCF focused policies in the
11 literature. This discussion partly emerged from different assumptions on possible reductions in the
12 absence of ambitious climate policy and the uncertain global climate benefit from aerosol (black
13 carbon) (Rogelj et al. 2014). The latter is now assessed to be smaller than originally thought (Smith et
14 al. 2020b; Takemura and Suzuki 2019) (see also WGI Chapter 6, Section 6.4). Reducing SLCF
15 emissions is critical to meet long-term climate goals and might help reduce the rate of climate change
16 in the short term. Deep SLCF emission reductions also increase the remaining carbon budget for a
17 specific temperature goal (Rogelj et al. 2015a; Reisinger et al. 2021) (see Box 3.4). A more detailed
18 discussion can be found in WGI Chapter 5 and 6.
19
20 For accounting of emissions and the substitution of different gases as part of a mitigation strategy,
21 typically, emission metrics are used to compare the climate impact of different gases. Most policies
22 currently use Global Warming Potentials with a 100-year time horizon as this is also mandated for
23 emissions reporting in the Paris Rulebook (for a wider discussion of GHG metrics, see Box 2.2 in
24 Chapter 2 and WGI, Chapter 7, Section 7.6). Alternative metrics have also been proposed, such as
25 those using a shorter or longer time horizon or those that focus directly on the consequences of
26 reaching a certain temperature target (Global Temperature Change Potential, GTP), allowing a more
27 direct comparison with cumulative CO2 emissions (Allen et al. 2016; Lynch et al. 2020) or focusing
28 on damages (Global Damage Potential) (an overview is given in Chapter 2, and Cross-Chapter Box
29 3). Depending on the metric, the value attributed to reducing short-lived forcers like methane can be
30 lower in the near term (e.g., in the case of GTP) or higher (GWP with short reference period). For
31 most metrics, however, the impact on mitigation strategies is relatively small, among others, due to
32 the marginal abatement cost curve of methane (low costs for low to medium mitigation levels;
33 expensive for high levels). The timing of reductions across different gases impacts warming and the
34 co-benefits (Harmsen et al. 2016; Cain et al. 2019). Nearly all scenarios in the literature use GWP-100
35 in cost-optimisation, reflecting the existing policy approach; the use of GWP-100 deviates from cost-
36 optimal mitigation pathways by at most a few percent for temperature goals of likely below 2°C and
37 lower (see Box 2.2 in Chapter 2).
38
39 Cumulative CO2 emissions and temperature goals
40 The dominating role of CO2 and its long lifetime in the atmosphere and some critical characteristics of
41 the earth system implies that there is a strong relationship between cumulative CO2 emissions and
42 temperature outcomes (MacDougall and Friedlingstein 2015; Meinshausen et al. 2009; Allen et al.
43 2009; Matthews et al. 2009). This is illustrated in Figure 3.13 that plots the cumulative CO2 emissions
44 against the projected outcome for global mean temperature, both until a temperature peak and full
45 century. The deviations from in linear relationship in Figure 3.13 are mostly caused by different non-
46 CO2 emission and forcing levels (see also Rogelj et al. (2015b)). This means that reducing non-CO2
47 emissions can play an important role in limiting peak warming: the smaller the residual non-CO2
48 warming, the larger the carbon budget. This impact on carbon budgets can be substantial for stringent
49 warming limits. For 1.5°C pathways, variations in non-CO2 warming across different emission
50 scenarios have been found to vary the remaining carbon budget by approximately 220 GtCO 2 (see
51 WGI Chapter 5, Subsection 5.5.2.2). In addition to reaching net zero CO2 emissions, a strong
52 reduction in methane emissions is the most critical component in non-CO2 mitigation to keep the Paris
53 climate goals in reach (van Vuuren et al. 2018; Collins et al. 2018) (see also WGI, Chapter 5, 6 and
54 7). It should be noted that the temperature categories (C1-C7) generally aligned with the horizontal
55 axis, except for the end-of-century values for C1 and C2 that coincide.
Do Not Cite, Quote or Distribute 3-31 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 START BOX HERE
2
3 Box 3.4 Consistency of remaining carbon budgets in the WGI assessment and cumulative CO2
4 emissions in WGIII mitigation pathways
5 Introduction
6 The WGI assessment has shown that the increase in global mean temperature has a near-linear
7 relationship with cumulative CO2 emissions (Chapter 5, Section 5.5, Box 5.3). Consistently, WGI has
8 confirmed that net zero CO2 emissions are required to halt CO2-induced warming. This permits the
9 estimation of carbon budgets consistent with specific temperature goals. In Chapter 3, we present the
10 temperature outcomes and cumulative CO2 emissions associated with different warming levels for
11 around 1200 scenarios published in the literature and which were classified according to different
12 warming levels (see Section 3.2 and Annex III, Part II.2.5). In this box, we discuss the consistency of
13 the assessments presented here and in IPCC WGI. The box summarises how the remaining carbon
14 budgets assessed by WGI relate to the remaining cumulative CO2 emissions until the time of net zero
15 CO2 emissions in mitigation pathways (Table 3.2, Table SPM1) assessed by WGIII.
16
17 In its assessment, WGI uses a framework in which the various components of the remaining carbon
18 budget are informed by various lines of evidence and assessed climate system characteristics. WGIII,
19 instead, uses around 1200 emission scenarios with estimated warming levels that cover the scenario
20 range presented in WGI but also contain many more intermediate projections with varying emission
21 profiles and a combination of CO2 emissions and other greenhouse gases. In order to assess their
22 climate outcomes, climate model emulators are used. The emulators are reduced complexity climate
23 models that are provided by WGI, and which are calibrated to the WGI assessment of future warming
24 for various purposes (a detailed description of the use of climate model emulators in the WGI and
25 WGIII assessments can be found in Cross-chapter Box 7.1 in the WGI report, with the connection of
26 WGI and WGII discussed in Annex III.2.5.1).
27
28 Remaining carbon budgets estimated by WGI
29 WGI estimated the remaining carbon budgets from their assessment of (i) the transient climate
30 response to cumulative emissions of carbon dioxide (TCRE), and estimates of (ii) the historical
31 human-induced warming, (iii) the temperature change after reaching net zero CO2 emissions, (iv) the
32 contribution of future non-CO2 warming (derived from the emissions scenarios assessed in the Special
33 Report on 1.5°C Warming using WGI-calibrated emulators), and (v) the earth system feedbacks (WGI
34 Chapter 5.5, Box 5.2). For a given warming level, WGI assessed the remaining carbon budget from
35 the beginning of 2020 onwards. These are 650 / 500 / 400 GtCO2 for limiting warming to 1.5°C with
36 33% / 50% / 67% chance and 1350 / 1150 GtCO2 for limiting warming to 2°C with 50% / 67%
37 chance. The estimates are subject to considerable uncertainty related to historical warming, future
38 non-CO2 forcing, and poorly quantified climate feedbacks. For instance, variation in non-CO2
39 emissions across scenarios are estimated to either increase or decrease the remaining carbon budget
40 estimates by 220 GtCO2. The estimates of the remaining carbon budget assume that non-CO2
41 emissions are reduced consistently with the tight temperature targets for which the budgets are
42 estimated.
43
44 Cumulative CO2 emissions until net zero estimated by WGIII
45 WGIII provides estimates of cumulative net CO2 emissions (from 2020 inclusive) until the time of
46 reaching net zero CO2 emissions (henceforth called “peak cumulative CO2 emissions”) and until the
47 end of the century for eight temperature classes that span a range of warming levels. The numbers can
48 be found in Table 3.2 (330-710 GtCO2 for C1; 540-930 for C2 and 640-1160 for C3).
49
50 Comparing the WGI remaining carbon budgets and remaining cumulative CO2 emissions of the
51 WGIII scenarios
52 A comparison between WGI and WGIII findings requires recognising that, unlike in WGI, cumulative
53 emissions in WGIII are not provided for a specific peak warming threshold or level but are instead
54 provided for a set of scenarios in a category, representing a specific range of peak temperature
Do Not Cite, Quote or Distribute 3-32 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 outcomes (for instance the C4 category contains scenarios with a median peak warming anywhere
2 between approximately 1.8°C and up to 2°C). When accounting for this difference, the WGI and
3 WGIII findings are very consistent for temperature levels below 2oC. Figure 1 compares the peak
4 temperatures and associated cumulative CO2 emissions (i.e., peak cumulative CO2 emissions) for the
5 WGIII scenarios to the remaining carbon budgets assessed by WGI. This shows only minor
6 differences between the WGI and WGIII approaches.
7
8
9
10 Box 3.4 Figure 1: Cumulative CO2 emissions from AR6 scenario categories (coloured dots), adjusted for
11 distinct 0.1°C warming levels (black bars) in comparison to the WGI remaining carbon budgets (grey
12 bars). The cumulative carbon emissions for the AR6 scenarios are shown for the median peak warming
13 (panel a), the 33rd-percentile peak warming (panel b) and the upper 67th-percentile peak warming (panel
14 c) calculated with the WGI-calibrated emulator MAGICC7 (IPCC AR6 WGI, Cross-Chapter Box 7.1).
15 The adjustment to the nearest 0.1°C intervals is made using WGI TCRE (at the relevant percentile, e.g.,
16 the 67th-percentile TCRE is used to adjust the 67th-percentile peak warming), with the 5% to 95% range
17 of adjusted scenarios provided by the black bar. The WGI remaining carbon budget is shown, including
18 the WGI estimate of at least a ±220 GtCO2 uncertainty due to non-CO2 emissions variations across
19 scenarios (grey bars). For median peak warming (panel a) projections below 2°C relative to 1850-1900,
20 the WGIII assessment of cumulative carbon emissions tends to be slightly smaller than the remaining
21 carbon budgets provided by WGI but well within the uncertainties. Note that only a few scenarios in
22 WGIII limit warming to below 1.5°C with a 50% chance, thus statistics for that specific threshold have
23 low confidence.
24
25 After correcting for the categorisation, some (small) differences between the WGI and WGIII
26 numbers arise from remaining differences between the outcomes of the climate emulators and their
27 set-up (see Cross-Chapter Box 7.1 in IPCC WGI AR6) and the differences in the underlying
28 scenarios. Moreover, the WGI assessment estimated the non-CO2 warming at the time of net zero CO2
29 emissions based on a relationship derived from the SR1.5 scenario database with historical emission
30 estimates as in Meinshausen et al. (2020) (WGI chapter 5). The WGIII assessment uses the same
31 climate emulator with improved historical emissions estimates (Nicholls et al. 2021) (WGI Cross-
32 Chapter Box 7.1). Annex III.2.5.1 further explores the effects of these factors on the relationship
33 between non-CO2 warming at peak cumulative CO2 and peak surface temperature.
34
35 Estimates of the remaining carbon budgets thus vary with the assumed level of non-CO2 emissions,
36 which are a function of policies and technology development. The linear relationship used in the WGI
37 assessment between peak temperature and the warming as a result of non-CO2 emissions (based on
38 the SR1.5 data) is shown in the right panel of Figure 2 (dashed line). In the WG3 approach, the non-
39 CO2 warming for each single scenario is based on the individual scenario characteristics. This shown
Do Not Cite, Quote or Distribute 3-33 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 in the same figure by plotting the outcomes of scenario outcomes of a range models (dots). The lines
2 show the fitted data for individual models, emphasizing the clear differences across models and the
3 relationship with peak warming (policy level). In some scenarios stringent non-CO2 emission
4 reductions provide an option to reach more stringent climate goals with the same carbon budget. This
5 is especially the case for scenarios with a very low non-CO2 warming, for instance as a result of
6 methane reductions through diet change. The left panel shows how these differences impact estimates
7 of the remaining carbon budget. While the WGIII AR6 scenario database includes a broad range of
8 non-CO2 emission projections the overall range is still very consistent with the WGI relationship and
9 the estimated uncertainty with a ±220 GtCO2 range (see also panel B of Figure II.2 in Annex III.2.5).
10
11
12 Box 3.4 Figure 2: Panel A) Differences in regressions of the relationship between peak surface
13 temperature and associated cumulative CO2 emissions from 2020 derived from scenarios of eight
14 integrated assessment model frameworks. The coloured lines show the regression at median for scenarios
15 of the 8 modelling frameworks, each with more than 20 scenarios in the database and a detailed land-use
16 representation. The red dotted lines indicate the non-CO2 uncertainty range of WGI Chapter 5 (± 220
17 GtCO2), here visualised around the median of the 8 model framework lines. Carbon budgets from 2020
18 until 1.5C (0.43K above 2010-2019 levels) and 2.0C (0.93K above 2010-2019 levels) are shown for
19 minimum and maximum model estimates at the median, rounded to the nearest 10GtCO2. Panel B)
20 Shows the relationship between the estimated non-CO2 warming in mitigation scenarios that reach net
21 zero and the associated peak surface temperature outcomes. The coloured lines show the regression at
22 median for scenarios of the 8 modelling frameworks with more than 20 scenarios in the database and a
23 detailed land-use representation. The black dashed line indicates the non-CO2 relationship based on the
24 scenarios and climate emulator setup as was assessed in WGI Chapter 5.
25 Overall, the slight differences between the cumulative emissions in WGIII and the carbon budget in
26 WGI are because the non-CO2 warming in the WGIII AR6 scenarios is slightly lower than in the
27 SR1.5 scenarios that are used for the budget estimates in WGI (Annex III.2.5.1). In addition,
28 improved consistency with Cross-Chapter Box 7.1 in WGI results in a non-CO2 induced temperature
29 difference of about ~0.05K between the assessments. Re-calculating the remaining carbon budget
30 using the WGI methodology combined with the full WGIII AR6 scenario database results in a
31 reduction of the estimated remaining 1.5C carbon budget by about 100 GtCO2 (-20%), and a reduction
32 of ~40 GtCO2 (-3%) for 2C. Accounting also for the categorisation effect, the difference between the
33 WGI and WGIII estimates is found to be small and well within the uncertainty range (Figure 1). This
34 means that the cumulative CO2 emissions presented in WGIII and the WGI carbon budgets are highly
35 consistent.
Do Not Cite, Quote or Distribute 3-34 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 A detailed comparison of the impact of different assessment steps, i.e., the new emulators, scenarios,
2 and harmonisation methods, has been made and is presented in Annex III.
3
4 Policy implications
5 The concept of a finite carbon budget means that the world needs to get to net zero CO2, no matter
6 whether global warming is limited to 1.5°C or well below 2°C (or any other level). Moreover,
7 exceeding the remaining carbon budget will have consequences by overshooting temperature levels.
8 Still, the relationship between the timing of net zero and temperature targets is a flexible one, as
9 discussed further in Cross-Chapter Box 3. It should be noted that national level inventory as used by
10 UNFCCC for the land use, land-use change and forestry sector are different from the overall concept
11 of anthropogenic emissions employed by IPCC WG1. For emissions estimates based on these
12 inventories, the remaining carbon budgets must be correspondingly reduced by approximately 15%,
13 depending on the scenarios (Grassi et al., 2021) (see also Chapter 7).
14
15 One of the uncertainties of the remaining carbon budget is the level of non-CO2 emissions which is a
16 function of policies and technology development. This represents a point of leverage for policies
17 rather than an inherent geophysical uncertainty. Stringent non-CO2 emission reductions hence can
18 provide – to some degree – an option to reach more stringent climate goals with the same carbon
19 budget.
20
21 END BOX HERE
22
23 Figure 3.13: The near-linear relationship between cumulative CO2 emissions and temperature. The left
24 panel shows cumulative emissions until net zero emission is reached. The right panel shows cumulative
25 emissions until the end of the century, plotted against peak and end-of-century temperature, respectively.
26 Both are shown as a function of non-CO2 forcing and cumulative net negative CO2 emissions. Position
27 temperature categories (circles) and IPs are also indicated, including two 2oC sensitivity cases for Neg
28 (Neg-2.0) and Ren (Ren-2.0).
29 The near-linear relationship implies that cumulative CO2 emissions are critically important for climate
30 outcomes (Collins et al. 2013). The maximum temperature increase is a direct function of the
31 cumulative emissions until net zero CO2 emissions is reached (the emission budget) (Figure 3.13, left
32 side). The end-of-century temperature correlates well with cumulative emissions across the century
Do Not Cite, Quote or Distribute 3-35 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 (right panel). For long-term climate goals, positive emissions in the first half of the century can be
2 offset by net removal of CO2 from the atmosphere (net negative emissions) at the cost of a temporary
3 overshoot of the target (Tokarska et al. 2019). The bottom panels of Figure 3.13 show the contribution
4 of net negative CO2 emissions.
5
6 Focusing on cumulative emissions, the right-hand panel of Figure 3.12 shows that for high-end
7 scenarios (C6-C7), most emissions originate from fossil fuels, with a smaller contribution from net
8 deforestation. For C5 and lower, there is also a negative contribution to emissions from both AFOLU
9 emissions and energy systems. For the energy systems, these negative emissions originate from bio-
10 energy-and-carbon-capture-and-storage (BECCS), while for AFOLU, they originate from re- and
11 afforestation. For C3-C5, reforestation has a larger CDR contribution than BECCS, mostly due to
12 considerably lower costs (Rochedo et al. 2018). For C1 and C2, the tight carbon budgets imply in
13 many scenarios more CDR use (Riahi et al. 2021). Please note that net negative emissions are not so
14 relevant for peak temperature targets, and thus the C1 category, but CDR can still be used to offset the
15 remaining positive emissions (Riahi et al. 2021). While positive CO2 emissions from fossil fuels are
16 significantly reduced, inertia and hard-to-abate sectors imply that in many C1-C3 scenarios, around
17 800-1000 GtCO2 of net positive cumulative CO2 emissions remain. This is consistent with literature
18 estimates that current infrastructure is associated with 650 GtCO2 (best estimate) if operated until the
19 end of its lifetime (Tong et al. 2019). These numbers are considerably above the estimated carbon
20 budgets for 1.5 °C estimated in WGI, hence explaining CDR reliance (either to offset emissions
21 immediately or later in time).
22
23 Creating net negative emissions can thus be an important part of a mitigation strategy to offset
24 remaining emissions or compensate for emissions earlier in time. As indicated above, there are
25 different ways to potentially achieve this, including re- and afforestation and BECCS (as often
26 covered in IAMs) but also soil carbon enhancement, direct air carbon capture and storage (DACCS)
27 and ocean alkalinization (see Chapter 12). Except for reforestation, these options have not been tested
28 at large scale and often require more R&D. Moreover, the reliance on CDR in scenarios has been
29 discussed given possible consequences of land use related to biodiversity loss and food security
30 (BECCS and afforestation), the reliance on uncertain storage potentials (BECCS and DACCS), water
31 use (BECCS), energy use (DACCS), the risks of possible temperature overshoot and the
32 consequences for meeting sustainable development goals (Venton 2016; Peters and Geden 2017;
33 Smith et al. 2016; van Vuuren et al. 2017; Anderson and Peters 2016; Honegger et al. 2021). In the
34 case of BECCS, it should be noted that bio-energy typically is associated with early-on positive CO2
35 emissions and net-negative effects are only achieved in time (carbon debt), and its potential is limited
36 (Cherubini et al. 2013; Hanssen et al. 2020) (most IAMs have only a very limited representation of
37 these time dynamics). Several scenarios have therefore explored how reliance on net negative CO2
38 emissions can be reduced or even avoided by alternative emission strategies (Grubler et al. 2018; van
39 Vuuren et al. 2018) or early reductions by more stringent emission reduction in the short-term (Rogelj
40 et al. 2019b; Riahi et al. 2021). A more in-depth discussion of land-based mitigation options can be
41 found in Chapter 7. It needs to be emphasized that even in strategies with net negative CO2 emissions,
42 the emission reduction via more conventional mitigation measures (efficiency improvement,
43 decarbonisation of energy supply) is much larger than the CDR contribution (Tsutsui et al. 2020).
44
45 3.3.2.3 The timing of net zero emissions
46
47 In addition to the constraints on change in global mean temperature, the Paris Agreement also calls for
48 reaching a balance of sources and sinks of GHG emissions (Art. 4). Different interpretations of the
49 concept related to balance have been published (Rogelj et al. 2015c; Fuglestvedt et al. 2018). Key
50 concepts include that of net zero CO2 emissions (anthropogenic CO2 sources and sinks equal zero)
51 and net zero greenhouse gas emission (see also Annex I Glossary and Box 3.3). The same notion can
52 be used for all GHG emissions, but here ranges also depend on the use of equivalence metrics
53 (Chapter 2, Box 2.2). Moreover, it should be noted that while reaching net zero CO2 emissions
54 typically coincides with the peak in temperature increase; net zero GHG emissions (based on GWP-
55 100) implies a decrease in global temperature (Riahi et al. 2021) and net zero GHG emission typically
Do Not Cite, Quote or Distribute 3-36 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 requires negative CO2 emissions to compensate for the remaining emissions from other GHGs. Many
2 countries have started to formulate climate policy in the year that net zero emissions (either CO2 or all
3 greenhouse gases) are reached – although, at the moment, formulations are often still vague (Rogelj et
4 al. 2021). There has been increased attention on the timing of net zero emissions in the scientific
5 literature and ways to achieve it.
6
7 Figure 3.14 shows that there is a relationship between the temperature target, the cumulative CO2
8 emissions budget, and the net zero year for CO2 emissions (left) and the sum of greenhouse gases
9 (right) for the scenarios published in the literature. In other words, the temperature targets from the
10 Paris Agreement can, to some degree, be translated into a net zero emission year (Tanaka and O’Neill
11 2018). There is, however, a considerable spread. In addition to the factors influencing the emission
12 budget (see WGI and section 3.3.2.2), this is influenced by the emission trajectory until net zero is
13 reached, decisions related to temperature overshoot and non-CO2 emissions (especially for the
14 moment CO2 reaches net zero emissions). Scenarios with limited or no net negative emissions and
15 rapid near-term emission reductions can allow small positive emissions (e.g., in hard-to-abate-
16 sectors). They may therefore have a later year that net zero CO2 emissions are achieved. High
17 emissions in the short-term, in contrast, require an early net zero year.
18
19 For the scenarios in the C1 category (warming below 1.5°C (50% probability) with limited
20 overshoot), the net zero year for CO2 emissions is typically around 2035-2070 For scenarios in C3
21 (likely limiting warming to below 2°C), CO2 emissions reach net zero around 2060-2100. Similarly,
22 also the years for net zero GHG emissions can be calculated (see right graph). The GHG net zero
23 emissions year is typically around 10-20 years later than the carbon neutrality. Residual non-CO2
24 emissions at the time of reaching net zero CO2 range between 4-11 GtCO2-eq in pathways that likely
25 limit warming to 2.0°C or below. In pathways likely limiting warming to 2°C, methane is reduced by
26 around 20% (1-46%) in 2030 and almost 50% (26-64%) in 2050, and in pathways limiting warming to
27 1.5°C with no or limited overshoot by around 33% (19-57) in 2030 and a similar 50% (33-69%) in
28 2050. Emissions reduction potentials assumed in the pathways become largely exhausted when
29 limiting warming to below 2°C. N2O emissions are reduced too, but similar to CH4, emission
30 reductions saturate for stringent climate goals. In the mitigation pathways, the emissions of cooling
31 aerosols are reduced due to reduced use of fossil fuels. The overall impact on non-CO2-related
32 warming combines these factors.
33 In cost-optimal scenarios, regions will mostly achieve net zero emissions as a function of options for
34 emission reduction, CDR, and expected baseline emission growth (van Soest et al. 2021b). This
35 typically implies relatively early net zero emission years in scenarios for the Latin America region and
36 relatively late net zero years for Asia and Africa (and average values for OECD countries). However,
37 an allocation based on equity principles (such as responsibility, capability and equality) might result
38 in different carbon neutrality years, based on the principles applied – with often earlier net zero years
39 for the OECD (Fyson et al. 2020; van Soest et al. 2021b). Therefore, the emission trajectory until net
40 zero emissions is a critical determinant of future warming (see Section 3.5). The more CO 2 is emitted
41 until 2030, the less CO2 can be emitted after that to stay below a warming limit (Riahi et al. 2015). As
42 discussed before, also non-CO2 forcing plays a key role in the short term.
43
44 START CCB HERE
45
46 Cross-Chapter Box 3: Understanding net zero CO2 and net zero GHG emissions
47 Authors: Elmar Kriegler (Germany), Alaa Al Khourdajie (United Kingdom/Syria), Edward Byers
48 (Ireland/Austria), Katherine Calvin (the United States of America), Leon Clarke (the United States of
49 America), Annette Cowie (Australia), Navroz Dubash (India), Jae Edmonds (the United States of
50 America), Jan S. Fuglestvedt (Norway), Oliver Geden (Germany), Giacomo Grassi (Italy/European
51 Union), Anders Hammer Strømman (Norway), Frank Jotzo (Australia), Alexandre Köberle
52 (Brazil/United Kingdom), Frank Lecocq (France), Yun Seng Lim (Malaysia), Eric Masanet (the
53 United States of America), Toshihiko Masui (Japan), Catherine Mitchell (United Kingdom), Gert-Jan
54 Nabuurs (the Netherlands), Anthony Patt (the United States of America/Switzerland), Glen Peters
Do Not Cite, Quote or Distribute 3-37 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 (Australia/Norway), Andy Reisinger (New Zealand), Keywan Riahi (Austria), Joeri Rogelj (United
2 Kingdom/Belgium), Yamina Saheb (France/ Algeria), Jim Skea (United Kingdom), Detlef van
3 Vuuren (the Netherlands), Harald Winkler (Republic of South Africa)
4
5 This cross-chapter box surveys scientific, technical and policy aspects of net zero carbon dioxide
6 (CO2) and net zero greenhouse gas (GHG) emissions, with a focus on timing, the relationship with
7 warming levels, and sectoral and regional characteristics of net zero emissions. Assessment of net
8 zero GHG emissions additionally requires consideration of non-CO2 gases and choice of GHG
9 emission metrics used to aggregate emissions and removals of different GHGs (Cross-Chapter Box 2
10 in Chapter 2; Cross-Chapter Box 7 in Chapter 10). The following considers net zero CO2 and GHG
11 emissions globally, followed by regional and sectoral dimensions.
12
13 Net zero CO2
14 Reaching net zero CO2 emissions globally is necessary for limiting global warming to any level.
15 At the point of net zero CO2, the amount of CO2 human activity is putting into the atmosphere equals
16 the amount of CO2 human activity is removing from the atmosphere (see Glossary). Reaching and
17 sustaining net zero CO2 emissions globally stabilizes CO2-induced warming. Reaching net zero CO2
18 emissions and then moving to net negative CO2 emissions globally leads to a peak and decline in
19 CO2-induced warming (WGI AR6 Chapter 5.5, 5.6).
20 Limiting warming to 1.5°C or likely to 2°C requires deep, rapid, and sustained reductions of
21 other greenhouse gases including methane alongside rapid reductions of CO2 emissions to net
22 zero. This ensures that the warming contributions from non-CO2 forcing agents as well as from CO2
23 emissions are both limited at low levels. WGI estimated remaining carbon budgets until the time of
24 reaching net zero CO2 emissions for a range of warming limits, taking into account historical CO2
25 emissions and projections of the warming from non-CO2 forcing agents (WGI AR6 Chapter 5.5,
26 Cross-chapter box 3).
27 The earlier global net zero CO2 emissions are reached, the lower the cumulative net amount of
28 CO2 emissions and human-induced global warming, all else being equal (Figure 1a in this Box).
29 For a given net zero date, a variation in the shape of the CO2 emissions profile can lead to a variation
30 in the cumulative net amount of CO2 emissions until the time of net zero CO2 and as a result to
31 different peak warming levels. For example, cumulative net CO2 emissions until the time of reaching
32 net zero CO2 will be smaller, and peak warming lower, if emissions are reduced steeply and then more
33 slowly compared to reducing emissions slowly and then more steeply (Figure 1b in this Box).
34 Net zero CO2 emissions are reached between 2050-2055 (2035-2070) in global emissions
35 pathways limiting warming to 1.5°C with no or limited overshoot, and between 2070-2075
36 (2060-…) in pathways likely limiting warming to 2°C as reported in the AR6 scenario database
37 (median five-year interval and 5th-95th percentile ranges)3. The variation of non-CO2 emissions in 1.5-
38 2°C pathways varies the available remaining carbon budget which can move the time of reaching net
39 zero CO2 in these pathways forward or backward4. The shape of the CO2 emissions reduction profile
FOOTNOTE 3 A small fraction of pathways in the AR6 scenarios database that likely limit warming to 2°C
(9%) or are as likely as not to limit warming to 2°C (14%) do not reach net-zero CO2 emissions during the 21st
century. This is not inconsistent with the fundamental scientific requirement to reach net-zero CO2 emissions for
a stable climate, but reflects that in some pathways, concurrent reductions in non-CO2 emissions temporarily
compensate for on-going warming from CO2 emissions. These would have to reach net-zero CO2 emissions
eventually after 2100 to maintain these warming limits. For the two classes of pathways, the 95th percentile
cannot be deduced from the scenario database as more than 5% of them do not reach net zero CO2 by 2100.
FOOTNOTE 4 WGI Chapter 5.5 estimates a variation of the remaining carbon budget by ± 220 GtCO2 due to
variations of the non-CO2 warming contribution in 1.5-2°C pathways. This translates to a shift of the timing of
Do Not Cite, Quote or Distribute 3-38 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 also affects the time of reaching net zero CO2 (Figure 1c in this Box). Global emission pathways that
2 more than halve CO2 emissions from 2020 to 2030 can follow this rapid reduction by a more gradual
3 decline towards net zero CO2 and still limit warming to 1.5°C with no or limited overshoot, reaching
4 the point of net zero after 2050. The literature since SR1.5 included a larger fraction of such pathways
5 than were available at the time of SR1.5. This is the primary reason for the small backward shift in the
6 median estimate of reaching global net zero CO2 emissions in 1.5°C pathways collected in the AR6
7 scenario database compared to SR1.5. This does not mean that the world is assessed to have more
8 time to rapidly reduce current emissions levels compared to SR1.5. The assessment of emissions
9 reductions by 2030 and 2040 in pathways limiting warming to 1.5°C with no or limited overshoot has
10 not changed substantially. It only means that the exact timing of reaching net zero CO2 after a steep
11 decline of CO2 emissions until 2030 and 2040 can show some variation, and the SR1.5 median value
12 of 2050 is still close to the middle of the current range (Figure 1c in this Box ).
13
14 Cross-Chapter Box 3 Figure 1: Selected global CO2 emissions trajectories with similar shape and
15 different net zero CO2 date (Panel a), different shape and similar net zero CO 2 date (Panel b), and similar
16 peak warming, but varying shapes and net zero CO 2 dates (Panel c). Funnels show pathways limiting
17 warming to 1.5°C with no or limited overshoot (light blue) and likely limiting warming to 2°C (beige).
18 Historic CO2 emissions from Chapter 2.2 (EDGAR v6).
19 Pathways coinciding with emissions levels projected from the implementation of current NDCs
20 would result in substantially (>0.1°C) exceeding 1.5°C. They would have to reach net zero CO2
net zero CO2 by about ±10 years, assuming global CO2 emissions decrease linearly from current levels of
around 40 GtCO2 to net zero.
Do Not Cite, Quote or Distribute 3-39 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 around 5-10 years later5 than in pathways with no or limited overshoot in order to reach the net
2 negative emissions that would then be required to return warming to 1.5°C by 2100. Those high
3 overshoot pathways have higher transient warming and higher reliance on net-negative CO2 emissions
4 towards the end of the 21st century. As they need to reach net zero CO2 only a few years later, with
5 2030 CO2 emission levels being around twice as high, they imply post-2030 CO2 emissions reduction
6 rates that are almost double that of pathways limiting warming to 1.5°C with no or limited overshoot
7 (Section 3.5).
8 Pathways following emissions levels projected from the implementation of current NDCs until
9 2030 would have to reach net zero CO2 around 10 years earlier6 than cost-effective pathways to
10 likely limit warming to 2°C. While cost-effective pathways take around 50-55 years to reach net
11 zero CO2 emissions, those pathways would only have 35-40 years left for transitioning to net zero
12 CO2 from 2030 onwards, close to the transition times that 1.5°C pathways are faced with today.
13 Current CO2 emissions and 2030 emission levels projected under the current NDCs are in a similar
14 range. (3.5, 4.2)
15 Net zero GHG emissions
16 The amount of CO2-equivalent emissions and the point when net zero GHG emissions are
17 reached in multi-GHG emissions pathways depends on the choice of GHG emissions metric.
18 Various GHG emission metrics are available for this purpose7. GWP100 is the most commonly used
19 metric for reporting CO2-equivalent emissions and is required for emissions reporting under the
20 Rulebook of the Paris Agreement. (Cross-Chapter Box 2 in Chapter 2, Annex II section 9, Annex I)
21 For most choices of GHG emissions metric, reaching net zero GHG emissions requires net
22 negative CO2 emissions in order to balance residual CH4, N2O and F-gas emissions. Under
23 foreseen technology developments, some CH4, N2O and F-gas emissions from, e.g., agriculture and
24 industry will remain over the course of this century. Net negative CO2 emissions will therefore be
25 needed to balance these remaining non-CO2 GHG emissions to obtain net zero GHG emissions at a
26 point in time after net zero CO2 has been reached in emissions pathways. Both the amount of net
27 negative CO2 emissions and the time lag to reaching net zero GHG depend on the choice of GHG
28 emission metric.
29 Reaching net zero GHG emissions globally in terms of GWP100 leads to a reduction in global
30 warming from an earlier peak. This is due to net negative CO2 emissions balancing the GWP100-
31 equivalent emissions of short lived GHG emissions, which by themselves do not contribute to further
32 warming if sufficiently declining (Fuglestvedt et al. 2018; Rogelj et al. 2021). Hence, 1.5-2°C
FOOTNOTE 5 Pathways following emissions levels of current NDCs to 2030 and then returning median
warming to below 1.5°C in 2100 reach net zero during 2055-2060 (2045-2070) (median five year interval and
5th-95th percentile range).
FOOTNOTE 6 Pathways that follow emission levels projected from the implementation of current NDCs until
2030 and that still likely limit warming to 2°C reach net zero CO2 emissions during 2065 - 2070 (2060 - ...)
compared with 2075 - 2080 (2060 - …) in cost-effective pathways acting immediately to likely limit warming to
2°C (median five year interval and 5th-95th percentile range). See Footnote 1 for the lack of 95th percentile.
(Chapter 3.3 Table 3.1.)
FOOTNOTE 7 Defining net zero GHG emissions for a basket of greenhouse gases (GHGs) relies on a metric to
convert GHG emissions including methane (CH4), nitrous oxide (N2O), fluorinated gases (F-gases), and
potentially other gases, to CO2-equivalent emissions. The choice of metric ranges from Global Warming
Potentials (GWPs) and Global Temperature change Potentials (GTP) to economically-oriented metrics. All
metrics have advantages and disadvantages depending on the context in which they are used (Cross-Chapter
Box 2 in Chapter 2).
Do Not Cite, Quote or Distribute 3-40 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 emissions pathways in the AR6 scenario database that reach global net zero GHG emissions in the
2 second half of the century show warming being halted at some peak value followed by a gradual
3 decline towards the end of the century (WGI Chapter 1 Box 1.4).
4 Global net zero GHG emissions measured in terms of GWP100 are reached between 2095-2100
5 (2055 - ...)8 in emission pathways limiting warming to 1.5°C with no or limited overshoot
6 (median and 5th-95th percentile). Around 50% of pathways limiting warming to 1.5°C with no or
7 limited overshoot and 70% of pathways limiting likely warming to 2°C do not reach net zero GHG
8 emissions in terms of GWP100 before 2100. These pathways tend to show less reduction in warming
9 after the peak than pathways that reach net zero GHG emissions. For the subset of pathways that
10 reach net zero GHG emissions before 2100, including 90% of pathways that return warming to 1.5°C
11 by 2100 after a high (>0.1°C) overshoot, the time lag between reaching net zero CO2 and net zero
12 GHG is 11-14 (6-40) years and the amount of net negative CO2 emissions deployed to balance non-
13 CO2 emissions at the time of net zero is -6 to -7 (-10 to -4) GtCO2 (range of medians and lowest 5th
14 to highest 95 percentile across the four scenario classes that limit median warming to 2°C or lower).
15 (section 3.3, Table 3.1)
16 Sectoral and regional aspects of net zero
17 The timing of net zero CO2 or GHG emissions may differ across regions and sectors. Achieving
18 net zero emissions globally implies that some sectors and regions must reach net zero CO 2 or
19 GHG ahead of the time of global net zero CO2 or GHG if others reach it later. Similarly, some
20 sectors and regions would need to achieve net negative CO2 or GHG emissions to compensate for
21 continued emissions by other sectors and regions after the global net zero year. Differences in the
22 timing to reach net zero emissions between sectors and regions depend on multiple factors, including
23 the potential of countries and sectors to reduce GHG emissions and undertake carbon dioxide
24 removal, the associated costs, and the availability of policy mechanisms to balance emissions and
25 removals between sectors and countries (Fyson et al. 2020; Strefler et al. 2021a; van Soest et al.
26 2021b). A lack of such mechanisms could lead to higher global costs to reach net zero emissions
27 globally, but less interdependencies and institutional needs (Fajardy and Mac Dowell 2020). Sectors
28 will reach net zero CO2 and GHG emissions at different times if they are aiming for such targets with
29 sector-specific policies or as part of an economy-wide net zero emissions strategy integrating
30 emissions reductions and removals across sectors. In the latter case, sectors with large potential for
31 achieving net-negative emissions would go beyond net zero to balance residual emissions from
32 sectors with low potential which in turn would take more time compared to the case of sector-specific
33 action. Global pathways project global AFOLU emissions to reach global net zero CO2 the earliest,
34 around 2030-2035 in pathways likely to limit warming to 2°C and below, by rapid reduction of
35 deforestation and enhancing carbon sinks on land, although net zero GHG emissions from global
36 AFOLU are typically reached 30 years later, if at all. The ability of global AFOLU CO 2 emissions to
37 reach net zero as early as in the 2030s in modelled pathways hinges on optimistic assumptions about
38 the ability to establish global cost-effective mechanisms to balance emissions reductions and removals
39 across regions and sectors. These assumptions have been challenged in the literature and the Special
40 Report on Climate Change and Land (IPCC SRCCL).
41 The adoption and implementation of net zero CO2 or GHG emission targets by countries and
42 regions also depends on equity and capacity criteria. The Paris Agreement recognizes that peaking
43 of emissions will occur later in developing countries (Article 4.1). Just transitions to net zero CO 2 or
44 GHG could be expected to follow multiple pathways, in different contexts. Regions may decide about
45 net zero pathways based on their consideration of potential for rapid transition to low-carbon
46 development pathways, the capacity to design and implement those changes, and perceptions of
47 equity within and across countries. Cost-effective pathways from global models have been shown to
FOOTNOTE 8 The 95th percentile cannot be deduced from the scenario database as more than 5% of pathways
do not reach net zero GHG by 2100 (Chapter 3.3 Table 3.1.), hence denoted by -…
Do Not Cite, Quote or Distribute 3-41 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 distribute the mitigation effort unevenly and inequitably in the absence of financial support
2 mechanisms and capacity building (Budolfson et al. 2021), and hence would require additional
3 measures to become aligned with equity considerations (Fyson et al. 2020; van Soest et al. 2021b).
4 Formulation of net zero pathways by countries will benefit from clarity on scope, roadmaps and
5 fairness (Rogelj et al. 2021; Smith 2021). Achieving net zero emission targets relies on policies,
6 institutions and milestones against which to track progress. Milestones can include emissions levels,
7 as well as markers of technological diffusion.
8 The accounting of anthropogenic carbon dioxide removal on land matters for the evaluation of
9 net zero CO2 and net zero GHG strategies. Due to the use of different approaches between national
10 inventories and global models, the current net CO2 emissions are lower by 5.5 GtCO2, and cumulative
11 net CO2 emissions in modelled 1.5-2°C pathways would be lower by 104-170 GtCO2, if carbon
12 dioxide removals on land are accounted based on national GHG inventories. National GHG
13 inventories typically consider a much larger area of managed forest than global models, and on this
14 area additionally consider the fluxes due to human-induced global environmental change (indirect
15 effects) to be anthropogenic, while global models consider these fluxes to be natural. Both approaches
16 capture the same land fluxes, only the accounting of anthropogenic vs. natural emission is different.
17 Methods to convert estimates from global models to the accounting scheme of national GHG
18 inventories will improve the use of emission pathways from global models as benchmarks against
19 which collective progress is assessed. (7.2.2.5, Cross-Chapter Box 3 in this chapter)
20 Net zero CO2 and carbon neutrality have different meanings in this assessment, as is the case for
21 net zero GHG and GHG neutrality. They apply to different boundaries in the emissions and
22 removals being considered. Net zero (GHG or CO2) refers to emissions and removals under the
23 direct control or territorial responsibility of the reporting entity. In contrast, (GHG or carbon)
24 neutrality includes anthropogenic emissions and anthropogenic removals within and also those
25 beyond the direct control or territorial responsibility of the reporting entity. At the global scale, net
26 zero CO2 and carbon neutrality are equivalent, as is the case for net zero GHG and GHG neutrality.
27 The term “climate neutrality” is not used in this assessment because the concept of climate neutrality
28 is diffuse, used differently by different communities, and not readily quantified.
29 END BOX HERE
30
31
32 Figure 3.14: Net zero year for CO2 and all GHGs (based on AR6 GWP-100) as a function of remaining
33 carbon budget and temperature outcomes (not that stabilize (near) zero are also included in determining
34 the net zero year)
35 Table 3.2 summarizes the key characteristics for all temperature categories in terms of cumulative
36 CO2 emissions, near-term emission reductions, and the years of peak emission and net zero CO2 and
37 GHG emissions. The table shows again that many pathways in the literature likely limit global
38 warming to 2°C or limit warming to 1.5°C with limited overshoot compared to preindustrial levels.
Do Not Cite, Quote or Distribute 3-42 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 Cumulative net CO2 emissions from the year 2020 until the time of net zero CO2 in pathways that
2 limit warming to 1.5°C with no or limited overshoot are 510 (330-710) GtCO2 and in pathways likely
3 to limit warming to below 2.0°C 890 (640-1160) GtCO2 (see also Cross-Chapter Box 3 in this
4 chapter). Mitigation pathways likely to limit global warming to 2°C compared to pre-industrial levels
5 are associated with net global GHG emissions of 40 (32–55) GtCO2-eq yr-1 by 2030 and 20 (13-26)
6 GtCO2-eq yr-1 in 2050. These correspond to GHG emissions reductions of 21 (1-42) % by 2030, and
7 64 (53-77) % by 2050 relative to 2019 emission levels. Pathways that limit global warming to below
8 1.5°C with no or limited overshoot require a further acceleration in the pace of the transformation,
9 with GHG emissions reductions of 43 (34–60) % by 2030 and 84 (74–98) % in 2050 relative to
10 modelled 2019 emission levels. The likelihood of limiting warming to below 1.5°C with no or limited
11 overshoot of the most stringent mitigation pathways in the literature (C1) has declined since SR1.5.
12 This is because emissions have risen since 2010 by about 9 GtCO2 yr-1, resulting in relatively higher
13 near-term emissions of the AR6 pathways by 2030 and slightly later dates for reaching net zero CO 2
14 emissions compared to SR1.5
15
16 Given the larger contribution of scenarios in the literature that aim to reduce net-negative emissions,
17 emission reductions are somewhat larger in the short-term compared to similar categories in the IPCC
18 SR1.5. At the same time, the year of net zero emissions is somewhat later (but only if these rapid,
19 short-term emission reductions are achieved). The scenarios in the literature in C1-C3 show a peak in
20 global emissions before 2025. Not achieving this requires a more rapid reduction after 2025 to still
21 meet the Paris goals (see Section 3.5).
22
Do Not Cite, Quote or Distribute 3-43 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 Table 3.2 GHG, CO2 emissions and warming characteristics of different mitigation pathways submitted to the AR6 scenarios database and as categorized in the
2 climate assessment.
3
4
0 Values in the table refer to the 50th and (5th-95th) percentile values. For emissions-related columns this relates to the distribution of all the scenarios in that category. For Temperature Change and Likelihood
columns, single upper row values are the 50th percentile value across scenarios in that Category for the MAGICC climate model emulator.
For the bracketed ranges for temperatures and likelihoods, the median warming for every scenario in that category is calculated for each of the three climate model emulators (MAGICC, FaIR and CICERO-SCM).
Subsequently, the 5th and 95th percentile values across all scenarios is calculated. The coolest and warmest outcomes (i.e., the lowest p5 of three emulators, and the highest p95, respectively) are shown in the
brackets. Thus, these ranges cover the extent of scenario and climate model emulator uncertainty.
1 Category definitions consider at peak warming and warming at the end-of-century (2100).
C1: Below 1.5°C in 2100 with a greater than 50% probability and a peak warming higher than 1.5°C with less than 67% probability.
C2: Below 1.5°C in 2100 with a greater than 50% probability but peak warming higher than 1.5°C with greater than or equal to 67% probability.
C3: Likely below 2 °C throughout the century with greater than 67% probability.
C4, C5, C6, C7: Below 2 °C, 2.5 °C, 3 °C and 4 °C throughout the century, respectively, with greater than 50% probability.
C8: Peak warming above 4 °C with greater than or equal to 50% probability.
2 All warming levels are relative to the pre-industrial temperatures from the 1850-1900 period.
3 The warming profile of Neg peaks around 2060 and declines to below 1.5°C (50% likelihood) shortly after 2100. Whilst technically classified as a C3, it strongly exhibits the characteristics of C2 high overshoot
scenarios.
Do Not Cite, Quote or Distribute 3-44 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
4 C3 scenarios are sub-categorized according to policy ambition and consistent with Figure SPM 6. Hence the subtotals of C3a & C3b do not match the total of C3 scenarios, as there areC3 scenarios with policy
categorizations not covered by C3a and C3b.
5 Percentage GHG reduction ranges shown here compare 2019 estimates from historical emissions assessed in Chapter 2 (58 Gt CO2) to the harmonized and infilled projections from the models. Negative values (e.g.,
in C7, C8) represent an increase in emissions.
6 Gross, % reductions and emissions milestones are based on model data for CO2 & GHG emissions, which has been harmonized to 2015 values. See also Footnote 6.
7 Percentiles reported across all pathways in that category including pathways that do not reach net zero before 2100 (fraction in square brackets). If the fraction of pathways that reach net zero before 2100 (one
minus fraction in square brackets) is lower than the fraction of pathways covered by a percentile (e.g., 0.95 for the 95th percentile), the percentile is not defined and denoted with "…".
8 For cases where models do not report all GHGs, missing GHG species are infilled and calculated as Kyoto basket with AR6 GWP-100 CO2-equivalent factors. For each scenario, a minimum of native reporting of CO2,
CH4, and N2O emissions to 2100 was required for the assessment of the climate response and assignment to a climate category. Emissions scenarios without climate assessment are not included in the ranges. See
Annex III.
9 For better comparability with the WGI assessment of the remaining carbon budget, the cumulative CO2 emissions of the pathways are harmonized to the 2015 CO2 emissions levels used in the WGI assessment and
are calculated for the future starting on 1 January 2020.
10 Temperature change (Global Surface Air Temperature - GSAT) for category (at peak and in 2100), based on the median warming for each scenario assessed using the probabilistic climate model emulators.
11 Probability of staying below the temperature thresholds for the scenarios in each category, taking into consideration the range of uncertainty from the climate model emulators consistent with the WGI AR6
assessment. The probabilities refer to the probability at peak temperature. Note that in the case of temperature overshoot (E.g., category C2 and some scenarios in C1), the probabilities at the end of the century are
higher than the probability at peak temperature.
Do Not Cite, Quote or Distribute 3-45 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 3.3.2.4 Mitigation strategies
2 Detailed sectoral implications are discussed in Section 3.4 and Chapters 5-11 (see also Table 3.3). The
3 stringency of climate policy has clear implications for mitigation action (Figure 3.15). There are a
4 number of important commonalities of pathways likely limiting warming to 2C and below: for
5 instance, they all rely on significant improvement of energy efficiency, rapid decarbonisation of
6 supply and, many of them, CDR (in energy supply or AFOLU), either in terms of net negative
7 emissions or to compensate residual emissions. Still, there are also important differences and the
8 (IMPs) show how different choices can steer the system into alternative directions with different
9 combinations of response options. For decarbonisation of energy supply many options exist, including
10 CCS, nuclear power, and renewables (see Chapter 6). In the majority of the scenarios reaching low
11 greenhouse targets, a considerable amount of CCS is applied (panel d). The share of renewables is
12 around 30-70% in the scenarios reaching a global average temperature change of likely below 2oC and
13 clearly above 40% for scenarios reaching 1.5oC (panel c). Scenarios have been published with 100%
14 renewable energy systems even at a global scale, partly reflecting the rapid progress made for these
15 technologies in the last decade (Breyer and Jefferson 2020; Creutzig et al. 2017; Jacobson et al. 2018).
16 These scenarios do not show in the graph due to a lack of information from non-energy sources. There
17 is a debate in the literature on whether it is possible to achieve a 100% renewable energy system by
18 2050 (Brook et al. 2018). This critically depends on assumptions made on future system integration,
19 system flexibility, storage options, consequences for material demand and the ability to supply high-
20 temperature functions and specific mobility functions with renewable energy. The range of studies
21 published showing 100% renewable energy systems show that it is possible to design such systems in
22 the context of energy system models (Lehtveer and Hedenus 2015a,b; Hong et al. 2014a,b; Zappa et
23 al. 2019; Pfenninger and Keirstead 2015; Sepulveda et al. 2018; IEA 2021b) (see also Box 6.6 in
24 Chapter 6 on 100% renewables in net zero CO2 systems). Panel e and f, finally, show the contribution
25 of CDR – both in terms of net negative emissions and gross CDR. The contribution of total CDR
26 obviously exceeds the net negative emissions. It should be noted that while a majority of scenarios
27 relies on net negative emissions to reach stringent mitigation goals – this is not the case for all of
28 them.
Do Not Cite, Quote or Distribute 3-46 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1
2
3 Figure 3.15: Characteristics of scenarios as a function of the remaining carbon budget (mean
4 decarbonisation rate is shown as the average reduction in the period 2010-2050 divided by 2010
5 emissions). The categories C1-C7 are explained in Table 3.1
6 The spread shown in Figure 3.15 implies different mitigation strategies that could all lead to
7 emissions levels consistent with the Paris Agreement (and reach zero emissions). The IMPs illustrate
8 some options for different decarbonization pathways with heavy reliance on renewables (Ren), strong
9 emphasis on energy demand reductions (LD), wide-spread deployment of CDR methods coupled with
10 CCS (BECCS and DACCS) (Neg), mitigation in the context of sustainable development (SP) (Figure
11 3.16). For example, in some scenarios, a small part of the energy system is still based on fossil fuels in
12 2100 (Neg), while in others, fossil fuels are almost or completely phased out (Ren). Nevertheless, in
13 all scenarios, fossil fuel use is greatly reduced and unabated coal use is completely phased out by
14 2050. Also, nuclear power can be part of a mitigation strategy (however, the literature only includes
15 some scenarios with high-nuclear contributions, such as Berger et al. (2017)). This is explored further
16 in Section 3.5. The different strategies are also clearly apparent in the way they scenarios reach net
17 zero emissions. While GS and Neg rely significantly on BECCS and DACCS, their use is far more
18 restricted in the other IMPs. Consistently, in these IMPs also residual emissions are also significantly
19 lower.
Do Not Cite, Quote or Distribute 3-47 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1
2 Figure 3.16: Primary energy use and net emissions at net zero year for the different IMPS
3 Mitigation pathways also have a regional dimension. In 2010, about 40% of emissions originated
4 from the Developed Countries and Eastern Europe and west-central Asia regions. According to the
5 projections shown in
6
7 Figure 3.17, the share of the latter regions will further increase to about 70% by 2050. In the scenarios
8 in the literature, emissions are typically almost equally reduced across the regions.
9
Do Not Cite, Quote or Distribute 3-48 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1
2 Figure 3.179: Emissions by region (including 5-95th percentile range)
3
4 3.3.3 Climate impacts on mitigation potential
5 At the moment, climate change impact on mitigation potential is hardly considered in model-based
6 scenarios. While a detailed overview of climate impacts is provided in IPCC WGII and Section 3.6
7 discusses the economic consequences, here we concentrate on the implications for mitigation
8 potential. Climate change directly impacts the carbon budget via all kinds of feedbacks – which is
9 included in the ranges provided for the carbon budget (e.g., 300-900 GtCO2 for 17th-83rd percentile for
10 not exceeding 1.5 °C; see Chapter 5 IPCC, 2021). Climate change, however, alters the production and
11 consumption of energy (see also Chapter 6.5). An overview of the literature is provided by Yalew et
12 al. (2020). In terms of supply, impacts could influence the cooling capacity of thermal plants, the
13 potential and predictability of renewable energy, and energy infrastructure (Cronin et al. 2018a; van
14 Vliet et al. 2016; Turner et al. 2017; Lucena et al. 2018; Gernaat et al. 2021; Yalew et al. 2020).
15 Although the outcomes of these studies differ, they seem to suggest that although impacts might be
16 relatively small at the global scale, they could be substantial at the regional scale (increasing or
17 decreasing potential). Climate change can also impact energy demand, with rising temperatures
18 resulting in decreases in heating demand and increases in cooling demand (Isaac and van Vuuren
19 2009; Zhou et al. 2014; Labriet et al. 2015; McFarland et al. 2015; Auffhammer et al. 2017; Clarke et
20 al. 2018; van Ruijven et al. 2019; Yalew et al. 2020). As expected, the increase in cooling demand
21 dominates the impact in warm regions and decreases in heating demand in cold regions (Clarke et al.
22 2018; Zhou et al. 2014; Isaac and van Vuuren 2009). Globally, most studies show a net increase in
23 energy demand at the end of the century due to climate impacts (van Ruijven et al. 2019; Isaac and
24 van Vuuren 2009; Clarke et al. 2018); however, one study shows a net decrease (Labriet et al. 2015).
25 Only a few studies quantify the combined impacts of climate change on energy supply and energy
26 demand (Emodi et al. 2019; Steinberg et al. 2020; McFarland et al. 2015; Mima and Criqui 2015).
27 These studies show increases in electricity generation in the USA (McFarland et al. 2015; Steinberg et
28 al. 2020) and increases in CO2 emissions in Australia (Emodi et al. 2019) or the USA (McFarland et
29 al. 2015).
30
31 Climate change can impact the potential for AFOLU mitigation action by altering terrestrial carbon
32 uptake, crop yields and bioenergy potential (see also Chapter 7). Carbon sequestration in forests may
33 be positively or adversely affected by climate change and CO2 fertilization. On the one hand, elevated
34 CO2 levels and higher temperatures could enhance tree growth rates, carbon sequestration, and timber
35 and biomass production (Beach et al. 2015; Kim et al. 2017; Anderegg et al. 2020). On the other hand,
FOOTNOTE9 The countries and areas classification in this figure deviate from the standard classification
scheme adopted by WGIII as set out in Annex II, section 1.
Do Not Cite, Quote or Distribute 3-49 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 climate change could lead to greater frequency and intensity of disturbance events in forests, such as
2 fires, prolonged droughts, storms, pests and diseases (Kim et al. 2017; Anderegg et al. 2020). The
3 impact of climate change on crop yields could also indirectly impact the availability of land for
4 mitigation and AFOLU emissions (Meijl et al. 2018; Calvin et al. 2013; Kyle et al. 2014; Bajželj and
5 Richards 2014; Beach et al. 2015). The impact is, however, uncertain, as discussed in WGII Chapter
6 5. A few studies estimate the effect of climate impacts on AFOLU on mitigation, finding increases in
7 carbon prices or mitigation costs by 1-6% in most scenarios (Calvin et al. 2013; Kyle et al. 2014).
8
9 In summary, a limited number of studies quantify the impact of climate on emissions pathways. The
10 most important impact in energy systems might be through the impact on demand, although climate
11 change could also impact renewable mitigation potential -certainly at the local and regional scale.
12 Climate change might be more important for land-use related mitigation measures, including
13 afforestation, bioenergy and nature-based solutions. The net effect of changes in climate and CO2
14 fertilization are uncertain but could be substantial (see also Chapter 7).
15
16 3.4 Integrating sectoral analysis into systems transformations
17 This section describes the role of sectors in long-term emissions pathways (see Table 3.3). We discuss
18 both sectoral aspects of IAM pathways and some insights from sectoral studies. Sectoral studies
19 typically include more detail and additional mitigation options compared to IAMs. However, sectoral
20 studies miss potential feedbacks and cross-sectoral linkages that are captured by IAMs. Additionally,
21 since IAMs include all emissions sources, these models can be used to identify pathways to a
22 particular climate goals. In such pathways, emissions are balanced across sectors typically based on
23 relative marginal abatement costs; as a result, some sectors are sources and some are sinks at the time
24 of net zero CO2 emissions. For these reasons, the mitigation observed in each sector in an IAM may
25 differ from the potential in sectoral studies. Given the strengths and limitations of each type of model,
26 IAMs and sectoral models are complementary, providing different perspectives.
27
28 Table 3.3: Section 3.4 structure, definitions, and relevant chapters
Section Sector What is included Relevant
chapter(s)
3.4.1 Cross-sector Supply and demand, bioenergy, timing of net zero CO2, Chapter 5, 12
other interactions among sectors
3.4.2 Energy supply Energy resources, transformation (e.g., electricity Chapter 6
generation, refineries, etc.)
3.4.3 Buildings1 Residential and commercial buildings, other non- Chapter 9
specified2
3.4.4 Transportation1 Road, rail, aviation, and shipping Chapter 10
3.4.5 Industry1 Industrial energy use and industrial processes Chapter 11
3.4.6 AFOLU Agriculture, forestry, and other land use Chapter 7
3.4.7 Other CDR CDR options not included in individual sectors (e.g., Chapter 12
direct air carbon capture and sequestration, enhanced
weathering)
1
29 Direct energy use and direct emissions only; emissions do not include those associated with energy
30 production
2
31 Other non-specified fuel use, including military. Some models report this category in the buildings
32 sector, while others report it in the “Other” sector
33
Do Not Cite, Quote or Distribute 3-50 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 3.4.1 Cross-sector linkages
2 3.4.1.1 Demand and supply strategies
3 Most IAM pathways rely heavily on supply-side mitigation strategies, including fuel switching,
4 decarbonization of fuels, and CDR (Creutzig et al. 2016; Bertram et al. 2018; Rogelj et al. 2018b;
5 Mundaca et al. 2019). For demand-side mitigation, IAMs incorporate changes in energy efficiency,
6 but many other demand-side options (e.g., behaviour and lifestyle changes) are often excluded from
7 models (van Sluisveld et al. 2015; Creutzig et al. 2016; Wilson et al. 2019; van den Berg et al. 2019).
8 In addition, this mitigation is typically price-driven and limited in magnitude (Yeh et al. 2017;
9 Sharmina et al. 2020; Luderer et al. 2018; Wachsmuth and Duscha 2019). In contrast, bottom-up
10 modelling studies show considerable potential for demand-side mitigation (Yeh et al. 2017;
11 Wachsmuth and Duscha 2019; Creutzig et al. 2016; Mundaca et al. 2019) (see also Chapter 5), which
12 can slow emissions growth and/or reduce emissions (Samadi et al. 2017; Creutzig et al. 2016).
13
14 A small number of mitigation pathways include stringent demand-side mitigation, including changes
15 in thermostat set points (van Vuuren et al. 2018; van Sluisveld et al. 2016), more efficient or smarter
16 appliances (Grubler et al. 2018; Napp et al. 2019; van Sluisveld et al. 2016), increased recycling or
17 reduced industrial goods (van de Ven et al. 2018; Liu et al. 2018; van Sluisveld et al. 2016; Grubler et
18 al. 2018; Napp et al. 2019), telework and travel avoidance (van de Ven et al. 2018; Grubler et al.
19 2018), shifts to public transit (van Vuuren et al. 2018; van Sluisveld et al. 2016; Grubler et al. 2018),
20 reductions in food waste (van de Ven et al. 2018) and less meat intensive diets (van de Ven et al.
21 2018; van Vuuren et al. 2018; Liu et al. 2018). These pathways show reduced dependence on CDR
22 and reduced pressure on land (van de Ven et al. 2018; van Vuuren et al. 2018; Grubler et al. 2018;
23 Rogelj et al. 2018a) (Chapter 5.3.3). However, the representation of these demand-side mitigation
24 options in IAMs is limited, with most models excluding the costs of such changes (van Sluisveld et al.
25 2016), using stylised assumptions to represent them (van den Berg et al. 2019), and excluding
26 rebound effects (Brockway et al. 2021; Krey et al. 2019). Furthermore, there are questions about the
27 achievability of such pathways, including whether the behavioural changes included are feasible
28 (Azevedo et al. 2021) and the extent to which development and demand can be decoupled (Semieniuk
29 et al. 2021; Steckel et al. 2013; Keyßer and Lenzen 2021; Brockway et al. 2021).
30
31 Figure 3.18: Indicators of demand and supply-side mitigation in the Illustrative Pathways (lines) and
32 the 5-95% range of Reference, 1.5C and 2C scenarios (shaded areas). shows indicators of supply-
33 and demand-side mitigation in the IMPs, as well as the range across the database. Two of these IMPs
34 (SP, LD) show strong reductions in energy demand, resulting in less reliance on bioenergy and
35 limited CDR from energy supply. In contrast, Neg has higher energy demand, depending more on
36 bioenergy and net negative CO2 emissions from energy supply.
37
Do Not Cite, Quote or Distribute 3-51 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1
2 Figure 3.18: Indicators of demand and supply-side mitigation in the Illustrative Pathways (lines) and the
3 5-95% range of Reference, 1.5C and 2C scenarios (shaded areas).
4 3.4.1.2 Sectoral emissions strategies and the timing of net zero
5 Mitigation pathways show differences in the timing of decarbonization (Figure 3.20) and the timing of
6 net zero (Figure 3.19) across sectors and regions (high confidence); the timing in a given sector
7 depends on the cost of abatement in it, the availability of CDR options, the scenario design, near-term
8 emissions levels, and the amount of non-CO2 abatement (Yeh et al. 2017; Emmerling et al. 2019;
9 Rogelj et al. 2019a,b; Johansson et al. 2020; van Soest et al. 2021b; Ou et al. 2021; Azevedo et al.
10 2021) (Cross-Chapter Box 3 in this chapter). However, delaying emissions reductions, or more
11 limited emissions reductions in one sector or region, involves compensating reductions in other
12 sectors or regions if warming is to be limited (high confidence) (Rochedo et al. 2018; Price and Keppo
13 2017; Grubler et al. 2018; van Soest et al. 2021b).
Do Not Cite, Quote or Distribute 3-52 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 Table 3.4: Energy, emissions and CDR characteristics of the pathways by climate category for 2030, 2050, 2100. Source: AR6 scenarios database
2
Footnotes
0 Values in the table refer to the 50th and (5th-95th) percentile values.
1 Category definitions consider at peak warming and warming at the end-of-century (2100).
C1: Below 1.5°C in 2100 with a greater than 50% probability and a peak warming higher than 1.5°C with less than 67% probability.
C2: Below 1.5°C in 2100 with a greater than 50% probability but peak warming higher than 1.5°C with greater than or equal to 67% probability.
C3: Likely below 2 °C throughout the century with greater than 67% probability.
C4, C5, C6, C7: Below 2 °C, 2.5 °C, 3 °C and 4 °C throughout the century, respectively, with greater than 50% probability.
C8: Peak warming above 4 °C with greater than or equal to 50% probability.
2 The warming profile of Neg peaks around 2060 and declines thereafter to below 1.5°C (50% likelihood) shortly after 2100. Whilst technically classified as a C3, it
strongly exhibits the characteristics of C2 high overshoot scenarios.
3 Primary Energy as calculated in 'Direct Equivalent' terms according to IPCC reporting conventions.
4 Low carbon energy here defined to include: renewables (including biomass, solar, wind, hydro, geothermal, ocean); fossil fuels when used with CCS; and, nuclear power.
Do Not Cite, Quote or Distribute 3-53 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1
2 Figure 3.19: Decade in which sectoral CO2 emissions first reach net negative values. Each panel is a
3 different temperature level. The colours indicate the decade in which CO 2 emissions go negative; the y-
4 axis indicates the share of scenarios achieving net zero in that decade. Only scenarios that pass the vetting
5 criteria are included (see Section 3.2). Scenarios achieving net zero prior to 2020 are excluded.
6 At the time of net zero global CO2 emissions, emissions in some sectors are positive and some
7 negative. In cost-effective mitigation pathways, the energy supply sector typically reaches net zero
8 CO2 before the economy as a whole, while the demand sectors reach net zero CO2 later, if at all
9 (Pietzcker et al. 2014; Price and Keppo 2017; Luderer et al. 2018; Rogelj et al. 2018a,b; Méjean et al.
10 2019; Azevedo et al. 2021) (Chapter 6.7). CO2 emissions from transport, industry, and buildings are
11 positive, and non-CO2 GHG emissions are also positive at the time of global net zero CO2 emissions
12 (Figure 3.20).
13
14
15 Figure 3.20: Greenhouse gas emissions, including CO2 emissions by sector and total non-CO2 GHGs in
16 2050 (top left), 2100 (top middle), year of global net zero CO2 (top right), cumulative CO2 emissions from
Do Not Cite, Quote or Distribute 3-54 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 2020-2100 (bottom left), and cumulative CO2 emissions from 2020 until the year of net zero CO 2 for
2 scenarios that limit warming to below 2C. Scenarios are grouped by their temperature category.
3 “Industry” includes CO2 emissions associated with industrial energy use only; sectors shown in this figure
4 do not necessarily sum to total CO2. In this, and other figures in Section 3.4, unless stated otherwise, only
5 scenarios that pass the vetting criteria are included (see Section 3.2). Boxes indicate the interquartile
6 range, the median is shown with a horizontal black line, while vertical lines show the 5 to 95% interval.
7 So, while pathways indicate some flexibility in emissions reductions across sectors, all pathways
8 involve substantial CO2 emissions reductions in all sectors and regions (high confidence) (Rogelj et al.
9 2018a,b; Luderer et al. 2018; Méjean et al. 2019; Azevedo et al. 2021). Projected CO2 emissions
10 reductions between 2019 and 2050 in 1.5C pathways with no or limited overshoot are around 77%
11 for energy demand, with a 5-95% range of 31 to 96%,10 115% for energy supply (90 to 167%), and
12 148% for AFOLU (94 to 387%). In likely 2C pathways, projected CO2 emissions are reduced
13 between 2019 and 2050 by around 49% for energy demand, 97% for energy supply, and 136% for
14 AFOLU (see also 3.4.2-3.4.6). Almost 75% of GHG reductions at the time of net zero GHG are from
15 the energy system, 13% are from AFOLU CO2, and 13% from non-CO2 (Figure 3.21). These
16 reductions are achieved through a variety of sectoral strategies, illustrated in Figure 3.21 (right panel),
17 and described in Sections 3.4.2 to 3.4.7; the primary strategies include declines in fossil energy,
18 increases in low carbon energy use, and CDR to address residual emissions.
19
20
21 Figure 3.21: Left panel: Greenhouse gas emissions reductions from 2019 by sector at the year of net zero
22 GHG for all scenarios that reach net zero GHG. Emissions reductions by sector for direct (demand) and
23 indirect (upstream supply) are shown as the percent of total GHG reductions.
24 Right panel: key indicators in 2050 for the IMPs. Definitions of significant and very significant are
25 defined relative to 2019 and vary between indicators, as follows: fossil energy (significant >10%, very
26 significant >50%), renewables (>150 EJ yr-1, >200 EJ yr-1), bioenergy (>100%, >200%), BECCS (>2.0
27 GtCO2 yr-1, >3.5 GtCO2 yr-1), AFOLU (>100% decline, >130%), energy crops (>150 million ha, >400),
28 forest (>5% increase, >15%).
29
30 In the context of mitigation pathways, only a few studies have examined solar radiation modification
31 (SRM), typically focusing on Stratospheric Aerosol Injection (Emmerling and Tavoni 2018a,b;
32 Arinoa et al. 2016; Belaia et al. 2021; Rickels et al. 2020; Heutel et al. 2018; Helwegen et al. 2019).
33 These studies find that substantial mitigation is required to limit warming to a given level, even if
FOOTNOTE 10 Unless otherwise specified, the values in parentheses in Section 3.4 from this point forward
indicate the 5-95% range.
Do Not Cite, Quote or Distribute 3-55 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 SRM is available (Moreno-Cruz and Smulders 2017; Emmerling and Tavoni 2018b; Belaia et al.
2 2021). SRM may reduce some climate impacts, reduce peak temperatures, lower mitigation costs, and
3 extend the time available to achieve mitigation; however, SRM does not address ocean acidification
4 and may involve risks to crop yields, economies, human health, or ecosystems (WGII Chapter 16;
5 WGI TS; WGI Ch 5; SR1.5 SPM; Cross-Working Group Box 4 in Chapter 14). There also are
6 significant uncertainties surrounding SRM, including uncertainties on the costs and risks, which can
7 substantially alter the amount of SRM used in modelled pathways (Tavoni et al. 2017; NASEM 2021;
8 Heutel et al. 2018; Helwegen et al. 2019; IPCC 2018). Furthermore, the degree of international
9 cooperation can influence the amount of SRM deployed in scenarios, with uncoordinated action
10 resulting in larger SRM deployment and consequently larger risks/impacts from SRM (Emmerling
11 and Tavoni 2018a). Bridging research and governance involves consideration of the full range of
12 societal choices and ramifications (Sugiyama et al. 2018). More information on SRM, including the
13 caveats, risks, uncertainties, and governance issues is found in WGI Chapter 4, WGIII Chapter 14,
14 and Cross-Working Group Box 4 in Chapter 14).
15
16 3.4.1.3 Linkages among sectors
17 Mitigation in one sector can be dependent upon mitigation in another sector, or may involve trade-offs
18 between sectors. Mitigation in energy demand often includes electrification (Luderer et al. 2018;
19 Pietzcker et al. 2014; Sharmina et al. 2020; DeAngelo et al. 2021), however such pathways only result
20 in reduced emissions if the electricity sector is decarbonized (Zhang and Fujimori 2020) (see also
21 Chapter 12). Relatedly, the mitigation potential of some sectors (e.g., transportation) depends on the
22 decarbonization of liquid fuels, e.g., through biofuels (Wise et al. 2017; Pietzcker et al. 2014;
23 Sharmina et al. 2020); Chapter 12). In other cases, mitigation in one sector results in reduced
24 emissions in another sector. For example, increased recycling can reduce primary resource extraction;
25 planting trees or green roofs in urban areas can reduce the energy demand associated with space
26 cooling (Chapter 12).
27
28 Mitigation in one sector can also result in additional emissions in another. One example is
29 electrification of end use which can result in increased emissions from energy supply. However, one
30 comparitively well-researched example of this linkage is bioenergy. An increase in demand for
31 bioenergy within the energy system has the potential to influence emissions in the AFOLU sector
32 through the intensification of land and forest management and/or via land use change (Smith et al.
33 2019; Daioglou et al. 2019; Smith et al. 2020a; IPCC 2019a). The effect of bioenergy and BECCS on
34 mitigation depends on a variety of factors in modelled pathways. In the energy system, the emissions
35 mitigation depends on the scale of deployment, the conversion technology, and the fuel displaced
36 (Calvin et al. 2021). Limiting or excluding bioenergy and/or BECCS increases mitigation cost and
37 may limit the ability of a model to reach a low warming level (Edmonds et al. 2013; Calvin et al.
38 2014b; Muratori et al. 2020; Luderer et al. 2018). In AFOLU, bioenergy can increase or decrease
39 terrestrial carbon stocks and carbon sequestration, depending on the scale, biomass feedstock, land
40 management practices, and prior land use (Calvin et al. 2014c; Wise et al. 2015; Smith et al. 2019,
41 2020a; IPCC 2019a; Calvin et al. 2021).
42
43 Pathways with very high biomass production for energy use typically include very high carbon prices
44 in the energy system (Popp et al. 2017; Rogelj et al. 2018b), little or no land policy (Calvin et al.
45 2014b), a high discount rate (Emmerling et al. 2019), and limited non-BECCS CDR options (e.g.,
46 afforestation, DACCS) (Fuhrman et al. 2020; Realmonte et al. 2019; Chen and Tavoni 2013;
47 Marcucci et al. 2017; Calvin et al. 2014b). Higher levels of bioenergy consumption are likely to
48 involve trade-offs with mitigation in other sectors, notably in construction (i.e., wood for material and
49 structural products) and AFOLU (carbon stocks and future carbon sequestration), as well as trade-offs
50 with sustainability (Section 3.7) and feasibility concerns (Section 3.8). Not all of these trade-offs are
Do Not Cite, Quote or Distribute 3-56 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 fully represented in all IAMs. Based on sectoral studies, the technical potential for bioenergy, when
2 constraints for food security and environmental considerations are included, are 5-50 and 50-250 EJ
3 yr-1 in 2050 for residues and dedicated biomass production systems, respectively (Chapter 7).
4 Bioenergy deployment in IAMs is within the range of these potentials, with between 75 and 248 EJ
5 yr-1 in 2050 in pathways that limit warming to 1.5C with no or limited overshoot. Finally, IAMs do
6 not include all potential feedstock and management practices, and have limited representation of
7 institutions, governance, and local context (Butnar et al. 2020; Brown et al. 2019; Calvin et al. 2021).
8 The inclusion of CDR options, like BECCS, can affect the timing of emissions mitigation in IAM
9 scenarios, i.e., delays in mitigations actions are compensated by net negative emissions in the second
10 half of the century. However, studies with limited net negative emissions in the long-term require very
11 rapid declines in emissions in the near-term (van Vuuren et al. 2017). Especially in forest-based
12 systems, increased harvesting of forests can perturb the carbon balance of forestry systems, increasing
13 emissions for some period; the duration of this period of increased emissions, preceding net emissions
14 reductions, can be very variable (Mitchell et al. 2012; Lamers and Junginger 2013; Röder et al. 2019;
15 Hanssen et al. 2020; Cowie et al. 2021). However, the factors contributing to differences in recovery
16 time are known (Zanchi et al. 2012; Laganière et al. 2017; Mitchell et al. 2012; Lamers and Junginger
17 2013; Röder et al. 2019). Some studies that consider market-mediated effects find that an increased
18 demand for biomass from forests can provide incentives to maintain existing forests and potentially to
19 expand forest areas, providing addition carbon sequestration as well as additional biomass (Kim et al.
20 2018; Dwivedi et al. 2014; Baker et al. 2019; Favero et al. 2020). However, these responses are
21 uncertain and likely to vary geographically.
22
23 3.4.2 Energy supply
24 Without mitigation, energy consumption and supply emissions continue to rise (high confidence)
25 (Riahi et al. 2017; Bauer et al. 2017; Kriegler et al. 2016; Mcjeon et al. 2021) (see also Chapter 6.7).
26 While the share of renewable energy continues to grow in reference scenarios, fossil fuel accounts for
27 the largest share of primary energy (Riahi et al. 2017; Bauer et al. 2017; Price and Keppo 2017).
28 In scenarios likely to limit warming to 2℃ and below, transition of the energy supply sector to a low
29 or no carbon system is rapid (Rogelj et al. 2016, 2018b; Luderer et al. 2018; Grubler et al. 2018; van
30 Vuuren et al. 2018). CO2 emissions from energy supply reach net zero around 2041 (2033 to 2057) in
31 pathways limiting warming to below 1.5℃ with no or limited overshoot and around 2053 (2040 to
32 2066) in pathways likely to limit warming to 2℃. Emissions reductions continue, with emissions
33 reaching -7.1 GtCO2 per year (-15 to -2.3 GtCO2 per year) in 2100 in all pathways likely to limit
34 warming to 2℃ and below.
35 All pathways likely to limit warming to 2C and below show substantial reductions in fossil fuel
36 consumption and a near elimination of the use of coal without CCS (high confidence) (Welsby et al.
37 2021; Bauer et al. 2017; van Vuuren et al. 2018; Rogelj et al. 2018a,b; Grubler et al. 2018; Luderer et
38 al. 2018; Azevedo et al. 2021; Mcjeon et al. 2021) (Figure 3.22). In these pathways, the use of coal,
39 gas and oil is reduced by 90, 25, and 41%, respectively, between 2019 and 2050 and 91, 39, and 78%
40 between 2019 and 2100; coal without CCS is further reduced to 99% below its 2019 levels in 2100.
41 These pathways show an increase in low carbon energy, with 88% (69-97%) of primary energy from
42 low carbon sources in 2100, with different combinations of low carbon fuels (e.g., non-biomass
43 renewables, biomass, nuclear, and CCS) (Rogelj et al. 2018a,b; van Vuuren et al. 2018) (Chapter 6.7,
44 Section 3.4.1). Across all pathways likely to limit warming to 2℃ and below, non-biomass
45 renewables account for 52% (24 to 77%) of primary energy in 2100 (Pietzcker et al. 2017; Creutzig et
46 al. 2017; Rogelj et al. 2018b); Figure 3.22, Chapter 6). There are some studies analysing the potential
47 for 100% renewable energy systems (Hansen et al. 2019); however, there are a range of issues around
48 such systems (see Chapter 6.6, Box 6.6).
Do Not Cite, Quote or Distribute 3-57 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1
2 Figure 3.22 Primary energy consumption across scenarios: total primary energy (top left), fossil fuels (top
3 middle), coal without CCS (top right), non-biomass renewables (bottom left), and biomass (bottom
4 middle). Scenarios are grouped by their temperature category. Primary energy is reported in direct
5 equivalent, where one unit of nuclear or non-biomass renewable energy output is reported as one unit of
6 primary energy. Not all subcategories of primary energy are shown.
7 Stringent emissions reductions at the level required to limit warming to 2C or 1.5C are achieved
8 through increased electrification of end-use, resulting in increased electricity generation in all
9 pathways (high confidence) (Figure 3.23) (Rogelj et al. 2018a; Azevedo et al. 2021). Nearly all
10 electricity in pathways likely to limit warming to 2℃ and below is from low or no carbon fuels
11 (Rogelj et al. 2018a; Azevedo et al. 2021), with different shares of nuclear, biomass, non-biomass
12 renewables, and fossil CCS across pathways. Low emissions scenarios also show increases in
13 hydrogen use (Figure 3.23).
Do Not Cite, Quote or Distribute 3-58 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1
2 Figure 3.23: Electricity (top left), share of low carbon electricity (top right), and hydrogen (bottom left)
3 production across all scenarios, grouped by the categories introduced in section 3.2. Low carbon includes
4 non-biomass renewables, biomass, nuclear, and CCS.
5 3.4.3 Buildings
6 Global final energy use in the building sector increases in all pathways as a result of population
7 growth and increasing affluence (Figure 3.24). There is very little difference in final energy intensity
8 for the buildings sector across scenarios. Direct CO2 emissions from the buildings sector vary more
9 widely across temperature stabilization levels than energy consumption. In 2100, scenarios above 3ºC
10 [C7-C8] still show an increase of CO2 emissions from buildings around 29% above 2019, while all
11 scenarios likely to limit warming to 2ºC and below have emission reductions of around 85% (8-
12 100%). Carbon intensity declines in all scenarios, but much more sharply as the warming level is
13 reduced.
Do Not Cite, Quote or Distribute 3-59 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1
2 Figure 3.24: Buildings final energy (top left), CO2 emissions (top middle), carbon intensity (top right),
3 energy intensity (bottom left), share of final energy from electricity (bottom middle), and share of final
4 energy from gases (bottom right). Energy intensity is final energy per unit of GDP. Carbon intensity is
5 CO2 emissions per EJ of final energy. The first four indicators are indexed to 2019 11, where values less
6 than 1 indicate a reduction.
7 In all scenarios, the share of electricity in final energy use increases, a trend that is accelerated by
8 2050 for the scenarios likely to limit warming to 2ºC and below (Figure 3.23). By 2100, the low
9 warming scenarios show large shares of electricity in final energy consumption for buildings. The
10 opposite is observed for gases.
11
12 While several global IAM models have developed their buildings modules considerably over the past
13 decade (Daioglou et al. 2012; Knobloch et al. 2017; Clarke et al. 2018; Edelenbosch et al. 2021;
14 Mastrucci et al. 2021), the extremely limited availability of key sectoral variables in the AR6
15 scenarios database (such as floor space and energy use for individual services) prohibit a detailed
16 analysis of sectoral dynamics. Individual studies in the literature often focus on single aspects of the
17 buildings sector, though collectively providing a more comprehensive overview (Edelenbosch et al.
18 2020; Ürge-Vorsatz et al. 2020). For example, energy demand is driven by economic development
19 that fulfills basic needs (Mastrucci et al. 2019; Rao et al. 2019a), but also drives up floorspace in
20 general (Daioglou et al. 2012; Levesque et al. 2018; Mastrucci et al. 2021) and ownership of energy
21 intensive appliances such as air conditioners (Isaac and van Vuuren 2009; Colelli and Cian 2020;
22 Poblete-Cazenave et al. 2021). These dynamics are heterogeneous and lead to differences in energy
23 demand and emission mitigation potential across urban/rural buildings and income levels (Krey et al.
24 2012; Poblete-Cazenave et al. 2021). Mitigation scenarios rely on fuel switching and technology
25 (Dagnachew et al. 2020; Knobloch et al. 2017), efficiency improvement in building envelopes
26 (Edelenbosch et al. 2021; Levesque et al. 2018) and behavioural changes (Niamir et al. 2018, 2020;
FOOTNOTE 11 2019 values are from model results and interpolated from other years when not directly
reported.
Do Not Cite, Quote or Distribute 3-60 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 van Sluisveld et al. 2016). The in-depth dynamics of mitigation in the building sector are explored in
2 Chapter 9.
3 3.4.4 Transport
4
5 Figure 3.25: Transport final energy (top left), CO 2 emissions (top middle), carbon intensity (top right),
6 and share of final energy from electricity (bottom left), hydrogen (bottom middle), and biofuels (bottom
7 right). See Chapter 10 for a discussion of energy intensity. Carbon intensity is CO 2 emissions per EJ of
8 final energy. The first three indicators are indexed to 2019 12, where values less than 1 indicate a
9 reduction.
10 Reference scenarios show growth in transport demand, particularly in aviation and freight (Yeh et al.
11 2017; Sharmina et al. 2020; Müller-Casseres et al. 2021b). Energy consumption continues to be
12 dominated by fossil fuels in reference scenarios, with some increases in electrification (Edelenbosch
13 et al. 2020; Yeh et al. 2017). CO2 emissions from transport increase for most models in reference
14 scenarios (Yeh et al. 2017; Edelenbosch et al. 2020).
15 The relative contribution of demand-side reduction, energy efficiency improvements, fuel switching,
16 and decarbonisation of fuels, varyies by model, level of mitigation, mitigation options available, and
17 underlying socio-economic pathway (Longden 2014; Wise et al. 2017; Luderer et al. 2018; Yeh et al.
18 2017; Edelenbosch et al. 2020; Müller-Casseres et al. 2021a,b). IAMs typically rely on technology-
19 focused measures like energy efficiency improvements and fuel switching to reduce carbon emissions
20 (Pietzcker et al. 2014; Edelenbosch et al. 2017a; Yeh et al. 2017; Zhang et al. 2018a,b; Rogelj et al.
21 2018b; Sharmina et al. 2020). Many mitigation pathways show electrification of the transport system
22 (Luderer et al. 2018; Pietzcker et al. 2014; Longden 2014; Zhang et al. 2018a); however, without
23 decarbonization of the electricity system, transport electrification can increase total energy system
24 emissions (Zhang and Fujimori 2020). A small number of pathways include demand-side mitigation
25 measures in the transport sector; these studies show reduced carbon prices and reduced dependence on
FOOTNOTE 12 2019 values are from model results and interpolated from other years when not directly
reported.
Do Not Cite, Quote or Distribute 3-61 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 CDR (Grubler et al. 2018; Méjean et al. 2019; van de Ven et al. 2018; Zhang et al. 2018c) (Section
2 3.4.1).
3 Across all IAM scenarios assessed, final energy demand for transport continues to grow, including in
4 many stringent mitigation pathways (Figure 3.25). The carbon intensity of energy declines
5 substantially by 2100 in likely 2°C and below scenarios, leading to substantial declines in transport
6 sector CO2 emissions with increased electrification of the transport system (Figure 3.23).
7 The transport sector has more detail than other sectors in many IAMs (Edelenbosch et al. 2020);
8 however, there is considerable variation across models. Some models (e.g., GCAM, IMAGE,
9 MESSAGE-GLOBIOM) represent different transport modes with endogenous shifts across modes as
10 a function of income, price, and modal speed (Edelenbosch et al. 2020).13 However, IAMs, including
11 those with detailed transport, exclude several supply-side (e.g., synthetic fuels) and demand-side (e.g.,
12 behaviour change, reduced shipping, telework and automation) mitigation options (Davis et al. 2018;
13 Köhler et al. 2020; Mittal et al. 2017; Gota et al. 2019; Wilson et al. 2019; Creutzig et al. 2016;
14 Sharmina et al. 2020; Pietzcker et al. 2014; Lefèvre et al. 2021; Müller-Casseres et al. 2021a,b).
15
16 As a result of these missing options and differences in how mitigation is implemented, IAMs tend to
17 show less mitigation than the potential from national transport/energy models (Wachsmuth and
18 Duscha 2019; Gota et al. 2019; Yeh et al. 2017; Edelenbosch et al. 2020). For the transport sector as a
19 whole, studies suggest a mitigation potential of 4-5 GtCO2 per year in 2030 (Edelenbosch et al. 2020)
20 with complete decarbonization possible by 2050 (Gota et al. 2019; Wachsmuth and Duscha 2019).
21 However, in the scenarios assessed in this chapter that limit warming to below 1.5C with no or
22 limited overshoot, transport sector CO2 emissions are reduced by only 59% (28% to 81%) in 2050
23 compared to 2015. IAM pathways also show less electrification than the potential from other studies;
24 pathways that limit warming to below 1.5C with no or limited overshoot show a median of 25% (7 to
25 43%) of final energy from electricity in 2050, while the IEA NZE scenario includes 45% (IEA
26 2021a).
27
FOOTNOTE 13 Some of these models are treated as global transport energy sectoral models (GTEMs) in
Chapter 10.
Do Not Cite, Quote or Distribute 3-62 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 3.4.5 Industry
2
3 Figure 3.26: Industrial final energy, including feedstocks (top left), CO 2 emissions (top middle), carbon
4 intensity (top right), energy intensity (bottom left), share of final energy from electricity (bottom middle),
5 and share of final energy from gases (bottom right). Energy intensity is final energy per unit of GDP.
6 Carbon intensity is CO2 emissions per EJ of final energy. The first four indicators are indexed to 2019,14
7 where values less than 1 indicate a reduction. Industrial sector CO 2 emissions include fuel combustion
8 emissions only.
9 Reference scenarios show declines in energy intensity, but increases in final energy use in the
10 industrial sector (Edelenbosch et al. 2017b). These scenarios show increases in CO2 emissions both
11 for the total industrial sector (Edelenbosch et al. 2020; Luderer et al. 2018; Edelenbosch et al. 2017b)
12 and individual subsectors like cement and iron and steel (van Ruijven et al. 2016; van Sluisveld et al.
13 2021) or chemicals (Daioglou et al. 2014; van Sluisveld et al. 2021).
14 In mitigation pathways, CO2 emissions reductions are achieved through a combination of energy
15 savings (via energy efficiency improvements and energy conservation), structural change, fuel
16 switching, and decarbonization of fuels (Grubler et al. 2018; Luderer et al. 2018; Edelenbosch et al.
17 2017b, 2020). Mitigation pathways show reductions in final energy for industry compared to the
18 baseline (Edelenbosch et al. 2017b; Luderer et al. 2018; Edelenbosch et al. 2020) and reductions in
19 the carbon intensity of the industrial sector through both fuel switching and the use of CCS (Paltsev et
20 al. 2021; Luderer et al. 2018; Edelenbosch et al. 2017b, 2020; van Ruijven et al. 2016; van Sluisveld
21 et al. 2021). The mitigation potential differs depending on the industrial subsector and the availability
22 of CCS, with larger potential reductions in the steel sector (van Ruijven et al. 2016) and cement
23 industry (Sanjuán et al. 2020) than in the chemicals sector (Daioglou et al. 2014). Many scenarios,
24 including stringent mitigation scenarios, show continued growth in final energy; however, the carbon
25 intensity of energy declines in all mitigation scenarios (Figure 3.26).
26 The representation of the industry sector is very aggregate in most IAMs, with only a small subset of
27 models disaggregating key sectors such as cement, fertilizer, chemicals, and iron and steel (Pauliuk et
FOOTNOTE 14 2019 values are from model results and interpolated from other years when not directly
reported.
Do Not Cite, Quote or Distribute 3-63 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 al. 2017; Edelenbosch et al. 2017b; Daioglou et al. 2014; van Sluisveld et al. 2021) (Napp et al. 2019).
2 IAMs often account for both energy combustion and feedstocks (Edelenbosch et al. 2017b), but IAMs
3 typically ignore material flows and miss linkages between sectors (Kermeli et al. 2019; Pauliuk et al.
4 2017). By excluding these processes, IAMs misrepresent the mitigation potential of the industry
5 sector, e.g. by overlooking mitigation from material efficiency and circular economies (Sharmina et
6 al. 2020), which can have substantial mitigation potential (Chapter 5.3.4, Chapter 11.3).
7 Sectoral studies indicate a large mitigation potential in the industrial sector by 2050, including the
8 potential for net zero CO2 emissions for steel, plastics, ammonia, and cement (Section 11.4.1).
9 Detailed industry sector pathways show emissions reductions between 39 and 94% by mid-century
10 compared to present day15 (Section 11.4.2) and a substantial increase in direct electrification (IEA
11 2021a). IAMs show comparable mitigation potential to sectoral studies with median reductions in
12 CO2 emissions between 2019 and 2050 of 70% in scenarios likely to limit warming to 2C and below
13 and a maximum reduction of 96% (Figure 3.26). Some differences between IAMs and sectoral models
14 can be attributed to differences in technology availability, with IAMs sometimes including more
15 technologies (van Ruijven et al. 2016) and sometimes less (Sharmina et al. 2020). Figure 3.27:
16 Reduction in AFOLU GHG emissions from 2019. The AFOLU CO2 estimates in this figure are not
17 necessarily comparable with country GHG inventories (see Chapter 7).
18 3.4.6 Agriculture, Forestry and Other Land Use (AFOLU)
19 Mitigation pathways show substantial reductions in CO2 emissions, but more modest reductions in
20 AFOLU CH4 and N2O emissions (high confidence) (Popp et al. 2017; Roe et al. 2019; Reisinger et al.
21 2021) (Figure 3.27). Pathways limiting warming to likely 2C or below are projected to reach net zero
22 CO2 emissions in the AFOLU sector around 2033 (2024-2060); however AFOLU CH4 and N2O
23 emissions remain positive in all pathways (Figure 3.27). While IAMs include many land-based
24 mitigation options, these models exclude several options with large mitigation potential, such as
25 biochar, agroforestry, restoration/avoided conversion of coastal wetlands, and restoration/avoided
26 conversion of peatland (Smith et al. 2019; IPCC 2019a) (see also Chapter 7, Section 3.4). Sectoral
27 studies show higher mitigation potential than IAM pathways, as these studies include more mitigation
28 options than IAMs (medium confidence) (Chapter 7).
29
FOOTNOTE 15 Some studies calculate emissions reductions in 2050 compared to 2014, while others note
emissions reductions in 2060 relative to 2018.
Do Not Cite, Quote or Distribute 3-64 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1
2 Figure 3.27: Reduction in AFOLU GHG emissions from 2019. The AFOLU CO 2 estimates in this figure
3 are not necessarily comparable with country GHG inventories (see Chapter 7).
4 Limiting warming to likely 2C or below can result in large scale transformation of the land surface
5 (high confidence) (Popp et al. 2017; Rogelj et al. 2018a,b; Brown et al. 2019; Roe et al. 2019). The
6 scale of land transformation depends, inter alia, on the temperature goal and the mitigation options
7 included (Popp et al. 2017; Rogelj et al. 2018a; IPCC 2019a). Pathways with more demand-side
8 mitigation options show less land transformation than those with more limited options (van Vuuren et
9 al. 2018; Grubler et al. 2018; IPCC 2019a). Most of these pathways show increases in forest cover,
10 with an increase of 322 million ha (-67 to 890 million ha) in 2050 in 1.5°C pathways with no or
11 limited overshoot, whereas bottom up models portray an economic potential of 300-500 million ha of
12 additional forest (Chapter 7). Many IAM pathways also include large amounts of energy cropland
13 area, to supply biomass for bioenergy and BECCS, with 199 (56-482) million ha in 2050 in 1.5°C
14 pathways with no or limited overshoot. Large land transformations, such as afforestation/reforestation
15 and widespread planting of energy crops, can have implications for biodiversity and sustainable
16 development (see Section 3.7, Chapter 7 - Subsection 7.7.4, Chapter 12 - Section 12.5).
Do Not Cite, Quote or Distribute 3-65 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1
2 Figure 3.28: Change in Land Cover from 2019 in million hectares. Positive values indicate an increase in
3 area.
4 Delayed mitigation has implications for land use transitions (Hasegawa et al. 2021a). Delaying
5 mitigation action can result in a temporary overshoot of temperature and large-scale deployment of
6 CDR in the second half of the century to reduce temperatures from their peak to a given level (Smith
7 et al. 2019; Hasegawa et al. 2021a). IAM pathways rely on afforestation and BECCS as CDR
8 measures, so delayed mitigation action results in substantial land use change in the second half of the
9 century with implications for sustainable development (Hasegawa et al. 2021a) (see also Section 3.7).
10 Shifting to earlier mitigation action reduces the amount of land required for this, though at the cost of
11 larger land use transitions earlier in the century (Hasegawa et al. 2021a). Earlier action could also
12 reduce climate impacts on agriculture and land-based mitigation options (Smith et al. 2019).
13
14 Some AFOLU mitigation options can enhance vegetation and soil carbon stocks such as reforestation,
15 restoration of degraded ecosystems, protection of ecosystems with high carbon stocks and changes to
16 agricultural land management to increase soil carbon (high confidence) (Fuss et al. 2018; Griscom et
17 al. 2017; de Coninck et al. 2018; Smith et al. 2019) (WGIII Chapter 7). The timescales associated
18 with these options indicate that carbon sinks in terrestrial vegetation and soil systems can be
19 maintained or enhanced so as to contribute towards long-term mitigation (high confidence); however,
20 many AFOLU mitigation options do not continue to sequester carbon indefinitely (IPCC 2019a; Fuss
21 et al. 2018; de Coninck et al. 2018)(WGIII Chapter 7). In the very long term (latter part of the century
22 and beyond), it will become more challenging to continue to enhance vegetation and soil carbon
23 stocks, so that the associated carbon sinks could diminish or even become sources (high confidence)
24 (IPCC 2019a; de Coninck et al. 2018) (WGI Chapter 5). Sustainable forest management, including
25 harvest and forest regeneration, can help to remediate and slow any decline in the forest carbon sink,
26 for example by restoring degraded forest areas, and so go some way towards addressing the issue of
27 sink saturation (IPCC 2019) (WGI Chapter 5; WGIII Chapter 7). The accumulated carbon resulting
28 from mitigation options that enhance carbon sequestration (e.g., reforestation, soil carbon
29 sequestration) is also at risk of future loss due to disturbances (e.g., fire, pests) (Anderegg et al. 2020;
30 Boysen et al. 2017; IPCC 2019a; Smith et al. 2019; Fuss et al. 2018; de Coninck et al. 2018)(WGI
Do Not Cite, Quote or Distribute 3-66 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 Chapter 5). Maintaining the resultant high vegetation and soil carbon stocks could limit future land
2 use options, as maintaining these carbon stocks would require retaining the land use and land cover
3 configuration implemented to achieve the increased stocks.
4
5 Anthropogenic land CO2 emissions and removals in IAM pathways cannot be directly compared with
6 those reported in national GHG inventories (high confidence) (Grassi et al. 2018, 2021) (Chapter 7.2).
7 Due to differences in definitions for the area of managed forests and what is emissions and removals
8 are considered anthropogenic, the reported anthropogenic land CO2 emissions and removals differ by
9 ~5.5 GtCO2 yr-1 between IAMs, which rely on bookkeeping approaches (e.g., (Houghton and
10 Nassikas 2017), and national GHG inventories (Grassi et al. 2021). Such differences in definitions can
11 alter the reported time at which anthropogenic net zero CO2 emissions are reached for a given
12 emission scenario. Using national inventories would lead to an earlier reported time of net zero (van
13 Soest et al. 2021b) or to lower calculated cumulative emissions until the time of net zero (Grassi et al.
14 2021) as compared to IAM pathways. The numerical differences are purely due to differences in the
15 conventions applied for reporting the anthropogenic emissions and do not have any implications for
16 the underlying land-use changes or mitigation measures in the pathways. Grassi et al. (Grassi et al.
17 2021) offers a methodology for adjusting to reconcile these differences and enable a more accurate
18 assessment of the collective progress achieved under the Paris Agreement (Chapter 7, Cross-Chapter
19 Box 6 in Chapter 7).
20
21 3.4.7 Other Carbon Dioxide Removal Options
22
23 Table 3.5: Carbon dioxide removal in assessed pathways. Scenarios are grouped by temperature
24 categories, as defined in section 3.2.4. Quantity indicates the median and 5-95% range of cumulative
25 sequestration from 2020 to 2100 in GtCO2. Count indicates the number of scenarios with positive values
26 for that option.
CDR Option Below 1.5°C with no or Below 1.5°C with high Likely below 2°C
limited OS OS
Quantity Count Quantity Count Quantity Count
Total CDR 584 (192 to 95 645 (333 to 123 533 (193 to 294
959) 1221) 895)
CO2 removal 262 (17 to 64 330 (28 to 82 209 (20 to 415) 196
on managed 397) 439)
land including
A/R
BECCS 334 (32 to 91 464 (226 to 122 291 (174 to 294
780) 842) 653)
Enhanced 0 (0 to 47) 2 0 (0 to 0) 1 0 (0 to 0) 1
weathering
DACCS 30 (0 to 308) 31 109 (0 to 24 19 (0 to 253) 91
539)
27
28 This subsection includes other CDR options not discussed in the previous subsections, including
29 direct air carbon capture and storage (DACCS), enhanced weathering, and ocean-based approaches,
30 focusing on the role of these options in long-term mitigation pathways, using both IAMs (Rickels et
31 al. 2018; Realmonte et al. 2019; Chen and Tavoni 2013; Marcucci et al. 2017; Strefler et al. 2021a;
32 Fuhrman et al. 2019, 2020, 2021; Akimoto et al. 2021) and non-IAMs (Fuss et al. 2013; González and
33 Ilyina 2016; Bednar et al. 2021; Shayegh et al. 2021). There are other options discussed in the
34 literature, like methane capture (Jackson et al. 2019), however, the role of these options in long-term
35 mitigation pathways has not been quantified and are thus excluded here. Chapter 12 includes a more
Do Not Cite, Quote or Distribute 3-67 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 detailed description of the individual technologies, including their costs, potentials, financing, risks,
2 impacts, maturity and upscaling.
3 Very few studies and pathways include other CDR options (Table 3.5). Pathways with DACCS
4 include potentially large removal from DACCS (up to 37 GtCO2 yr-1 in 2100) in the second half of the
5 century (Realmonte et al. 2019; Marcucci et al. 2017; Chen and Tavoni 2013; Fuhrman et al. 2020,
6 2021; Shayegh et al. 2021; Akimoto et al. 2021) and reduced cost of mitigation (Bistline and Blanford
7 2021; Strefler et al. 2021a). At large scales, the use of DACCS has substantial implications for energy
8 use, emissions, land, and water; substituting DACCS for BECCS results in increased energy usage,
9 but reduced land use change and water withdrawals (Fuhrman et al., 2020, 2021; Chapter 12.3.2;
10 IPCC WGI Chapter 5). The level of deployment of DACCS is sensitive to the rate at which it can be
11 scaled up, the climate goal or carbon budget, the underlying socioeconomic scenario, the availability
12 of other decarbonization options, the cost of DACCS and other mitigation options, and the strength of
13 carbon cycle feedbacks (Honegger and Reiner 2018; Fuss et al. 2013; Fuhrman et al. 2021; Bistline
14 and Blanford 2021; Chen and Tavoni 2013; Strefler et al. 2021a; Fuhrman et al. 2020; Realmonte et
15 al. 2019) (IPCC WGI Chapter 5). Since DACCS consumes energy, its effectiveness depends on the
16 type of energy used; the use of fossil fuels would reduce its sequestration efficiency (Creutzig et al.
17 2019; Babacan et al. 2020; NASEM 2019). Studies with additional CDR options in addition to
18 DACCS (e.g., enhanced weathering, BECCS, afforestation, biochar, and soil carbon sequestration)
19 find that CO2 removal is spread across available options (Holz et al. 2018; Strefler et al. 2021a).
20 Similar to DACCS, the deployment of deep ocean storage depends on cost and the strength of carbon
21 cycle feedbacks (Rickels et al. 2018).
22
23
24 3.5 Interaction between near-, medium- and long-term action in
25 mitigation pathways
26 This section assesses the relationship between long-term climate goals and short- to medium-term
27 emissions reduction strategies based on the mitigation pathway literature. After an overview of this
28 relationship (3.5.1), it provides an assessment of what currently planned near-term action implies for
29 limiting warming to 1.5-2°C (3.5.2), and to what extent pathways with accelerated action beyond
30 current NDCs can improve the ability to keep long-term targets in reach (3.5.3).
31 The assessment in this section shows that if mitigation ambitions in current NDCs16 are followed until
32 2030, leading to estimated emissions of 47-57 GtCO2-eq in 203017 (Chapter 4.2.2), it is no longer
33 possible to stay below 1.5°C warming with no or limited overshoot (high confidence). Instead, it
34 would entail high overshoot (typically >0.1oC) and reliance on net negative CO2 emissions with
35 uncertain potential to return warming to 1.5°C by the end of the century. It would also strongly
36 increase mitigation challenges to likely limit warming to 2°C (high confidence). GHG emissions
FOOTNOTE 16 The term “current NDCs” used in this section and throughout the report refers to the most
recent nationally determined contributions submitted to the UNFCCC as well as those publicly announced with
sufficient detail on targets, but not yet submitted, up to 11 October 2021, and reflected in studies published up to
11 October 2021. In contrast, “original NDCs” refers to nationally determined contributions that were initially
submitted to the Paris Agreement by parties, largely reflecting the state of submissions until the year 2019. See
Chapter 4.2.
FOOTNOTE 17 In this section, the emissions range associated with current (or original) NDCs refer to the
combined emissions ranges from the two cases of implementing only the unconditional elements of current
NDCs (50-57 GtCO2-eq) and implementing both unconditional and conditional elements of current NDCs (47-
53 GtCO2-eq), if not specified otherwise.
Do Not Cite, Quote or Distribute 3-68 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 reductions would need to abruptly increase after 2030 to an annual average rate of 1.3-2.1 GtCO2-eq
2 during the period 2030-2050, around 70% higher than in mitigation pathways assuming immediate
3 action18 to likely limit warming to 2°C. The higher post-2030 reduction rates would have to be
4 obtained in an environment of continued build-up of fossil fuel infrastructure and less development of
5 low carbon alternatives until 2030. A lock-in into fossil-fuel intensive production systems (carbon
6 lock-in) will increase the societal, economic and political strain of a rapid low-carbon transition after
7 2030 (high confidence).
8 The section builds on previous assessments in the IPCC’s Fifth Assessment Report (Clarke et al.
9 2014) and Special Report on 1.5°C Warming (Rogelj et al. 2018a). The literature assessed in these
10 two reports has focused on delayed action until 2030 in the context of limiting warming to 2°C (den
11 Elzen et al. 2010; van Vuuren and Riahi 2011; Kriegler et al. 2015; Luderer et al. 2013, 2016; Riahi et
12 al. 2015; Rogelj et al. 2013a) and 1.5°C (Strefler et al. 2018; Luderer et al. 2018; Rogelj et al. 2013b).
13 Here we provide an update of these assessments drawing on the most recent literature on global
14 mitigation pathways. New studies have focused, inter alia, on constraining near term developments by
15 peak warming limits (Rogelj et al. 2019b; Strefler et al. 2021b; Riahi et al. 2021) and updating
16 assumptions about near- and medium term emissions developments based on national plans and long-
17 term strategies (Roelfsema et al. 2020) (Chapter 4.2). Several studies have explored new types of
18 pathways with accelerated action bridging between current policy plans and the goal of limiting
19 warming below 2°C (Kriegler et al. 2018a; van Soest et al. 2021a) and looked at hybrid international
20 policy regimes to phase in global collective action (Bauer et al. 2020).
21
22 3.5.1 Relationship between long-term climate goals and near- to medium-term emissions
23 reductions
24 The close link between cumulative CO2 emissions and warming has strong implications for the
25 relationship between near-, medium-, and long-term climate action to limit global warming. The AR6
26 WGI Assessment has estimated a remaining carbon budget of 500 (400) GtCO2 from the beginning of
27 2020 onwards for staying below 1.5°C with 50% (67%) likelihood, subject to additional uncertainties
28 about historic warming and the climate response and variations in warming from non-CO2 climate
29 forcers (Canadell and Monteiro 2019) (see also WGI Chapter 5, Section 5.5). For comparison, if
30 current CO2 emissions of more than 40 GtCO2 are keeping up until 2030, more than 400 GtCO2 will
31 be emitted during 2021–2030, already exhausting the remaining carbon budget for 1.5°C by 2030.
32 The relationship between warming limits and near term action is illustrated in Figure 3.29, using a set
33 of 1.5-2°C scenarios with different levels of near term action, overshoot and non-CO2 warming
34 contribution from a recent study (Riahi et al. 2021). In general, the more CO2 is emitted until 2030,
35 the less CO2 can be emitted thereafter to stay within a remaining carbon budget and below a warming
36 limit. Scenarios with immediate action to observe the warming limit give the longest time to exhaust
37 the associated remaining carbon budget and reach net zero CO2 emissions (light blue line in Figure
38 3.29, Cross-Chapter Box 3 in this chapter). In comparison, following projected NDC emissions until
39 2030 would imply a more pronounced drop in emissions from 2030 levels to net zero to make up for
40 the additional near-term emissions (orange lines in Figure 3.29). If such a drop does not occur, the
41 remaining carbon budget is exceeded and net negative CO2 emissions are required to return global
42 mean temperature below the warming limit (black lines in Figure 3.29) (Fuss et al. 2014; Clarke et al.
43 2014; Rogelj et al. 2018a).
FOOTNOTE 18 In this section, near term emissions outcomes are often compared to near term emissions in
mitigation pathways with immediate action towards a warming limit. These are defined as pathways that
immediately after imposing the warming limit turn to global mitigation action tapping into all least cost
abatement options available globally. These pathways are often called cost-effective or least-cost pathways in
the literature. The onset of immediate action is defined by scenario design and is typically chosen to be the first
time step after 2020 in immediate action scenarios assessed in this report.
Do Not Cite, Quote or Distribute 3-69 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 The relationship between warming limits and near-term action is also affected by the warming
2 contribution of non-CO2 greenhouse gases and other short lived climate forcers (3.3, Working Group I
3 6.7). The estimated budget values for limiting warming to 1.5-2°C already assume stringent
4 reductions in non-CO2 greenhouse gases and non-CO2 climate forcing as found in 1.5-2°C pathways
5 (3.3, Cross-Working Group Box 1; Working Group I 5.5, Box 5.2). Further variations in non-CO2
6 warming observed across 1.5-2°C pathways can vary the median estimate for the remaining carbon
7 budget by 220 GtCO2 (Working Group I 5.5). In 1.5-2°C pathways, the non-CO2 warming
8 contribution differs strongly between the near-, medium and long term. Changes to the atmospheric
9 composition of short-lived climate forcers dominate the warming response in the near term (Working
10 Group I 6.7). CO2 reductions are combined with strong reductions in air pollutant emissions due to
11 rapid reduction in fossil fuel combustion and in some cases the assumption of stringent air quality
12 policies (Rao et al. 2017b; Smith et al. 2020c). As air pollutants exert a net cooling effect, their
13 reduction drives up non-CO2 warming in the near term, which can be attenuated by the simultaneous
14 reduction of methane and black carbon (Smith et al. 2020b; Shindell and Smith 2019) (Working
15 Group I 6.7). After 2030, the reduction in methane concentrations and associated reductions in
16 tropospheric ozone levels tend to dominate so that a peak and decline in non-CO2 forcing and non-
17 CO2 induced warming can occur before net zero CO2 is reached (Figure 3.29) (Rogelj et al. 2018a)
18 The more stringent the reductions in methane and other short-lived warming agents like black carbon,
19 the lower this peak and the earlier the decline of non-CO2 warming, leading to a reduction of warming
20 rates and overall warming in the near to medium term (Smith et al. 2020b; Harmsen et al. 2020). This
21 is important for keeping warming below a tight warming limit that is already reached around mid-
22 century as is the case in 1.5°C pathways (Xu and Ramanathan 2017). Early and deep reductions of
23 methane emissions, and other short-lived warming agents like black carbon, provide space for residual
24 CO2-induced warming until the point of net zero CO2 emissions is reached (purple lines in Figure
25 3.29). Such emissions reductions have also been advocated due to co-benefits for, e.g., reducing air
26 pollution (Rao et al. 2016; Shindell et al. 2017a, 2018; Shindell and Smith 2019; Rauner et al. 2020a;
27 Vandyck et al. 2020).
Do Not Cite, Quote or Distribute 3-70 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1
2 Figure 3.29: Illustration of emissions and climate response in four mitigation pathways with different
3 assumptions about near term policy developments, global warming limit and non-CO2 warming
4 contribution drawn from Riahi et al. (2021). Shown are (a) CO2 emissions trajectories, (b) cumulative
5 CO2 emissions, (c) effective non-CO2 radiative forcing, and (d) the resulting estimate of the 67 th percentile
6 of global mean temperature response relative to 1850-1900. Light blue lines show a scenario that acts
7 immediately on a remaining carbon budget of 900 GtCO 2 from 2020 without allowing net negative CO 2
8 emissions, i.e., temporary budget overshoot (COFFEE 1.1, Scenario EN_NPi2020_900). Orange and black
9 lines show scenarios drawn from the same model that follow the NDCs until 2030 and thereafter
10 introduce action to stay within the same budget – in one case excluding net negative CO2 emissions like
11 before (orange lines; COFFEE 1.1., Scenario EN-INDCi2030_900) and in the other allowing for a
12 temporary overshoot of the carbon budget until 2100 (black lines; COFFEE 1.1., Scenario EN-
13 INDCi2030_900f). Light blue lines describe a scenario following the NDCs until 2030, and then aiming for
14 a higher budget of 2300 GtCO2 without overshoot (AIM/CGE 2.2, Scenario EN-INDCi2030_1200). It is
15 drawn from another model which projects a lower anthropogenic non-CO2 forcing contribution and
16 therefore achieves about the same temperature outcome as the other two non-overshoot scenarios despite
17 the higher CO2 budget. Grey funnels include the trajectories from all scenarios that likely limit warming
18 to 2°C (Category C3). Historical CO2 emissions until 2019 are from Chapter SM.2.1 EDGAR v6.0.
19 The relationship between long-term climate goals and near term action is further constrained by
20 social, technological, economic and political factors (Aghion et al. 2019; Mercure et al. 2019; van
21 Sluisveld et al. 2018b; Cherp et al. 2018; Jewell and Cherp 2020; Trutnevyte et al. 2019b). These
22 factors influence path dependency and transition speed (Vogt-Schilb et al. 2018; Pahle et al. 2018).
23 While detailed integrated assessment modelling of global mitigation pathways accounts for
24 technology inertia (Bertram et al. 2015a; Mercure et al. 2018) and technology innovation and
25 diffusion (Wilson et al. 2013; van Sluisveld et al. 2018a; Luderer et al. 2021), there are limitations in
26 capturing socio-technical and political drivers of innovation, diffusion and transition processes
27 (Keppo et al. 2021; Gambhir et al. 2019; Hirt et al. 2020; Köhler et al. 2019). Mitigation pathways
28 show a wide range of transition speeds that have been interrogated in the context of socio-technical
Do Not Cite, Quote or Distribute 3-71 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 inertia (Gambhir et al. 2017; Kefford et al. 2018; Kriegler et al. 2018a; Brutschin et al. 2021) vs.
2 accelerating technological change and self-enforcing socio-economic developments (Creutzig et al.
3 2017; Zenghelis 2019) (Section 3.8). Diagnostic analysis of detailed IAMs found a lag of 8-20 years
4 between the convergence of emissions pricing and the convergence of emissions response after a
5 period of differentiated emission prices (Harmsen et al. 2021). This provides a measure of the inertia
6 to changing policy signals in the model response. It is about half the timescale of 20-40 years
7 observed for major energy transitions (Grubb et al. 2021). Hence, the mitigation pathways assessed
8 here capture socio-technical inertia in reducing emissions, but the limited modelling of socio-political
9 factors may alter the extent and persistence of this inertia.
10
11 3.5.2 Implications of near-term emission levels for keeping long-term climate goals within
12 reach
13 The implications of near-term climate action for long-term climate outcomes can be explored by
14 comparing mitigation pathways with different near-term emissions developments aiming for the same
15 climate target (Vrontisi et al. 2018; Riahi et al. 2015; Roelfsema et al. 2020). A particular example is
16 the comparison of cost-effective pathways with immediate action to limit warming to 1.5-2°C with
17 mitigation pathways pursuing more moderate mitigation action until 2030. After the adoption of the
18 Paris Agreement, near term action was often modelled to reflect conditional and unconditional
19 elements of originally submitted NDCs (2015-2019) (Fawcett et al. 2015; Fujimori et al. 2016a;
20 Kriegler et al. 2018a; Vrontisi et al. 2018; Roelfsema et al. 2020). The most recent modelling studies
21 also include submission of updated NDCs or announcements of planned updates in the first half of
22 2021 (Network for Greening the Financial System 2021; Riahi et al. 2021). Emissions levels under
23 current NDCs (see footnote 15) are assessed to range between 47–57 GtCO2-eq in 2030 (Chapter
24 4.2.2). This assessed range corresponds well to 2030 emissions levels in 2°C mitigation pathways in
25 the literature that are designed to follow the original or current NDCs until 203019. For the 139
26 scenarios of this kind that are collected in the AR6 scenario database and that still likely limit
27 warming to 2°C, the 2030 emissions range is 52.5 (44.5-57) GtCO2-eq (based on native model
28 reporting) and 52 (46.5-56) GtCO2-eq, respectively (based on harmonized emissions data for climate
29 assessment (Annex III part II section 2.4); median and 5th to 95th percentile). This close match allows
30 a robust assessment of the implications of implementing current NDCs for post-2030 mitigation
31 efforts and warming outcomes based on the literature and the AR6 scenarios database.
32 The assessed emission ranges from implementing the unconditional (unconditional and conditional)
33 elements of current NDCs implies an emissions gap to cost-effective mitigation pathways of 20-26
34 (16-24) GtCO2-eq in 2030 for limiting warming to 1.5°C with no or limited overshoot and 10-17 (7-
35 14) GtCO2-eq in 2030 for likely limiting warming to 2°C (Chapter 4, Cross-Chapter Box 2 in chapter
36 2). The emissions gap gives rise to a number of mitigation challenges (Rogelj et al. 2013a; Kriegler et
37 al. 2013a, 2018b,a; Riahi et al. 2015; Luderer et al. 2013, 2018; Fujimori et al. 2016b; Fawcett et al.
38 2015; Strefler et al. 2018; Winning et al. 2019; UNEP 2020; SEI et al. 2020): (i) larger transitional
39 challenges post-2030 to still remain under the warming limit (, in particular higher CO2 emissions
40 reduction rates and technology transition rates required during 2030-2050; (ii) larger lock-in into
41 carbon-intensive infrastructure and increased risk of stranded fossil fuel assets (3.5.2.2); and (iii)
42 larger reliance on CDR to reach net zero CO2 more rapidly and compensate excess emissions in the
43 second half of the century (3.5.2.1). All these factors exacerbate socio-economic strain of
FOOTNOTE 19 The intended design of mitigation pathways in the literature can be deduced from underlying
publications and study protocols. This information was collected as part of this assessment to establish a
categorization of policy assumptions underpinning the mitigation pathways collected in the AR6 scenario
database (3.2; Annex III Part II section 3.2).
Do Not Cite, Quote or Distribute 3-72 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 implementing the transition, leading to an increased risk of overshooting the warming and a higher
2 risk of climate change impacts (Drouet et al. 2021).
3 The challenges are illustrated in Table 3.6 and Figure 3.30, surveying global mitigation pathways in
4 the literature that were collected in the AR6 scenarios database. There is a clear trend of increasing
5 peak warming with increasing 2030 GHG emission levels (Figure 3.30a+b). In particular, there is no
6 mitigation pathway designed to follow the NDCs until 2030 in 2030 that can limit warming to 1.5°C
7 with no or limited overshoot. Our assessment confirms the finding of the IPCC Special Report on
8 1.5°C Warming for the case of current NDCs, including updates until 11 October 2021 that were
9 assessed in the literature, that pathways following the NDCs until 2030 “would not limit global
10 warming to 1.5°C, even if supplemented by very challenging increases in the scale and ambition of
11 emissions reductions after 2030” (SR1.5 SPM). This assessment is now more robust than in SR1.5 as
12 it is based on a larger set of 1.5-2°C pathways with better representation of current trends and plans
13 covering a wider range of post-2030 emissions developments. In particular, a recent multi-model
14 study limiting peak cumulative CO2 emissions for a wide range of carbon budgets and immediate vs
15 NDC-type action until 2030 established a feasibility frontier for the existence of such pathways across
16 participating models (Riahi et al. 2021).
17 2030 emissions levels in the NDC range also tighten the remaining space to likely limit warming 2°C.
18 As shown in Figure 3.30b, the 67th percentile of peak warming reaches values above 1.7°C warming
19 in pathways with 2030 emissions levels in this range. To still have a likely chance to stay below 2°C,
20 the global post-2030 GHG emission reduction rates would need to be abruptly raised in 2030 from 0-
21 0.8 GtCO2-eq yr-1 to an average of 1.3-2.1 GtCO2-eq yr-1 during the period 2030-2050 (Figure 3.30c),
22 around 70% of that in immediate mitigation pathways confirming findings in the literature (Winning
23 et al. 2019). Their average reduction rate of 0.6-1.4 GtCO2 yr-1 would already be unprecedented at the
24 global scale and with a few exceptions national scale for an extended period of time (Riahi et al.
25 2015). For comparison, the impact of COVID-19 on the global economy is projected to lead to a
26 decline of ca. 2.5-3 GtCO2 of global CO2 emissions from fossil fuel and industry in 2020
27 (Friedlingstein et al. 2020) (Chapter 2.2).
28 The increased post-2030 transition challenge in mitigation pathways with moderate near term action is
29 also reflected in the timing of reaching net zero CO2 emissions (Figure 3.30d and Table 3.6) (Cross-
30 Chapter Box 3 in this chapter). As 2030 emission levels and the cumulated CO2 emissions until 2030
31 increase, the remaining time for dropping to net zero CO2 and staying within the remaining carbon
32 budget shortens (Figure 3.29). This gives rise to an inverted v-shape of the lower bound on the year of
33 reaching net zero as a function of 2030 emissions levels. Reaching low in 2030 facilitates reaching net
34 zero early (left leg of the inverted v), but staying high until 2030 also requires to reach net zero CO 2
35 faster to compensate for higher emissions early on (right leg of the inverted v). Overall, there is a
36 considerable spread of the timing of net zero CO2 for any 2030 emissions level due to variation in the
37 timing of spending the remaining carbon budget and the non-CO2 warming contribution (Cross-
38 Chapter Box 3 in this chapter).
Do Not Cite, Quote or Distribute 3-73 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1
2 Figure 3.30: Relationship between level of global GHG emissions in 2030 and selected indicators as listed
3 in the panel titles for scenarios collected in the AR6 scenario database. Emissions data based on
4 harmonized emissions used for the climate assessment. All scenarios likely to limit warming to 2°C or
5 below are coloured blue or red (see p67 peak warming in panel b). The large majority of blue coloured
6 scenarios act immediately on the temperature target, while red coloured scenarios depict all those that
7 were designed to follow the NDCs or lesser action until 2030 and orange coloured scenarios comprise a
8 small set of pathways with additional regulatory action beyond NDCs (3.5.3). Grey coloured scenarios
9 exceed the 2°C (p67), either by temporary overshoot or towards the end of the century. Large markers
10 denote the 5 illustrative mitigation pathways (legend in Panel h; Section 3.2). Shaded yellow areas depict
11 the estimated range of 2030 emissions from current NDCs (Chapter 4.2.2). Dotted lines are inserted in
12 some panels to highlight trends in the dependency of selected output variables on 2030 GHG emissions
13 levels (see text)
14 There is also a profound impact on the underlying transition of energy and land use (Table 3.6 and
15 Figure 3.30f-h). Scenarios following NDCs until 2030 show a much smaller reduction in fossil fuel
16 use, only half of the growth in renewable energy use, and a smaller reduction in CO 2 and CH4 land
17 use emissions in 2030 compared to immediate action scenarios. This is then followed by a much faster
18 reduction of land use emissions and fossil fuels, and a larger increase of nuclear energy, bioenergy
19 and non-biomass renewable energy during the medium term in order to get close to the levels of the
Do Not Cite, Quote or Distribute 3-74 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 immediate action pathways in 2050. This is combined with a larger amount of net negative CO2
2 emissions that are used to compensate the additional emissions before 2030. The faster transition
3 during 2030-2050 is taking place from a greater investment in fossil fuel infrastructure and lower
4 deployment of low carbon alternatives in 2030, adding to the socio-economic challenges to realize the
5 higher transition rates (Section 3.5.2.2). Therefore, these pathways also show higher mitigation costs,
6 particularly during the period 2030-2050, than immediate action scenarios (3.6.1, Figure 3.34d) (Liu
7 et al. 2016; Vrontisi et al. 2018; Kriegler et al. 2018a).. Given these circumstances and the fact the
8 modelling of socio-political and institutional constraints is limited in integrated assessment models
9 (Keppo et al. 2021; Gambhir et al. 2019; Hirt et al. 2020; Köhler et al. 2019), the feasibility of
10 realizing these scenarios is assessed to be lower (Gambhir et al. 2017; Napp et al. 2017; Brutschin et
11 al. 2021) (cf. Section 3.8), increasing the risk of an overshoot of climate goals.
12
13 Table 3.6: Comparison of key scenario characteristics for four scenario classes (see Table 3.2): (i)
14 immediate action to limit warming to 1.5°C with no or limited overshoot, (ii) near team action following
15 the NDCs until 2030 and returning warming to below 1.5°C (50% chance) by 2100 with high overshoot,
16 (iii) immediate action to likely limit warming to 2°C, and (iv) near term action following the NDCs until
17 2030 followed by post-2030 action to likely limit warming to 2°C. The classes (ii) and (iv) comprise the
18 large majority of scenarios indicated by red dots, and the classes (i) and (iii) the scenarios depicted by
19 blue dots in Figure 3.30. Shown are median and interquartile ranges (in brackets) for selected global
20 indicators. Emissions ranges are based on harmonized emissions data for the climate assessment with the
21 exception of land use CO2 emissions for which uncertainty in historic estimates is large. Numbers are
22 rounded to the nearest 5.
23
1.5°C 1.5°C by 2100 Likely < 2°C
Global indicators Immediate action, NDCs until 2030, Immediate action NDCs until 2030
with no or limited with overshoot before (C3a, 204 (C3b; 97 scenarios)
overshoot 2100 (subset of 42 scenarios)
(C1, 97 scenarios) scenarios in C2)
Kyoto GHG emissions in 2030 (% rel to 2019) -45 (-50,-40) -5 (-5,0) -25 (-35,-20) -5 (-10,0)
in 2050 (% rel to 2019) -85 (-90,-80) -75 (-85,-70) -65 (-70,-60) -70 (-70,-60)
CO2 emissions change in 2030 (% rel to 2019) -50 (-60,-40) -5 (-5,0) -25 (-35,-20) -5 (-5,0)
in 2050 (% rel to 2019) -100 (-105,-95) -85 (-95,-80) -70 (-80,-65) -75 (-80,-65)
Net land use CO2 emissions in 2030 (% rel to 2019) -100 (-105,-95) -30 (-60,-20) -90 (-105,-75) -20 (-80,-20)
in 2050 (% rel to 2019) -150 (-200,-100) -135 (-165,-120) -135 (-185,-100) -130 (-145,-115)
CH4 emissions in 2030 (% rel to 2019) -35 (-40,-30) -5 (-5,0) -25 (-35,-20) -10 (-15,-5)
in 2050 (% rel to 2019) -50 (-60,-45) -50 (-60,-45) -45 (-50,-40) -50 (-65,-45)
Cumulative CCS until 2100 (GtCO2) 665 (520,900) 670 (535,865) 605 (490,895) 535 (440,725)
of which BECCS (GtCO2) 330 (250,560) 365 (280,590) 350 (240,455) 270 (240,400)
Cumulative net negative CO2 emissions
until 2100 (GtCO2) 190 (0,385) 320 (250,440) 10 (0,120) 70 (0,200)
Primary energy from coal in 2030 (% rel to 2019) -75 (-80,-65) -10 (-20,-5) -50 (-65,-35) -15 (-20,-10)
in 2050 (% rel to 2019) -95 (-100,-80) -90 (-100,-85) -85 (-100,-65) -80 (-90,-70)
Primary energy from oil in 2030 (% rel to 2019) -10 (-25,0) 5 (5,10) 0 (-10,10) 10 (5,10)
in 2050 (% rel to 2019) -60 (-75,-40) -50 (-65,-30) -30 (-45,-15) -40 (-55,-20)
Do Not Cite, Quote or Distribute 3-75 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
Primary energy from gas in 2030 (% rel to 2019) -10 (-30,0) 15 (10,25) 10 (0,15) 15 (10,15)
in 2050 (% rel to 2019) -45 (-60,-20) -45 (-55,-25) -10 (-35,15) -30 (-45,-5)
Primary energy from nuclear in 2030 (% rel to 2019)
40 (5,70) 10 (0,25) 35 (5,50) 10 (0,30)
in 2050 (% rel to 2019) 90 (10,305) 100 (40,135) 85 (30,200) 75 (30,120)
Primary energy from biomass in 2030 (% rel to 2019) 75 (55,130) 45 (20,75) 60 (35,105) 45 (10,80)
in 2050 (% rel to 2019) 290 (215,430) 230 (170,440) 240 (130,355) 260 (95,435)
Primary energy from non-biomass renewables
in 2030 (% rel to 2019) 225 (150,270) 100 (85,145) 150 (115,190) 115 (85,130)
in 2050 (% rel to 2019) 725 (540,955) 665 (515,925) 565 (415,765) 625 (545,705)
Carbon intensity of electricity in 2030 (% rel to 2019) -75 (-85,-70) -30 (-40,-30) -60 (-70,-50) -35 (-40,-30)
in 2050 (% rel to 2019) -100 (-100,-100) -100 (-100,-100) -95 (-100,-95) -100 (-100,-95)
Carbon intensity of non-electric final energy
consumption in 2030 (% rel to 2019) -40 (-50,-35) 0 (0,5) -20 (-30,-15) 0 (0,0)
in 2050 (% rel to 2019) -80 (-85,-75) -70 (-75,-70) -60 (-65,-55) -65 (-70,-55)
1
2
3 3.5.2.1 Overshoot and net negative CO2 emissions
4 If near to medium term emissions developments deplete the remaining carbon budget, the associated
5 warming limit will be overshot. Some pathways that return median warming to below 1.5°C by the
6 end of the century show mid-century overshoots of up to 1.8°C median warming. The overshoot tends
7 to be the higher, the higher 2030 emissions. Mitigation pathways with 2030 emissions levels in the
8 NDC range consistently overshoot 1.5°C by 0.15-0.3°C. This leads to higher risks from climate
9 change impacts during the time of overshoot compared to pathways that limit warming to 1.5°C with
10 no or limited overshoot (Tachiiri et al. 2019; Hofmann et al. 2019; Schleussner et al. 2016a; Mengel
11 et al. 2018; Drouet et al. 2021; Lenton et al. 2019). Furthermore, even if warming is reversed by net
12 negative emissions, other climate changes such as sea level rise would continue in their current
13 direction for decades to millennia (WGI Chapter 4.6 and Chapter 5.6).
14
15 Returning warming to lower levels requires net negative CO2 emissions in the second half of the
16 century (Fuss et al. 2014; Clarke et al. 2014; Rogelj et al. 2018a). The amount of net negative CO2
17 emissions in pathways limiting warming to 1.5-2°C climate goals varies widely, with some pathways
18 not deploying net negative CO2 emissions at all and others deploying up to -600 to -800 GtCO2. The
19 amount of net negative CO2 emissions tends to increase with 2030 emissions levels (Figure 3.30e and
20 Table 3.6). Studies confirmed the ability of net negative CO2 emissions to reduce warming, but
21 pointed to path dependencies in the storage of carbon and heat in the Earth system and the need for
22 further research particularly for cases of high overshoot (Keller et al. 2018a,b; Tokarska et al. 2019;
23 Zickfeld et al. 2016, 2021). WGI assessed the reduction in global surface temperature to be
24 approximately linearly related to cumulative CO2 removal and, with lower confidence, that the
25 amount of cooling per unit CO2 removed is approximately independent of the rate and amount of
26 removal (WGI TS.3.3.2). Still there remains large uncertainty about a potential asymmetry between
27 the warming response to CO2 emissions and the cooling response to net negative CO2 emissions
28 (Zickfeld et al. 2021). It was also shown that warming can adversely affect the efficacy of carbon
29 dioxide removal measures and hence the ability to achieve net negative CO2 emissions (Boysen et al.
30 2016).
31
Do Not Cite, Quote or Distribute 3-76 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 Obtaining net negative CO2 emissions requires massive deployment of carbon dioxide removal (CDR)
2 in the second half of the century, on the order of 220 (160-370) GtCO2 for each 0.1°C degree of
3 cooling (based on the assessment of the likely range of the transient response to cumulative CO2
4 emissions in WGI Chap. 5.5, not taking into account potential asymmetries in the temperature
5 response to CO2 emissions and removals). CDR is assessed in detail in Chapter 12.3 of this report (see
6 also Cross-Chapter Box 8 in Chapter 12). Here we only point to the finding that CDR ramp-up rates
7 and absolute deployment levels are tightly limited by techno-economic, social, political, institutional
8 and sustainability constraints (Smith et al. 2016; Boysen et al. 2017; Nemet et al. 2018; Fuss et al.
9 2018, 2020; Hilaire et al. 2019; Jia et al. 2019) (Chapter 12.3). CDR therefore cannot be deployed
10 arbitrarily to compensate any degree of overshoot. A fraction of models was not able to compute
11 pathways that would follow the mitigation ambition in unconditional and conditional NDCs until
12 2030 and return warming to below 1.5°C by 2100 (Luderer et al. 2018; Roelfsema et al. 2020; Riahi
13 et al. 2021).There exists a three-way trade-off between near-term emissions developments until 2030,
14 transitional challenges during 2030-50, and long-term CDR deployment post 2050 (Sanderson et al.
15 2016; Holz et al. 2018; Strefler et al. 2018). For example, Strefler et al. (2018) find that if CO2
16 emission levels stay around 40 GtCO2 until 2030, within the range of what is projected for current
17 unconditional and conditional NDCs, rather than being halved to 20 GtCO2 until 2030, CDR
18 deployment in the second half of the century would have to increase by 50%-100%, depending on
19 whether the 2030-2050 CO2 emissions reduction rate is doubled from 6% to 12% or kept at 6% per
20 year. This three-way trade-off has also been identified at the national level (Pan et al. 2020).
21
22 In addition to enabling a temporary budget overshoot by net negative CO2 emissions in the second
23 half of the century, CDR can also be used to compensate – on an annual basis - residual CO2
24 emissions from sources that are difficult to eliminate and to reach net zero CO2 emissions more
25 rapidly if deployed before this point (Kriegler et al. 2013b; Rogelj et al. 2018a). This explains its
26 continued deployment in pathways that exclude overshoot and net negative CO2 emissions (Riahi et
27 al. 2021). However, given the timescales that would likely be needed to ramp-up CDR to Gigaton
28 scale (Nemet et al. 2018), it can be expected to only make a limited contribution to reaching net zero
29 CO2 as fast as possible. In the vast majority (95%) of 1.5-2°C mitigation pathways assessed in this
30 report, cumulative CDR deployment did not exceed 100 GtCO2 until mid-century. This adds to the
31 risk of excessively relying on CDR to compensate for weak mitigation action until 2030 by either
32 facilitating massive net CO2 emissions reduction rates during 2030-2050 or allowing a high temporary
33 overshoot of 1.5°C until the end of the century. If international burden sharing considerations are
34 taken into account, the CDR penalty for weak action could increase further, in particular for
35 developed countries (Fyson et al. 2020). Further assessment of CDR deployment in 1.5-2°C
36 mitigation pathways is found in Section 3.4.7.
37
38 3.5.2.2 Carbon lock-in and stranded assets
39 There already exists a substantial and growing carbon lock-in today, as measured by committed
40 emissions associated with existing long-lived infrastructure {Chapter 2.7, Figure 2.31}. If existing
41 fossil-fuel infrastructure would continue to be operated as historically, they would entail CO 2
42 emissions exceeding the carbon budget for 1.5°C {Chapter 2.7.2, Figure 2.32}. However, owner-
43 operators and societies may choose to retire existing infrastructure earlier than in the past, and
44 committed emissions are thus contingent on the competitiveness of non-emitting alternative
45 technologies and climate policy ambition. Therefore, in mitigation pathways, some infrastructure may
46 become stranded assets. Stranded assets have been defined as “assets that have suffered from
47 unanticipated or premature write-downs, devaluations or conversion to liabilities” (Caldecott 2017).
48 A systematic map of the literature on carbon lock-in has synthesized quantification of stranded-assets
49 in the mitigation pathways literature, and showed that (i) coal power plants are the most exposed to
Do Not Cite, Quote or Distribute 3-77 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 risk of becoming stranded, (ii) delayed mitigation action increases stranded assets and (iii) sectoral
2 distribution and amount of stranded assets differ between countries (Fisch-Romito et al. 2020). There
3 is high agreement that existing fossil fuel infrastructure would need to be retired earlier than
4 historically, used less, or retrofitted with CCS, to stay within the remaining carbon budgets of limiting
5 warming to 1.5°C or 2°C (Johnson et al. 2016; Kefford et al. 2018; Pfeiffer et al. 2018; Cui et al.
6 2019; Fofrich et al. 2020; Rogelj et al. 2018a). Studies estimate that cumulative early retired power
7 plant capacities by 2060 can be up to 600 GW for gas and 1700 GW for coal (Iyer et al. 2015a;
8 Kefford et al. 2018), that only 42% of the total capital stock of both operating and planned coal-fired
9 powers plants can be utilized to be compatible with the 2°C target (Pfeiffer et al. 2018), and that coal-
10 fired power plants in scenarios consistent with keeping global warming below 2°C or 1.5°C retire one
11 to three decades earlier than historically has been the case (Cui et al. 2019; Fofrich et al. 2020). After
12 coal, electricity production based on gas is also projected to be phased out, with some capacity
13 remaining as back-up (van Soest et al. 2017a). Kefford et al. (2018) find USD541 billion worth of
14 stranded fossil fuel power plants could be created by 2060, with China and India the most exposed.
15 Some publications have suggested that stranded long-lived assets may be even more important outside
16 of the power sector. While stranded power sector assets by 2050 could reach up to USD1.8 trillion in
17 scenarios consistent with a 2°C target, Saygin et al. (2019) found a range of USD5-11 trillion in the
18 buildings sectors. Muldoon-Smith and Greenhalgh (2019) have even estimated a potential value at
19 risk for global real estate assets up to USD21 trillion. More broadly, the set of economic activities that
20 are potentially affected by a low carbon transition is wide and includes also energy-intensive
21 industries, transport and housing, as reflected in the concept of climate policy relevant sectors
22 introduced in Battiston et al. (2017). The sectoral distribution and amount of stranded assets differ
23 across countries (Fisch-Romito et al. 2020). Capital for fossil fuel production and distribution
24 represents a larger share of potentially stranded assets in fossil fuel producing countries such as the
25 United States and Russia. Electricity generation would be a larger share of total stranded assets in
26 emerging countries because this capital is relatively new compared to its operational lifetime.
27 Conversely, buildings could represent a larger part of stranded capital in more developed countries
28 such as the United States, EU or even Russia because of high market value and low turnover rate.
29 Many quantitative estimates of stranded assets along mitigation pathways have focused on fossil fuel
30 power plants in pathways characterized by mitigation ambition until 2030 corresponding to the
31 current NDCs followed by strengthened action afterwards to limit warming to 2°C or lower (Bertram
32 et al. 2015a; Iyer et al. 2015b; Lane et al. 2016; Farfan and Breyer 2017; van Soest et al. 2017a;
33 Luderer et al. 2018; Kriegler et al. 2018a; Cui et al. 2019; Saygin et al. 2019; SEI et al. 2020).
34 Pathways following current NDCs until 2030 do not show a significant reduction of coal, oil and gas
35 use (Figure 3.30f-h, Table 3.6) compared to immediate action pathways. Stranded coal power assets
36 are evaluated to be higher by a factor of 2-3 if action is strengthened after 2030 rather than now (Iyer
37 et al. 2015b; Cui et al. 2019). There is high agreement that the later climate policies are implemented,
38 the higher the expected stranded assets and the societal, economic and political strain of strengthening
39 action. Associated price increases for carbon-intensive goods and transitional macro-economic costs
40 have been found to scale with the emissions gap in 2030 (Kriegler et al. 2013a). At the aggregate level
41 of the whole global economy, Rozenberg et al. (2015) showed that each year of delaying the start of
42 mitigation decreases the required CO2 intensity of new production by 20-50 gCO2/USD. Carbon lock-
43 in can have a long-lasting effect on future emissions trajectories after 2030. Luderer et al. (2018)
44 compared cost-effective pathways with immediate action to limit warming to 1.5-2°C with pathways
45 following the NDCs until 2030 and adopting the pricing policy of the cost-effective pathways
46 thereafter and found that the majority of additional CO2 emissions from carbon lock-in occurs after
47 2030, reaching a cumulative amount of 290 (160–330) GtCO2 by 2100 (2.7.2). Early action and
48 avoidance of investments in new carbon-intensive assets can minimize these risks.
Do Not Cite, Quote or Distribute 3-78 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 The risk of stranded assets has implications for workers depending from those assets, asset owners,
2 assets portfolio managers, financial institutions and the stability of the financial system. Chapter 6
3 assesses the risks and implications of stranded assets for energy systems (6.7.3, Box 6.11) and fossil
4 fuels (6.7.4). The implications of stranded assets for inequality and just transition are assessed in
5 Chapter 17 (17.3.2.3). Chapter 15 assesses the literature on those implications for the financial system
6 as well as on coping options (15.5.2, 15.6.1).
7 On the other hand, mitigation, by limiting climate change, reduces the risk of destroyed or stranded
8 assets from the physical impacts of climate change on natural and human systems, from more
9 frequent, intense or extended extreme events and from sea level rise (O’Neill et al. 2020a). The
10 literature on mitigation pathways rarely includes an evaluation of stranded assets from climate change
11 impacts. Unruh (Unruh 2019) suggest that these are the real stranded assets of carbon lock-in and
12 could prove much more costly.
13
14 3.5.2.3 Global accelerated action towards long-term climate goals
15 A growing literature explores long-term mitigation pathways with accelerated near-term action going
16 beyond the NDCs (Jiang et al. 2017; Roelfsema et al. 2018; Graichen et al. 2017; Kriegler et al.
17 2018a; van Soest et al. 2021a; Fekete et al. 2021). Global accelerated action pathways are designed to
18 transition more gradually from current policies and planned implementation of NDCs onto a 1.5-2°C
19 pathway and at the same time alleviate the abrupt transition in 2030 that would be caused by
20 following the NDCs until 2030 and strengthening towards limiting warming to 2°C thereafter (Section
21 3.5.2). Therefore they have sometimes been called bridging scenarios / pathways in the literature (IEA
22 2011; Spencer et al. 2015; van Soest et al. 2021a). They rely on regionally differentiated regulatory
23 and pricing policies to gradually strengthening regional and sectoral action beyond the mitigation
24 ambition in the NDCs. There are limitations to this approach. The tighter the warming limit, the more
25 is disruptive action becoming inevitable to achieve the speed of transition that would be required
26 (Kriegler et al. 2018a). Cost effective pathways already have abrupt shifts in deployments,
27 investments and prices at the time a stringent warming limit is imposed reflecting the fact that the
28 overall response to climate change has so far been misaligned with long-term climate goals (Geiges et
29 al. 2019; Rogelj et al. 2016; Fawcett et al. 2015; Schleussner et al. 2016b). Disruptive action can help
30 to break lock-ins and enable transformative change (Vogt-Schilb et al. 2018).
31
32 The large literature on accelerating climate action was assessed in the IPCC Special Report on 1.5°C
33 Warming (de Coninck et al. 2018) and is taken up in this report primarily in Chapters 4, 13, and 14.
34 Accelerating climate action and facilitating transformational change requires a perspective on socio-
35 technical transitions (Geels et al. 2016b; Geels 2018; Geels et al. 2016a), a portfolio of policy
36 instruments to manage technological and environmental change (Goulder and Parry 2008; Fischer and
37 Newell 2008; Acemoglu et al. 2012, 2016), a notion of path dependency and policy sequencing
38 (Meckling et al. 2017; Pahle et al. 2018; Pierson 2000) and the evolvement of poly-centric governance
39 layers of institutions and norms in support of the transformation (Messner 2015; Leach et al. 2007;
40 Dietz et al. 2003). This subsection is focused on an assessment of the emerging quantitative literature
41 on global accelerated action pathways towards 1.5-2°C, which to a large extent abstracts from the
42 underlying processes and uses a number of stylized approaches to generate these pathways. A
43 representative of accelerated action pathways has been identified as one of the illustrative mitigation
44 pathways in this assessment (GS, Figure 3.31).
45
46 One approach relies on augmenting initially moderate emissions pricing policies with robust
47 anticipation of ratcheting up climate action in the future (Spencer et al. 2015). If announcements of
48 strong future climate policies are perceived to be credible, they can help to prevent carbon lock-in as
49 investors anticipating high future costs of GHG emissions would reduce investment into fossil fuel
50 infrastructure, such as coal power plants (Bauer et al. 2018b). However, the effectiveness of such
Do Not Cite, Quote or Distribute 3-79 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 announcements strongly hinges on their credibility. If investors believe that policy makers could drop
2 them if anticipatory action did not occur, they may not undertake such action.
3
4 Another approach relies on international cooperation to strengthen near term climate action. These
5 studies build on international climate policy architectures that could incentivize a coalition of like-
6 minded countries to raise their mitigation ambition beyond what is stated in their current NDC
7 (Graichen et al. 2017). Examples are the idea of climate clubs characterized by harmonized carbon
8 and technology markets (Pihl 2020; Nordhaus 2015; Keohane et al. 2017; Paroussos et al. 2019) and
9 the Powering Past Coal Alliance (Jewell et al. 2019). Paroussos et al. (2019) find economic benefits of
10 joining a climate club despite the associated higher mitigation effort, in particular due to access to
11 technology and climate finance. Graichen et al. (2017) find an additional reduction of 5-11 GtCO2eq
12 compared to the mitigation ambition in the NDCs from the successful implementation of international
13 climate initiatives. Other studies assess benefits from international transfers of mitigation outcomes
14 (Stua 2017; Edmonds et al. 2021). Edmonds et al. (2021) find economic gains from sharing NDC
15 emissions reduction commitments compared to purely domestic implementation of NDCs. If
16 reinvested in mitigation efforts, the study projects an additional reduction of 9 billion tonnes of CO2 in
17 2030.
18
19 The most common approach relies on strengthening regulatory policies beyond current policy trends,
20 also motivated by the finding that such policies have so far been employed more often than
21 comprehensive carbon pricing (Roelfsema et al. 2018; Kriegler et al. 2018a; van Soest et al. 2021a;
22 Fekete et al. 2021; IEA 2021a). Some studies have focused on generic regulatory policies such as low
23 carbon support policies, fossil fuel sunset policies, and resource efficiency policies (Bertram et al.
24 2015b; Hatfield-Dodds et al. 2017). Bertram et al. (2015b) found that a moderate carbon price
25 combined with a coal moratorium and ambitious low carbon support policies can limit efficiency
26 losses until 2030 if emissions pricing is raised thereafter to limit warming to 2°C. They also showed
27 that all three components are needed to achieve this outcome. Hatfield-Dodds et al. (2017) found that
28 resource efficiency can lower 2050 emissions by an additional 15-20% while boosting near-term
29 economic growth. The International Energy Agency (IEA 2021a) developed a detailed net zero
30 scenario for the global energy sector characterized by a rapid phase out of fossil fuels, a massive clean
31 energy and electrification push, and the stabilization of energy demand, leading to 10 GtCO 2 lower
32 emissions from energy use in 2030 than in a scenario following the announced pledges.
33
34 The Paris Agreement has spurred the formulation of NDCs for 2030 and mid-century strategies
35 around the world (cf. Chapter 4). This is giving researchers a rich empirical basis to formulate
36 accelerated policy packages taking national decarbonisation pathways as a starting point (van Soest et
37 al. 2017b; Waisman et al. 2019; Graichen et al. 2017; Jiang et al. 2017). The concept is to identify
38 good practice policies that had demonstrable impact on pushing low carbon options or reducing
39 emissions in a country or region and then consider a wider roll-out of these policies taking into
40 account regional specificities (den Elzen et al. 2015; Kuramochi et al. 2018; Roelfsema et al. 2018;
41 Kriegler et al. 2018a; Fekete et al. 2015, 2021). A challenge for this approach is to account for the fact
42 that policy effectiveness varies with different political environments in different geographies. As a
43 result, a global roll-out of good practice policies to close the emissions gap will still be an idealized
44 benchmark, but it is useful to understand how much could be gained from it.
45
46 Accelerated action pathways derived with this approach show considerable scope for narrowing the
47 emissions gap between pathways reflecting the ambition level of the NDCs and cost-effective
48 mitigation pathways in 2030. Kriegler et al. (2018a) find around 10 GtCO2eq lower emissions
49 compared to original NDCs from a global roll-out of good practice plus net zero policies and a
50 moderate increase in regionally differentiated carbon pricing. Fekete et al. (2021) show that global
Do Not Cite, Quote or Distribute 3-80 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 replication of sector progress in five major economies would reduce GHG emissions in 2030 by about
2 20% compared to a current policy scenario. These findings were found in good agreement with a
3 recent model comparison study based on results from 9 integrated assessment models (van Soest et al.
4 2021a). Based on these three studies, implementing accelerated action in terms of a global roll out of
5 regulatory and moderate pricing policies is assessed to lead to global GHG emissions of 47 (38-51)
6 GtCO2-eq in 2030 (median and 5th to 95th percentile based on 10 distinct modelled pathways). This
7 closes the implementation gap for the NDCs, and in addition falls below the emissions range implied
8 by implementing unconditional and conditional elements of NDCs by 3-9 GtCO2-eq. However, it
9 does not close the emissions gap to immediate action pathways likely limiting warming to 2°C, and,
10 based on our assessment in Section 3.5.2, emission levels above 40 GtCO2-eq in 2030 still have a very
11 low prospect for keeping limiting warming to 1.5°C with no or limited overshoot in reach.
12
13 Figure 3.31 shows the intermediate position of accelerated action pathways derived by van Soest et al.
14 (2021a) between pathways that follow the NDCs until 2030 and immediate action pathways likely
15 limiting warming to 2°C. Accelerated action is able to reduce the abrupt shifts in emissions, fossil fuel
16 use and low carbon power generation in 2030 and also limits peak warming more effectively than
17 NDC pathways. But primarily due to the moderate carbon price assumptions (Fig. 3.31b), the
18 reductions in emissions and particular fossil fuel use are markedly smaller than what would be
19 obtained in the case of immediate action. The assessment shows that accelerated action until 2030 can
20 have significant benefits in terms of reducing the mitigation challenges from following the NDCs
21 until 2030. But putting a significant value on GHG emissions reductions globally remains a key
22 element of moving onto 1.5-2°C pathways. The vast majority of pathways that limit warming to 2°C
23 or below, independently of their differences in near term emission developments, converge to a global
24 mitigation regime putting such a significant value on GHG emission reductions in all regions and
25 sectors.
Do Not Cite, Quote or Distribute 3-81 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1
2
3 Figure 3.31: Comparison of (i) pathways with immediate action likely limiting warming to 2°C
4 (Immediate, light blue), (ii) pathways following the NDCs until 2030 and aiming to likely stay below 2°C
5 thereafter (NDC; orange) and (iii) pathways accelerating near term action until 2030 beyond NDC
6 ambition levels and aiming to likely stay below 2°C thereafter (Accelerated) for selected indicators as
7 listed in the panel titles, based on pathways from van Soest et al. (2021a). Low carbon electricity
8 comprises renewable and nuclear power. Indicator ranges are shown as boxplots (full range, interquartile
9 range, and median) for the years 2030, 2050 and 2100 (absolute values) and for the periods 2020-2030,
10 2030-2050 (change indicators). Ranges are based on nine models participating in (van Soest et al. 2021a)
11 with only 7 models reporting emissions and climate results and 8 models reporting carbon prices. The
12 purple dot denotes the illustrative mitigation pathway GS that was part of the study by van Soest et al.
13
14
15 3.6 Economics of long-term mitigation and development pathways,
16 including mitigation costs and benefits
17 A complete appraisal of economic effects and welfare effects at different temperature levels would
18 include the macroeconomic impacts of investments in low-carbon solutions and structural change
19 away from emitting activities, co-benefits and adverse side effects of mitigation, (avoided) climate
Do Not Cite, Quote or Distribute 3-82 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 damages, as well as (reduced) adaptation costs, with high temporal, spatial and social heterogeneity
2 using a harmonized framework. If no such complete appraisal in a harmonized framework exists, key
3 elements are emerging from the literature, and assessed in the following subsections: on aggregated
4 economy-wide global mitigation costs (section 3.6.1), on the economic benefits of avoiding climate
5 impacts (section 3.6.2), on economic benefits and costs associated with mitigation co-benefits and co-
6 harms (section 3.6.3) and on the distribution of economic implications between economic sectors and
7 actors (section 3.6.4).
8
9 3.6.1 Economy-wide implications of mitigation
10 3.6.1.1 Global economic effects of mitigation and carbon values in mitigation pathways
11
12 START BOX HERE
13
14 Box 3.5 Concepts and modelling frameworks used for quantifying macro-economic effects of mitigation
15 Most studies that have developed mitigation pathways have used a Cost-effectiveness analysis (CEA)
16 framework, which aim at comparing the costs of different mitigation strategies designed to meet a
17 given climate change mitigation goal (e.g., an emission-reduction target or a temperature stabilization
18 target) but does not represent economic impacts from climate change itself, nor the associated
19 economic benefits of avoided impacts. Other studies use modelling frameworks that represent the
20 feedback of damages from climate change on the economy in a Cost-benefit analysis (CBA)
21 approach, which balances mitigation costs and benefits. This second type of studies is represented in
22 section 3.6.2.
23 The marginal abatement cost of carbon, also called carbon price, is determined by the mitigation
24 target under consideration: it describes the cost of reducing the last unit of emissions to reach the
25 target at a given point in time. Total macro-economic mitigation costs (or gains) aggregate the
26 economy-wide impacts of investments in low-carbon solutions and structural changes away from
27 emitting activities. The total macro-economic effects of mitigation pathways are reported in terms of
28 variations in economic output or consumption levels, measured against a Reference Scenario, also
29 called Baseline, at various points in time or discounted over a given time period. Depending on the
30 study, the Reference Scenario reflects specific assumptions about patterns of socio-economic
31 development and assumes either no climate policies or the climate policies in place or planned at the
32 time the study was carried out. When available in the AR6 scenarios database, this second type of
33 Reference Scenario, with current policies, has been chosen for computation of mitigation costs. In the
34 vast majority of studies that have produced the body of work on the cost of mitigation assessed here,
35 and in particular in all studies that have submitted global scenarios to the AR6 scenarios database
36 except (Schultes et al. 2021), the feedbacks of climate change impacts on the economic development
37 pathways are not accounted for. This omission of climate impacts leads to overly optimistic economic
38 projections in the reference scenarios, in particular in reference scenarios with no or limited mitigation
39 action where the extent of global warming is the greatest. Mitigation cost estimates computed against
40 no or limited policy reference scenarios therefore omit economic benefits brought by avoided climate
41 change impact along mitigation pathways, and should be interpreted with care (Grant et al. 2020).
42 When aggregate economic benefits from avoided climate change impacts are accounted for,
43 mitigation is a welfare-enhancing strategy (see section 3.6.2).
44 If GDP or consumption in mitigation pathways are below the reference scenario levels, they are
45 reported as losses or macro-economic costs. Such cost estimates give an indication on how economic
46 activity slows relative to the reference scenario; they do not necessarily describe, in absolute terms, a
47 reduction of economic output or consumption levels relative to previous years along the pathway.
48 Aggregate mitigation costs depend strongly on the modelling framework used and the assumptions
Do Not Cite, Quote or Distribute 3-83 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 about the reference scenario against which mitigation costs are measured, in particular whether the
2 reference scenario is, or not, on the efficiency frontier of the economy. If the economy is assumed to
3 be at the efficiency frontier in the reference scenario, mitigation inevitably leads to actual costs, at
4 least in the short run until the production frontier evolves with technical and structural change.
5 Starting from a reference scenario that is not on the efficiency frontier opens the possibility to
6 simultaneously reduce emissions and obtain macroeconomic gains, depending on the design and
7 implementation of mitigation policies. A number of factors can result in reference scenarios below the
8 efficiency frontier, for instance distorting labour taxes and/or fossil fuel subsidies, misallocation or
9 under-utilization of production factors such as involuntary unemployment, imperfect information or
10 non-rational behaviours. Although these factors are pervasive, the modelling frameworks used to
11 construct mitigation pathways are often limited in their ability to represent them (Köberle et al. 2021).
12 The absolute level of economic activity and welfare also strongly depends on the socioeconomic
13 pathway assumptions regarding inter alia evolutions in demography, productivity, education levels,
14 inequality and technical change and innovation. The GDP or consumption indicators reported in the
15 database of scenarios, and synthesized below, represent the absolute level of aggregate economic
16 activity or consumption but do not reflect welfare and well-being (Roberts et al. 2020), that notably
17 depend on human needs satisfaction, distribution within society and inequality (see section 3.6.4).
18 Chapter 1 and Annex III give further elements on the economic concepts and on the modelling
19 frameworks, including their limitations, used in this report, respectively.
20 END BOX HERE
21
22 Estimates for the marginal abatement cost of carbon in mitigation pathways vary widely, depending
23 on the modelling framework used and socioeconomic, technological and policy assumptions.
24 However, it is robust across modelling frameworks that the marginal abatement cost of carbon
25 increases for lower temperature categories, with a higher increase in the short-term than in the longer-
26 term (Figure 3.32, left panel) (high confidence). The marginal abatement cost of carbon increases non-
27 linearly with the decrease of CO2 emissions level, but the uncertainty in the range of estimates also
28 increases (Figure 3.33). Mitigation pathways with low‐energy consumption patterns exhibit lower
29 carbon values (Méjean et al. 2019; Meyer et al. 2021). In the context of the COVID-19 pandemic
30 recovery, Kikstra et al. (2021a) also show that a low energy demand recovery scenario reduces carbon
31 prices for a 1.5°C consistent pathway by 19% compared to a scenario with energy demand trends
32 restored to pre-pandemic levels.
33
34 For optimization modelling frameworks, the time profile of marginal abatement costs of carbon
35 depends on the discount rate, with lower discount rates implying higher carbon values in the short
36 term but lower values in the long term (Emmerling et al. 2019) (see also Discounting in glossary and
37 Annex III part I section 2). In that case, the discount rate also influences the shape of the emissions
38 trajectory, with low discount rates implying more emission reduction in the short-term and, for low
39 temperature categories, limiting CDR and temperature overshoot.
40
41 Pathways that correspond to NDCs in 2030 and strengthen action after 2030 imply higher marginal
42 abatement costs of carbon in the longer run than pathways with stronger immediate global mitigation
43 action (Figure 3.32, right panel) (high confidence).
44
Do Not Cite, Quote or Distribute 3-84 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1
2 Figure 3.32: Marginal abatement cost of carbon in 2030, 2050 and 2100 for mitigation pathways with
3 immediate global mitigation action (left panel), and ratio in 2050 between pathways that correspond to
4 NDC in 2030 and strengthen action after 2030 and pathways with immediate global mitigation action, for
5 C3 and C4 temperature categories (right panel).
6
7 Figure 3.33: Marginal abatement cost of carbon with respect to CO 2 emissions for mitigation pathways
8 with immediate global mitigation action, in 2030 (left panel) and 2050 (right panel).
9 Aggregate economic activity and consumption levels in mitigation pathways are primarily determined
10 by socioeconomic development pathways but are also influenced by the stringency of the mitigation
11 goal and the policy choices to reach the goal (high confidence). Mitigation pathways in temperature
12 categories C1 and C2 entail losses in global consumption with respect to their baselines – not
13 including benefits of avoided climate change impacts nor co-benefits or co-harms of mitigation action
14 – that correspond to an annualized reduction of consumption growth by 0.04 (median value)
15 (interquartile range [0.02-0.06]) percentage points over the century. For pathways in temperature
16 categories C3 and C4 this reduction in global consumption growth is 0.03 (median value)
17 (interquartile range [0.01-0.05]) percentage points over the century. In the majority of studies that
18 focus on the economic effects of mitigation without accounting for climate damages, global economic
19 growth and consumption growth is reduced compared to baseline scenarios (that omit damages from
20 climate change), but mitigation pathways do not represent an absolute decrease of economic activity
21 level (Figure 3.34, panels b and c).
22
Do Not Cite, Quote or Distribute 3-85 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 However, the possibility for increased economic activity following mitigation action, and conversely
2 the risk of large negative economic effects, are not excluded. Some studies find that mitigation
3 increases the speed of economic growth compared to baseline scenarios (Pollitt and Mercure 2018;
4 Mercure et al. 2019). These studies are based on a macroeconomic modelling framework that
5 represent baselines below the efficiency frontier, based on non-equilibrium economic theory, and
6 assume that mitigation is undertaken in such a way that green investments do not crowd-out
7 investment in other parts of the economy – and therefore offers an economic stimulus. In the context
8 of the recovery from the COVID-19 crisis, it is estimated that a green investment push, would initially
9 boost the economy while also reducing GHG emissions (IMF 2020; Pollitt et al. 2021). Conversely,
10 several studies find that only a GDP non-growth/degrowth or post-growth approach allow to reach
11 climate stabilization below 2°C (Hardt and O’Neill 2017; D’Alessandro et al. 2020; Hickel and Kallis
12 2020; Nieto et al. 2020), or to minimize the risks of reliance on high energy-GDP decoupling, large-
13 scale CDR and large-scale renewable energy deployment (Keyßer and Lenzen 2021). Similarly,
14 feedbacks of financial system risk to amplify shocks induced by mitigation policy and lead to higher
15 impact on economic activity (Stolbova et al. 2018).
16
17 Mitigation cost increases with the stringency of mitigation (Figure 3.34, panels b and c) (Hof et al.
18 2017; Vrontisi et al. 2018), but are reduced when energy demand is moderated through energy
19 efficiency and lifestyle changes (Fujimori et al. 2014; Bibas et al. 2015; Liu et al. 2018; Méjean et al.
20 2019), when sustainable transport policies are implemented (Zhang et al. 2018c), and when
21 international technology cooperation is fostered (Schultes et al. 2018; Paroussos et al. 2019).
22 Mitigation costs also depend on assumptions on availability and costs of technologies (Clarke et al.
23 2014; Bosetti et al. 2015; Dessens et al. 2016; Creutzig et al. 2018; Napp et al. 2019; Giannousakis et
24 al. 2021), on the representation of innovation dynamics in modelling frameworks (Hoekstra et al.
25 2017; Rengs et al. 2020) (see also chapter 16), as well as the representation of investment dynamics
26 and financing mechanisms (Iyer et al. 2015c; Mercure et al. 2019; Battiston et al. 2021). In particular,
27 endogenous and induced innovation reduce technology cost over time, create path-dependencies and
28 reduce the macroeconomic cost of reaching a mitigation target (see also 1.7.1.2). Mitigation costs also
29 depend on the socioeconomic assumptions (van Vuuren et al. 2020; Hof et al. 2017).
30
Do Not Cite, Quote or Distribute 3-86 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1
2
3 Figure 3.34: Panel (a): Mean annual global consumption growth rate over 2020-2100 for the mitigation
4 pathways in the AR6 scenarios database. Panel (b): Global GDP loss compared to baselines (not
5 accounting for climate change damages) in 2030, 2050 and 2100 for mitigation pathways with immediate
6 global action. Panel (c): Total discounted consumption loss (with a 3% discount rate) in mitigation
7 scenarios with respect to their corresponding baseline (not accounting for climate change damages) as a
8 function of cumulative CO2 emissions until date of net zero CO2. Panel (d): Comparison of GDP losses
9 compared to baselines (not accounting for climate change damages) in 2030, 2050 and 2100 for pairs of
10 scenarios depicting immediate action pathways and delayed action pathways.
11 Mitigation pathways with early emissions reductions represent higher mitigation costs in the short run
12 but bring long-term gains for the economy compared to delayed transition pathways (high
13 confidence). Pathways with earlier mitigation action bring higher long-term GDP than pathways
14 reaching the same end-of-century temperature with weaker early action (Figure 3.34, panel d).
15 Comparing counterfactual history scenarios, Sanderson and O’Neill (2020) also find that delayed
16 mitigation action leads to higher peak costs. Rogelj et al. (2019b) and Riahi et al. (2021) also show
17 that pathways with earlier timing of net zero CO2 lead to higher transition costs but lower long term
18 mitigation costs, due to dynamic effects arising from lock-in avoidance and learning effects. For
19 example, Riahi et al (2021) find that for a 2˚C target, the GDP losses (compared to a reference
20 scenario without impacts from climate change) in 2100 are 5-70% lower in pathways that avoid net
21 negative CO2 emissions and temperature overshoot than in pathways with overshoot. Accounting also
22 for climate change damage, van der Wijst et al. (2021a) show that avoiding net negative emissions
23 leads to a small increase in total discounted mitigation costs over 2020-2100, between 5% and 14% in
24 their medium assumptions, but does not increase mitigation costs when damages are high and when
25 using a low discount rate, and becomes economically attractive if damages are not fully reversible.
26 The modelled cost-optimal balance of mitigation action over time strongly depends on the discount
27 rate used to compute or evaluate mitigation pathways: lower discount rates favour earlier mitigation,
Do Not Cite, Quote or Distribute 3-87 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 reducing both temperature overshoot and reliance on net negative carbon emissions (Emmerling et al.
2 2019; Riahi et al. 2021). Mitigation pathways with weak early action corresponding to NDCs in 2030
3 and strengthening action after 2030 to reach end-of-century temperature targets imply limited
4 mitigation costs in 2030, compared to immediate global action pathways, but faster increase in costs
5 post-2030, with implications for intergenerational equity (Aldy et al. 2016; Liu et al. 2016; Vrontisi et
6 al. 2018). Emissions trading policies reduce global aggregate mitigation costs, in particular in the
7 context of achieving NDCs (Fujimori et al. 2015, 2016a; Edmonds et al. 2021; Böhringer et al. 2021),
8 and change the distribution of mitigation costs between regions and countries (see section 3.6.1.2).
9
10 3.6.1.2 Regional mitigation costs and effort-sharing regimes
11 The economic repercussions of mitigation policies vary across countries (Hof et al. 2017; Aldy et al.
12 2016): regional variations exist in institutions, economic and technological development, and
13 mitigation opportunities. For a globally uniform carbon price, carbon intensive and energy exporting
14 countries bear the highest economic costs because of a deeper transformation of their economies and
15 of trade losses in the fossil markets (Stern et al. 2012; Tavoni et al. 2015; Böhringer et al. 2021). This
16 finding is confirmed in
17 Figure 3.35. Since carbon intensive countries are often poorer, uniform global carbon prices raises
18 equity concerns (Tavoni et al. 2015). On the other hand, the climate economic benefits of mitigating
19 climate change will be larger in poorer countries (Cross-Working Group Box 1). This reduces policy
20 regressivity but does not eliminate it (Taconet et al. 2020; Gazzotti et al. 2021). Together with co-
21 benefits, such as health benefits of improved air quality, the economic benefits of mitigating climate
22 change are likely to outweigh mitigation costs in many regions (Li et al. 2018, 2019; Scovronick et al.
23 2021).
24
25 Regional policy costs depend on the evaluation framework (Budolfson et al. 2021), policy design,
26 including revenue recycling, and on international coordination, especially among trade partners. By
27 fostering technological change and finance, climate cooperation can generate economic benefits, both
28 in large developing economies such as China and India (Paroussos et al. 2019) and industrialized
29 countries such as Europe (Vrontisi et al. 2020). International coordination is a major driver of regional
30 policy costs. Delayed participation in global mitigation efforts raises participation costs, especially in
31 carbon intensive economies (
32 Figure 3.35, right panel). Trading systems and transfers can deliver cost savings and improve equity
33 (Rose et al. 2017a). On the other hand, measures that reduce imports of energy intensive goods such
34 as carbon border tax adjustment may imply costs outside of the policy jurisdiction and have
35 international equity repercussions, depending on how they are designed (Böhringer et al. 2012, 2017;
36 Cosbey et al. 2019) (see also 13.6.6).
37
38 An equitable global emission trading scheme would require very large international financial
39 transfers, in the order of several hundred billion USD per year (Tavoni et al. 2015; van den Berg et al.
40 2020; Bauer et al. 2020). The magnitude of transfers depends on the stringency of the climate goals
41 and on the burden sharing principle. Equitable burden sharing compliant with the Paris Agreement
42 leads to negative carbon allowances for developed countries as well as China by mid-century (van den
43 Berg et al. 2020), more stringent than cost-optimal pathways. International transfers also depend on
44 the underlying socio-economic development (Leimbach and Giannousakis 2019), as these drive the
45 mitigation costs of meeting the Paris Agreement (Rogelj et al. 2018b). By contrast, achieving equity
46 without international markets would result in a large discrepancy in regional carbon prices, up to a
47 factor (Bauer et al. 2020). The efficiency-sovereignty trade-off can be partly resolved by allowing for
48 limited differentiation of regional carbon prices: moderate financial transfers substantially reduce
49 inefficiencies by narrowing the carbon price spread (Bauer et al. 2020).
Do Not Cite, Quote or Distribute 3-88 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1
2
3 Figure 3.35: Left panel: Regional mitigation costs in the year 2050 (expressed as GDP losses between
4 mitigation scenarios and corresponding baselines, not accounting for climate change damages), under the
5 assumption of immediate global action with uniform global carbon pricing and no international transfers,
6 by climate categories for the 2°C and 1.5°C (with and without overshoot) categories. Right panel: Policy
7 costs in 2050 (as in panel a) for 1.5-2°C climate categories for scenario pairs that represent either
8 immediate global action (‘immediate’) or delayed global action (‘delayed’) with weaker action in the
9 short-term, strengthening to reach the same end-of-century temperature target.
10
11 3.6.1.3 Investments in mitigation pathways
12 Figure 3.36 and Figure 3.37 show increased investment needs in the energy sector in lower
13 temperature categories, and a major shift away from fossil generation and extraction towards
14 electricity, including for system enhancements for electricity transmission, distribution and storage,
15 and low-carbon technologies. Investment needs in the electricity sector are USD2.3 trillion 2015 yr-1
16 over 2023-2050 on average for C1 pathways, USD2.0 trillion for C2 pathways, USD1.7 trillion for
17 C3, USD1.2 trillion for C4 and USD0.9-1.1 billion for C5/C6/C7 (mean values for pathways in each
18 temperature categories). The regional pattern of power sector investments broadly mirrors the global
19 picture. However, the bulk of investment requirements are in medium- and low-income regions.
20 These results from the AR6 scenarios database corroborate the findings from McCollum et al.
21 (2018a), Zhou et al. (2019) and Bertram et al. (2021).
22
23 In the context of the COVID-19 pandemic recovery, (Kikstra et al. 2021a) show that a low energy
24 demand recovery scenario reduces energy investments required until 2030 for a 1.5°C consistent
25 pathway by 9% (corresponding to reducing total required energy investment by USD1.8 trillion)
26 compared to a scenario with energy demand trends restored to pre-pandemic levels.
27
Do Not Cite, Quote or Distribute 3-89 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1
2 Figure 3.36: Global average yearly investments from 2023-2052 for 9 electricity supply subcomponents
3 and for extraction of fossil fuels (in billion USD2015), in pathways by temperature categories. T&D:
4 transmission and distribution of electricity. Bars show the median values (number of pathways at the
5 bottom), and whiskers the interquartile ranges.
6
7 Figure 3.37: Average yearly investments from 2023-2052 for the four subcomponents of the energy
8 system representing the larger amounts (in billion USD2015), by aggregate regions, in pathways by
9 temperature categories. T&D: transmissions and distribution of electricity. Extr.: extraction of fossil
Do Not Cite, Quote or Distribute 3-90 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 fuels. Bars show the median values (number of pathways at the bottom), and whiskers the interquartile
2 ranges.
3
4 Few studies extend the scope of the investment needs quantification beyond the energy sector. Fisch-
5 Romito and Guivarch (2019) and Ó Broin and Guivarch (2017) assess investment needs for
6 transportation infrastructures and find lower investment needs in low-carbon pathways, due to a
7 reduction in transport activity and a shift towards less road construction, compared to high-carbon
8 pathways. Rozenberg and Fay (2019) estimate the funding needs to close the service gaps in water
9 and sanitation, transportation, electricity, irrigation, and flood protection in thousands of scenarios,
10 showing that infrastructure investment paths compatible with full decarbonization in the second half
11 of the century need not cost more than more-polluting alternatives. Investment needs are estimated
12 between 2 percent and 8 percent of GDP, depending on the quality and quantity of services targeted,
13 the timing of investments, construction costs, and complementary policies.
14
15 Chapter 15 also reports investment requirements in global mitigation pathways in the near-term,
16 compares them to recent investment trends, and assesses financing issues.
17
18 3.6.2 Economic benefits of avoiding climate changes impacts
19
20 START BOX HERE
21
22 Cross-Working Group Box 1: Economic benefits from avoided climate impacts along long-term
23 mitigation pathways
24 Authors: Céline Guivarch (France), Steven Rose (the United States of America), Alaa Al Khourdajie
25 (United Kingdom/Syria), Valentina Bosetti (Italy), Edward Byers (Austria/Ireland), Katherine Calvin
26 (the United States of America), Tamma Carleton (the United States of America), Delavane Diaz (the
27 United States of America), Laurent Drouet (France/Italy), Michael Grubb (United Kingdom), Tomoko
28 Hasegawa (Japan), Alexandre C. Köberle (Brazil/ United Kingdom), Elmar Kriegler (Germany),
29 David McCollum (the United States of America), Aurélie Méjean (France), Brian O’Neill (the United
30 States of America), Franziska Piontek (Germany), Julia Steinberger (United Kingdom/Switzerland),
31 Massimo Tavoni (Italy)
32
33 Mitigation reduces the extent of climate change and its impacts on ecosystems, infrastructure, and
34 livelihoods. This box summarizes elements from the WGII report on aggregate climate change
35 impacts and risks, putting them into the context of mitigation pathways. AR6 Working Group II
36 provides an assessment of current lines of evidence regarding potential climate risks with future
37 climate change, and therefore, the avoided risks from mitigating climate change. Regional and
38 sectoral climate risks to physical and social systems are assessed (WGII Chapters 2-15). Over 100 of
39 these are identified as Key Risks (KRs) and further synthesized by WGII Chapter 16 into eight
40 overarching Representative Key Risks (RKRs) relating to low-lying coastal systems; terrestrial and
41 ocean ecosystems; critical physical infrastructure, networks and services; living standards; human
42 health; food security; water security; and peace and mobility (WGII 16.5.2). The RKR assessment
43 finds that risks increase with global warming level, and also depend on socioeconomic development
44 conditions, which shape exposure and vulnerability, and adaptation opportunities and responses.
45 “Reasons For Concern”, another WGII aggregate climate impacts risk framing, are also assessed to
46 increase with climate change, with increasing risk for unique and threatened systems, extreme weather
47 events, distribution of impacts, global aggregate impacts, and large-scale singular events (WGII
48 Chapter 16). For human systems, in general, the poor and disadvantaged are found to have greater
49 exposure level and vulnerability for a given hazard. With some increase in global average warming
Do Not Cite, Quote or Distribute 3-91 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 from today expected regardless of mitigation efforts, human and natural systems will be exposed to
2 new conditions and additional adaptation will be needed (WGII Chapter 18). The range of dates for
3 when a specific warming level could be reached depends on future global emissions, with significant
4 overlap of ranges across emissions scenarios due to climate system response uncertainties (WGI
5 Tables 4.2 and 4.5). The speed at which the climate changes is relevant to adaptation timing,
6 possibilities, and net impacts.
7
8 WGII also assesses the growing literature estimating the global aggregate economic impacts of
9 climate change and the social cost of carbon dioxide and other greenhouse gases (Cross-Working Box
10 Economic: Estimating Global Economic Impacts from Climate Change and the Social Cost of
11 Carbon”, in WGII Chapter 16). The former represents aggregate estimates that inform assessment of
12 the economic benefits of mitigation. This literature is characterized by significant variation in the
13 estimates, including for today's level of global warming, due primarily to fundamental differences in
14 methods, but also differences in impacts included, representation of socioeconomic exposure,
15 consideration of adaptation, aggregation approach, and assumed persistence of damages. WGII’s
16 assessment identifies different approaches to quantification of aggregated economic impacts of
17 climate change, including: physical modelling of impact processes, such as projected mortality rates
18 from climate risks such as heat, vector- or waterborne diseases that are then monetized; structural
19 economic modelling of impacts on production, consumption, and markets for economic sectors and
20 regional economies; and statistical estimation of impacts based on observed historical responses to
21 weather and climate. WGII finds that variation in estimated global economic impacts increases with
22 warming in all methodologies, indicating higher risk in terms of economic impacts at higher
23 temperatures (high confidence). Many estimates are nonlinear with marginal economic impacts
24 increasing with temperature, although some show declining marginal economic impacts with
25 temperature, and functional forms cannot be determined for all studies. WGII’s assessment finds that
26 the lack of comparability between methodologies does not allow for identification of robust ranges of
27 global economic impact estimates (high confidence). Further, WGII identifies evaluating and
28 reconciling differences in methodologies as a research priority for facilitating use of the different lines
29 of evidence (high confidence). However, there are estimates that are higher than AR5, indicating that
30 global aggregate economic impacts could be higher than previously estimated (low confidence due to
31 the lack of comparability across methodologies and lack of robustness of estimates) (Cross-Working
32 Box Economic).
33
34 Conceptually, the difference in aggregate economic impacts from climate change between two given
35 temperature levels represents the aggregate economic benefits arising from avoided climate change
36 impacts due to mitigation action. A subset of the studies whose estimates were evaluated by WGII (5
37 of 15) are used to derive illustrative estimates of aggregate economic benefits in 2100 arising from
38 avoided climate change (Howard and Sterner 2017; Burke et al. 2018; Pretis et al. 2018; Kahn et al.
39 2019; Takakura et al. 2019). Burke et al. (2018), Pretis et al. (2018) and Kahn et al. (2019) are
40 examples of statistical estimation of historical relationships between temperature and economic
41 growth, whereas Takakura et al. (2019) is an example of structural modelling, which evaluates
42 selected impact channels (impacts on agriculture productivity, undernourishment, heat-related
43 mortality, labour productivity, cooling/heating demand, hydroelectric and thermal power generation
44 capacity and fluvial flooding) with a general equilibrium model. Howard and Sterner (2017) and
45 Rose et al. (2017b) estimate damage functions that can be used to compute the economic benefits of
46 mitigation from avoiding a given temperature level for a lower one. Howard and Sterner (2017)
47 estimate a damage function from a meta-analysis of aggregate economic impact studies, while Rose et
48 al. (2017b) derive global functions by temperature and socioeconomic drivers from stylized aggregate
49 Cost-Benefit-Analysis integrated assessment models using diagnostic experiments. Cross-Working
50 Group Box 1 Figure 1 summarizes the global aggregate economic benefits in 2100 of avoided climate
Do Not Cite, Quote or Distribute 3-92 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 change impacts from individual studies corresponding to shifting from a higher temperature category
2 (above 3°C, below 3°C or below 2.5°C) to below 2°C, as well as from below 2°C to below 1.5°C.
3 Benefits are positive and increase with the temperature gap for any given study, and this result is
4 robust across socioeconomic scenarios. The Figure provides evidence of a wide range of
5 quantifications, and illustrates the important differences associated with methods. Panel a puts the
6 studies used to calculate aggregate economic benefits arising from avoided impacts into the context of
7 the broader set of studies assessed in WGII (WGII Cross-Working Group Box Economic, 16.6.2).
8 However, economic benefits in 2100 arising from avoided impacts cannot be directly computed from
9 damage estimates across this broader set of studies, due to inconsistencies - different socioeconomic
10 assumptions, scenario designs, and counterfactual reference scenarios across studies. Furthermore,
11 these types of estimates cannot be readily compared to mitigation cost estimates. The comparison
12 would require a framework that ensures consistency in assumptions and dynamics and allows for
13 consideration of benefits and costs along the entire pathway.
14
15 Cross-Working Group Box 1 Figure 1: Global aggregate economic benefits of mitigation from avoided
16 climate change impacts in 2100 corresponding to shifting from a higher temperature category (4°C
17 (3.75°C-4.25°C), 3°C(2.75°C-3.25°C) or above 2°C (2°C-2.5°C)) to below 2°C (1.5°C-2°C), as well as from
18 below 2°C to below 1.5°C (1°C-1.5°C), from the five studies discussed in the text. Panel a is adapted from
19 WGII CWGB ECONOMIC Figure 1, showing global aggregate economic impact estimates (% global
20 GDP loss relative to GDP without additional climate change) by temperature change level. All estimates
21 are shown in grey. Estimates used for the computation of estimated benefits in 2100 in Panel b are
22 coloured for the selected studies, which provide results for different temperature change levels. See the
23 WGII Chapter 16 box for discussion and assessment of the estimates in Panel a and the differences in
24 methodologies. For B18 and T19, median estimates in the cluster are considered. Shape distinguishes the
25 baseline scenarios. Temperature ranges are highlighted. HS17 estimates are based on their preferred
26 model - 50th percentile of non-catastrophic damage. Panel b shows the implied aggregate economic
27 benefits in 2100 of a lower temperature increase. Economic benefits for point estimates are computed as a
28 difference, while economic benefits from the curve HS17 are computed as ranges from the segment
29 differences.
30
31 Aggregate benefits from avoided impacts expressed in GDP terms, as in Figure 1, do not encompass
32 all avoided climate risks, adaptation possibilities, and does not represent their influence on well-being
33 and welfare (WGII Cross-Working Group Box Economic). Methodological challenges for economic
34 impact estimates include representing uncertainty and variability, capturing interactions and spill
35 overs, considering distributional effects, representing micro and macro adaptation processes,
36 specifying non-gradual damages and non-linearities, and improving understanding of potential long-
37 run growth effects. In addition, the economic benefits aggregated at the global scale provide limited
38 insights into regional heterogeneity. Global economic impact studies with regional estimates find
39 large differences across regions in absolute and percentage terms, with developing and transitional
40 economies typically more vulnerable. Furthermore, (avoided) impacts for poorer households and
41 poorer countries can represent a smaller share in aggregate quantifications expressed in GDP terms or
42 monetary terms, compared to their influence on well-being and welfare (Hallegatte et al. 2020;
43 Markhvida et al. 2020). Finally, as noted by WGII, other lines of evidence regarding climate risks,
44 beyond monetary estimates, should be considered in decision-making, including key risks and
45 Reasons for Concern.
46
47 END CCB HERE
48
49 Cost-benefit analyses (CBA) aim to balance all costs and benefits in a unified framework (Nordhaus,
50 2008). Estimates of economic benefits from avoided climate change impacts depend on the types of
51 damages accounted for, the assumed exposure and vulnerability to these damages as well as the
52 adaptation capacity, which in turn are based on the development pathway assumed (Cross-Working
Do Not Cite, Quote or Distribute 3-93 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 Group Box 1). CBA integrated assessment models raised critics, in particular for omitting elements of
2 dynamic realism, such as inertia, induced innovation and path dependence, in their representation of
3 mitigation (Grubb et al. 2021), and for underestimating damages from climate change, missing non-
4 monetary damages, the uncertain and heterogeneous nature of damages and the risk of catastrophic
5 damages (Stern 2013, 2016; Stern and Stiglitz 2021; Diaz and Moore 2017; Pindyck 2017; NASEM
6 2017; Stoerk et al. 2018). Emerging literature has started to address those gaps, and integrated into
7 cost-benefit frameworks the account of heterogeneity of climate damage and inequality (Dennig et al.
8 2015; Budolfson et al. 2017; Fleurbaey et al. 2019; Kornek et al. 2021), damages with higher
9 persistence, including damages on capital and growth (Dietz and Stern 2015; Moore and Diaz 2015;
10 Moyer et al. 2014; Guivarch and Pottier 2018; Piontek et al. 2019; Ricke et al. 2018), risks of tipping
11 points (Cai et al. 2015, 2016; Lontzek et al. 2015; Lemoine and Traeger 2016; van der Ploeg and de
12 Zeeuw 2018; Cai and Lontzek 2019; Nordhaus 2019; Yumashev et al. 2019; Taconet et al. 2021) and
13 damages to natural capital and non-market goods (Tol 1994; Sterner and Persson 2008; Bastien-
14 Olvera and Moore 2020; Drupp and Hänsel 2021).
15
16 Each of these factors, when accounted for in a CBA framework, tends to increase the welfare benefit
17 of mitigation, thus leading to stabilization at lower temperature in optimal mitigation pathways. The
18 limitations in CBA modelling frameworks remain significant, their ability to represent all damages
19 incomplete, and the uncertainty in estimates remains large. However, emerging evidence suggests
20 that, even without accounting for co-benefits of mitigation on other sustainable development
21 dimensions (see section 3.6.3 for elements on co-benefits), global benefits of pathways likely to limit
22 warming to 2°C outweigh global mitigation costs over the 21st century: depending on the study, the
23 reason for this result lies in assumptions of economic damages from climate change in the higher end
24 of available estimates (Moore and Diaz 2015; Ueckerdt et al. 2019; Brown and Saunders 2020;
25 Glanemann et al. 2020), in the introduction of risks of tipping-points (Cai and Lontzek 2019), in the
26 consideration of damages to natural capital and non-market goods (Bastien-Olvera and Moore 2020)
27 or in the combination of updated representations of carbon cycle and climate modules, updated
28 damage estimates and/or updated representations of economic and mitigation dynamics (Dietz and
29 Stern 2015; Hänsel et al. 2020; Wei et al. 2020; van der Wijst et al. 2021b). In the above studies that
30 perform a sensitivity analysis, this result is found to be robust to a wide range of assumptions on
31 social preferences (in particular on inequality aversion and pure rate of time preference) and holds
32 except if assumptions of economic damages from climate change are in the lower end of available
33 estimates and the pure rate of time preference is in the higher range of values usually considered
34 (typically above 1.5%). However, although such pathways bring net benefits over time (in terms of
35 aggregate discounted present value), they involve distributional consequences and transition costs
36 (Brown and Saunders 2020; Brown et al. 2020) (see also sections 3.6.1.2 and 3.6.4).
37
38 The standard discounted utilitarian framework dominates CBA, thus often limiting the analysis to the
39 question of discounting. CBA can be expanded to accommodate a wider variety of ethical values to
40 assess mitigation pathways (Fleurbaey et al. 2019). The role of ethical values with regard to inequality
41 and the situation of the worse-off (Adler et al. 2017), risk (van den Bergh and Botzen 2014; Drouet et
42 al. 2015), and population size (Scovronick et al. 2017; Méjean et al. 2020) has been explored. In most
43 of these studies, the optimal climate policy is found to be more stringent than the one obtained using a
44 standard discounted utilitarian criterion.
45
46 Comparing economic costs and benefits of mitigation raises a number of methodological and
47 fundamental difficulties. Monetizing the full range of climate change impacts is extremely hard, if not
48 impossible (WGII chapter 16), as is aggregating costs and benefits over time and across individuals
49 when values are heterogeneous (AR5, WGIII chapter 3; this assessment, Chapter 1). Other approaches
50 should thus be considered in supplement for decision making (Chapter 1, section 1.7), in particular
Do Not Cite, Quote or Distribute 3-94 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 cost-effectiveness approaches that analyse how to achieve a defined mitigation objective at least cost
2 or while also reaching other societal goals (Koomey 2013; Kaufman et al. 2020; Köberle et al. 2021;
3 Stern and Stiglitz 2021). In cost effectiveness studies too, incorporating benefits from avoided climate
4 damages influences the results and leads to more stringent mitigation in the short-term (Drouet et al.
5 2021; Schultes et al. 2021).
6
7 3.6.3 Aggregate economic implication of mitigation co-benefits and trade-offs
8 Mitigation actions have co-benefits and trade-offs with other sustainable development dimensions
9 (section 3.7), beyond climate change, which imply welfare effects and economic effects, as well as
10 other implications beyond the economic dimension. The majority of quantifications of mitigation
11 costs and benefits synthesized in sections 3.6.1 and 3.6.2 do not account for these economic benefits
12 and costs associated with co-benefits and trade-offs along mitigation pathways.
13
14 Systematic reviews of the literature on co-benefits and trade-offs from mitigation actions have shown
15 that only a small portion of articles provide economic quantifications (Deng et al. 2017; Karlsson et
16 al. 2020). Most economic quantifications use monetary valuation approaches. Improved air quality,
17 and associated health effects, are the co-benefit category dominating the literature (Markandya et al.
18 2018; Vandyck et al. 2018; Scovronick et al. 2019; Howard et al. 2020; Karlsson et al. 2020b; Rauner
19 et al. 2020a,b), but some studies cover other categories, including health effects from diet change
20 (Springmann et al. 2016b) and biodiversity impacts (Rauner et al. 2020a). Regarding health effects
21 from air quality improvement and from diet change, co-benefits are shown to be of the same order of
22 magnitude as mitigation costs (Thompson et al. 2014; Springmann et al. 2016a,b; Markandya et al.
23 2018; Scovronick et al. 2019b; Howard et al. 2020; Rauner et al. 2020a,b; Liu et al. 2021; Yang et al.
24 2021). Co-benefits from improved air quality are concentrated sooner in time than economic benefits
25 from avoided climate change impacts (Karlsson et al. 2020), such that when accounting both for
26 positive health impacts from reduced air pollution and for negative climate effect of reduced cooling
27 aerosols, optimal greenhouse gas mitigation pathways exhibit immediate and continual net economic
28 benefits (Scovronick et al. 2019a). However, WGI chapter 6 (section 6.7.3) shows a delay in air
29 pollution reduction benefits when they come from climate change mitigation policies compared with
30 air pollution reduction policies.
31
32 Achieving co-benefits is not automatic but results from coordinated policies and implementation
33 strategies (Clarke et al. 2014; McCollum et al. 2018a). Similarly, avoiding trade-offs requires targeted
34 policies (van Vuuren et al. 2015; Bertram et al. 2018). There is limited evidence of such pathways, but
35 the evidence shows that pathways mitigation pathways designed to reach multiple sustainable
36 development goals instead of focusing exclusively on emissions reductions, result in limited
37 additional costs compared to the increased benefits (Cameron et al. 2016; McCollum et al. 2018b;
38 Fujimori et al. 2020a; Sognnaes et al. 2021).
39
40 3.6.4 Structural change, employment and distributional issues along mitigation pathways
41
42 Beyond aggregate effects at the economy wide level, mitigation pathways have heterogeneous
43 economic implications for different sectors and different actors. Climate-related factors are only one
44 driver of the future structure of the economy, of the future of employment, and of future inequality
45 trends, as overarching trends in demographics, technological change (innovation, automation, etc.),
46 education and institutions will be prominent drivers. For instance, Rao et al. (2019b) and Benveniste
47 et al. (2021) have shown that income inequality projections for the 21st century vary significantly,
48 depending on socioeconomic assumptions related to demography, education levels, social public
49 spending and migrations. However, the sections below focus on climate-related factors, both climate
Do Not Cite, Quote or Distribute 3-95 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 mitigation actions themselves and the climate change impacts avoided along mitigation pathways,
2 effects on structural change, including employment, and distributional effects.
3
4 3.6.4.1 Economic structural change and employment in long-term mitigation pathways
5 Mitigation pathways entail transformation of the energy sector, with structural change away from
6 fossil energy and towards low-carbon energy (section 3.3), as well as broader economic structural
7 change, including industrial restructuring and reductions in carbon-intensive activities in parallel to
8 extensions in low-carbon activities.
9
10 Mitigation affects work through multiple channels, which impacts geographies, sectors and skill
11 categories differently (Fankhaeser et al. 2008; Bowen et al. 2018; Malerba and Wiebe 2021).
12 Aggregate employment impacts of mitigation pathways mainly depend on the aggregate
13 macroeconomic effect of mitigation (see 3.6.1, 3.6.2) and of mitigation policy design and
14 implementation (Freire-González 2018) (section 4.2.6.3). Most studies that quantify overall
15 employment implications of mitigation policies are conducted at the national or regional scales
16 (section 4.2.6.3), or sectoral scales (e.g., see chapter 6 for energy sector jobs). The evidence is limited
17 at the multi-national or global scale, but studies generally find small differences in aggregate
18 employment in mitigation pathways compared to baselines: the sign of the difference depends on the
19 assumptions and modelling frameworks used and the policy design tested, with some studies or policy
20 design cases leading to small increases in employment (Chateau and Saint-Martin 2013; Pollitt et al.
21 2015; Barker et al. 2016; Garcia-Casals et al. 2019; Fujimori et al. 2020a; Vrontisi et al. 2020;
22 Malerba and Wiebe 2021) and other studies or policy design cases leading to small decreases
23 (Chateau and Saint-Martin 2013; Vandyck et al. 2016). The small variations in aggregate employment
24 hide substantial reallocation of jobs across sectors, with jobs creation in some sectors and jobs
25 destruction in others. Mitigation action through thermal renovation of buildings, installation and
26 maintenance of low-carbon generation, the build-out of public transit lead to jobs creation, while jobs
27 are lost in fossil fuel extraction, energy supply and energy intensive sectors in mitigation pathways
28 (von Stechow et al. 2015, 2016; Barker et al. 2016; Fuso Nerini et al. 2018; Perrier and Quirion 2018;
29 Pollitt and Mercure 2018; Dominish et al. 2019; Garcia-Casals et al. 2019). In the energy sector, jobs
30 losses in the fossil fuel sector are found to be compensated by gains in wind and solar jobs, leading to
31 a net increase in energy sector jobs in 2050 in a mitigation pathway compatible with stabilization of
32 the temperature increase below 2°C (Pai et al. 2021). Employment effects also differ by geographies,
33 with energy-importing regions benefiting from net job creations but energy-exporting regions
34 experiencing very small gains or suffering from net job destruction (Barker et al. 2016; Pollitt and
35 Mercure 2018; Garcia-Casals et al. 2019; Malerba and Wiebe 2021). Coal phase-out raises acute
36 issues of just transition for the coal-dependent countries (Spencer et al. 2018; Jakob et al. 2020)
37 (section 4.5 and Box 6.2).
38
39 Mitigation action also affects employment through avoided climate change impacts. Mitigation
40 reduces the risks to human health and associated impacts on labour and helps protect workers from
41 the occupational health and safety hazards imposed by climate change (Kjellstrom et al. 2016, 2018,
42 2019 ; Levi et al. 2018; Day et al. 2019) (see WGII chapter 16).
43
44 3.6.4.2 Distributional implications of long-term mitigation pathways
45 Mitigation policies can have important distributive effects between and within countries, either
46 reducing or increasing economic inequality and poverty, depending on policy instruments design and
47 implementation (see section 3.6.1.2 for an assessment of the distribution of mitigation costs across
48 regions in mitigation pathways, section 3.7, Box 3.6 and chapter 4 section 4.2.2.6 for an assessment of
49 the fairness and ambition of NDCs, and section 4.5 for an assessment of national mitigation pathways
50 along the criteria of equity, including just transition, as well as chapter 17, section 17.4.5 for equity in
Do Not Cite, Quote or Distribute 3-96 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 a just transition). For instance, emissions taxation has important distributive effects, both between and
2 within income groups (Klenert et al. 2018; Cronin et al. 2018b; Pizer and Sexton 2019; Douenne
3 2020; Steckel et al. 2021). These effects are more significant in some sectors, such as transport, and
4 depend on country-specific consumption structures (Dorband et al. 2019; Fullerton and Muehlegger
5 2019; Ohlendorf et al. 2021). However, revenues from emissions taxation can be used to lessen their
6 regressive distributional impacts or even turn the policy into a progressive policy reducing inequality
7 and/or leading to gains for lower income households (Cameron et al. 2016; Jakob and Steckel 2016;
8 Fremstad and Paul 2019; Fujimori et al. 2020b; Böhringer et al. 2021; Steckel et al. 2021; Soergel et
9 al. 2021b; Budolfson et al. 2021). Mitigation policies may affect the poorest through effects on energy
10 and food prices (Hasegawa et al. 2015; Fujimori et al. 2019). Markkanen and Anger-Kraavi (2019)
11 and Lamb et al. (2020) synthesize evidence from the existing literature on social co-impacts of
12 climate change mitigation policy and their implications for inequality. They show that most policies
13 can compound or lessen inequalities depending on contextual factors, policy design and policy
14 implementation, but that negative inequality impacts of climate policies can be mitigated (and
15 possibly even prevented), when distributive and procedural justice are taken into consideration in all
16 stages of policy making, including policy planning, development and implementation, and when
17 focusing on the carbon intensity of lifestyles, sufficiency and equity, wellbeing and decent living
18 standards for all (see also 13.6).
19
20 Mitigation pathways also affect economic inequalities between and within countries, and poverty,
21 through the reduction of climate change impacts that fall more heavily on low-income countries,
22 communities and households and exacerbate poverty (WGII chapters 8 and 16). Higher levels of
23 warming are projected to generate higher inequality between countries as well as within them (WGII
24 chapter 16). Through avoiding impacts, mitigation thus reduces economic inequalities and poverty
25 (high confidence).
26
27 A few studies consider both mitigation policies distributional impacts and avoided climate change
28 impacts on inequalities along mitigation pathways. Rezai et al. (2018) find that unmitigated climate
29 change impacts increase inequality, whereas mitigation has the potential to reverse this effect.
30 Considering uncertainty in socioeconomic assumptions, emission pathways, mitigation costs,
31 temperature response, and climate damage, Taconet et al. (2020) show that the uncertainties
32 associated with socioeconomic assumptions and damage estimates are the main drivers of future
33 inequalities between countries and that in most cases mitigation policies reduce future inequalities
34 between countries. Gazzotti et al. (2021) show that inequality persists in 2°C consistent pathways due
35 to regressivity of residual climate damages. However, the evidence on mitigation pathways
36 implications for global inequality and poverty remains limited, and the modelling frameworks used
37 have limited ability to fully represent the different dimensions of inequality and poverty and all the
38 mechanisms by which mitigation affects inequality and poverty (Rao et al. 2017a; Emmerling and
39 Tavoni 2021; Jafino et al. 2021).
40
41
42 3.7 Sustainable development, mitigation and avoided impacts
43 3.7.1 Synthesis findings on mitigation and sustainable development
44 Rapid and effective climate mitigation is a necessary part of sustainable development (high
45 confidence) (see Cross-Chapter Box 5 in Chapter 4), but the latter can only be realized if climate
46 mitigation becomes integrated with sustainable development policies (high confidence). Targeted
47 policy areas must include healthy nutrition, sustainable consumption and production, inequality and
48 poverty alleviation, air quality and international collaboration (high confidence). Lower energy
Do Not Cite, Quote or Distribute 3-97 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 demand enables synergies between mitigation and sustainability, with lower reliance on CDR (high
2 confidence).
3
4 This section covers the long-term interconnection of SD and mitigation, taking forward the holistic
5 vision of SD described in the SDGs (Brandi 2015; Leal Filho et al. 2018). Recent studies have
6 explored the aggregated impact of mitigation for multiple sustainable development dimensions
7 (Hasegawa et al. 2014; Bertram et al. 2018; Grubler et al. 2018; McCollum et al. 2018b; Fuso Nerini
8 et al. 2018; van Vuuren et al. 2019; Soergel et al. 2021a). For instance, Figure 3.38 shows selected
9 mitigation co-benefits and trade-offs based on a subset of models and scenarios, since so far many
10 IAMs do not have a comprehensive coverage of sustainable development goals (Rao et al. 2017a; van
11 Soest et al. 2019). Figure 3.38 shows that mitigation likely leads to increased forest cover (SDG 15)
12 and reduced mortality from ambient PM2.5 pollution (SDG 3) compared to reference scenarios.
13 However, mitigation policies can also cause higher food prices and an increased population at risk of
14 hunger (SDG 2) and relying on solid fuels (SDG 3 and SDG 7) as side effects. These trade-offs can be
15 compensated through targeted support measures and/or additional SD policies (Cameron et al. 2016;
16 Bertram et al. 2018; Fujimori et al. 2019; Soergel et al. 2021a).
17
18
19 Figure 3.38: Effect of climate change mitigation on different dimensions of sustainable development:
20 shown are mitigation scenarios compatible with 1.5 oC target (blue) and reference scenarios (red). Blue
21 boxplots contain scenarios that include narrow mitigation policies from different studies (see below). This
22 is compared to a sustainable development scenario (SP, Soergel et al. (2021a), cyan diamonds) integrating
23 mitigation and SD policies (e.g., zero hunger in 2050 by assumption). Scenario sources for boxplots: single
24 scenarios from i) Fujimori et al. (2020a); ii) Soergel et al. (2021a); multi-model scenario set from CD-
25 LINKS (McCollum et al. 2018b; Roelfsema et al. 2020; Fujimori et al. 2019). For associated methods, see
26 also Cameron et al. (2016), Rafaj et al. (2021). The reference scenario for (Fujimori et al. 2020a) is no-
27 policy baseline; for all other studies, it includes current climate policies. In the “Food prices” and “Risk
28 of hunger” panels, scenarios from CD-LINKS include a price cap of 200 USD/tCO2eq for land-use
29 emissions (Fujimori et al. 2019). For the other indicators, CD-LINKS scenarios without price cap
30 Roelfsema et al. (2020) are used due to SDG indicator availability. In the “Premature deaths” panel, a
31 well-below 2oC scenario from Fujimori et al. (2020a) is used in place of a 1.5oC scenario due to data
32 availability, and all scenarios are indexed to their 2015 values due to a spread in reported levels between
33 models. SDG icons were created by the United Nations.
Do Not Cite, Quote or Distribute 3-98 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1
2 The synthesis of the interplay between climate mitigation and sustainable development is shown in
3 Figure 3.39. The left panel shows the reduction in population affected by climate impacts at 1.5°C
4 compared to 3°C according to sustainability domains (Byers et al. 2018). Reducing warming reduces
5 the population impacted by all impact categories shown (high confidence). The left panel does not
6 take into account any side effects of mitigation efforts or policies to reduce warming: only reductions
7 in climate impacts. This underscores that mitigation is an integral basis for comprehensive SD (Watts
8 et al. 2015).
9
10 The middle and right panels of Figure 3.39 show the effects of 1.5°C mitigation policies compared to
11 current national policies: narrow mitigation policies (averaged over several models, middle panel),
12 and policies integrating sustainability considerations (right panel of Figure 3.39, based on the
13 Illustrative Mitigation Pathway “Shifting Pathways” (SP) (Soergel et al. 2021a)). Policies integrating
14 sustainability and mitigation (right panel) have far fewer trade-offs (red bars) and more co-benefits
15 (green bars) than narrow mitigation policies (middle panel). Note that neither middle nor right panels
16 include climate impacts.
17
18 Areas of co-benefits include human health, ambient air pollution and other specific kinds of pollution,
19 while areas of trade-off include food access, habitat loss and mineral resources (medium confidence).
20 For example, action consistent with 1.5°C in the absence of energy demand reduction measures
21 require large quantities of CDR, which, depending on the type used, are likely to negatively impact
22 both food availability and areas for biodiversity (Fujimori et al. 2018; Ohashi et al. 2019; Roelfsema
23 et al. 2020).
24
25 Mitigation to 1.5°C reduces climate impacts on sustainability (left). Policies integrating sustainability
26 and mitigation (right) have far fewer trade-offs than narrow mitigation policies (middle).
27
28
Do Not Cite, Quote or Distribute 3-99 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 Figure 3.39: Sustainable development effects of mitigation to 1.5°C. Left: benefits of mitigation from
2 avoided impacts. Middle: sustainability co-benefits and trade-offs of narrow mitigation policies (averaged
3 over multiple models). Right: sustainability co-benefits and trade-offs of mitigation policies integrating
4 sustainable development goals. Scale: 0% means no change compared to 3oC (left) or current policies
5 (middle and right). Green values correspond to proportional improvements, red values to proportional
6 worsening. Note: only the left panel considers climate impacts on sustainable development; the middle
7 and right panels do not. “Res’ C&P” stands for Responsible Consumption and Production (SDG 12).
8 Data are from Byers et al. (2018) (left), SP/Soergel et al. (2021a) (right). Methods used in middle panel:
9 for biodiversity, Ohashi et al. (2019), for ecotoxicity and eutrophication, Arvesen et al. (2018) and Pehl et
10 al. (2017), for energy access, Cameron et al. (2016). “Energy services” on the right is a measure of useful
11 energy in buildings and transport. “Food prices” and “Risk of hunger” in the middle panel are the same
12 as in Figure 3.38.
13 3.7.1.1 Policies combining mitigation and sustainable development
14
15 These findings indicate that holistic policymaking integrating sustainability objectives alongside
16 mitigation will be important in attaining sustainable development goals (van Vuuren et al. 2015, 2018;
17 Bertram et al. 2018; Fujimori et al. 2018; Hasegawa et al. 2018; Liu et al. 2020a; Honegger et al.
18 2021; Soergel et al. 2021a). Mitigation policies which target direct sector-level regulation, early
19 mitigation action, and lifestyle changes have beneficial sustainable development outcomes across air
20 pollution, food, energy and water (Bertram et al. 2018).
21
22 These policies include ones around stringent air quality (Kinney 2018; Rafaj et al. 2018; Soergel et al.
23 2021a); efficient and safe demand-side technologies, especially cook stoves (Cameron et al. 2016);
24 lifestyle changes (Bertram et al. 2018; Grubler et al. 2018; Soergel et al. 2021a); industrial and
25 sectoral policy (Bertram et al. 2018); agricultural and food policies (including food waste) (van
26 Vuuren et al. 2019; Soergel et al. 2021a); international cooperation (Soergel et al. 2021a); as well as
27 economic policies described in section 3.6. Recent research shows that mitigation is compatible with
28 reductions in inequality and poverty (see Box 3.6 on poverty and inequality).
29
30 Lower demand – e.g., for energy and land-intensive consumption such as meat – represents a
31 synergistic strategy for achieving ambitious climate mitigation without compromising sustainable
32 development goals (Grubler et al. 2018; van Vuuren et al. 2018; Bertram et al. 2018; Kikstra et al.
33 2021b; Soergel et al. 2021a) (high confidence). This is especially true for reliance on BECCS (Hickel
34 et al. 2021; Keyßer and Lenzen 2021). Options that reduce agricultural demand (e.g., dietary change,
35 reduced food waste) can have co-benefits for adaptation through reductions in demand for land and
36 water (IPCC 2019a; Grubler et al. 2018; Bertram et al. 2018; Soergel et al. 2021a).
37
38 While the impacts of climate change on agricultural output are expected to increase the population at
39 risk of hunger, there is evidence suggesting population growth will be the dominant driver of hunger
40 and undernourishment in Africa in 2050 (Hall et al. 2017). Meeting SDG5 relating to gender equality
41 and reproductive rights could substantially lower population growth, leading to a global population
42 lower than the 95% prediction range of the UN projections (Abel et al. 2016). Meeting SDG5 (gender
43 equality, including via voluntary family planning (O’Sullivan 2018)) could thus minimise the risks to
44 SDG2 (hunger) that are posed by meeting SDG13 (climate action).
45
46 START BOX HERE
47 Box 3.6 Poverty and Inequality
48 There is high confidence (medium evidence, high agreement) that the eradication of extreme poverty
49 and universal access to energy can be achieved without resulting in significant greenhouse gas
50 emissions (Tait and Winkler 2012; Pachauri 2014; Chakravarty and Tavoni 2013; Rao 2014; Pachauri
51 et al. 2013; Hubacek et al. 2017b; Poblete-Cazenave et al. 2021). There is also high agreement in the
Do Not Cite, Quote or Distribute 3-100 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 literature that a focus on wellbeing and decent living standards for all can reduce disparities in access
2 to basic needs for services concurrently with climate mitigation (Chapter 5, 5.2). Mitigation pathways
3 in which national redistribution of carbon pricing revenues is combined with international climate
4 finance, achieve poverty reduction globally (Fujimori et al. 2020b; Soergel et al. 2021b). Carbon
5 pricing revenues in mitigation pathways consistent with limiting temperature increase to 2°C could
6 also contribute to finance investment needs for basic infrastructure (Jakob et al. 2016) and SDGs
7 achievement (Franks et al. 2018).
8
9 Several studies conclude that reaching higher income levels globally, beyond exiting extreme poverty,
10 and achieving more qualitative social objectives and well-being, are associated with higher emissions
11 (Ribas et al. 2017, 2019; Scherer et al. 2018; Fischetti 2018; Hubacek et al. 2017b). Studies give
12 divergent results on the effect of economic inequality reduction on emissions, with either an increase
13 or a decrease in emissions (Berthe and Elie 2015; Lamb and Rao 2015; Grunewald et al. 2017;
14 Hubacek et al. 2017a,b; Jorgenson et al. 2017; Knight et al. 2017; Mader 2018; Rao and Min 2018;
15 Liu et al. 2019; Sager 2019; Baležentis et al. 2020; Liobikienė 2020; Liobikienė and Rimkuvienė
16 2020; Liu et al. 2020b; Millward-Hopkins and Oswald 2021). However, the absolute effect of
17 economic inequality reduction on emissions remains moderate, under the assumptions tested. For
18 instance, (Sager 2019) finds that a full redistribution of income leading to equality among US
19 households in a counterfactual scenario for 2009 would raise emissions by 2.3%; and (Rao and Min
20 2018) limit to 8% the maximum plausible increase in emissions that would accompany the reduction
21 of the global Gini coefficient from its current level of 0.55 to a level of 0.3 by 2050. Similarly,
22 reduced income inequality would lead to a global energy demand increase of 7% (Oswald et al. 2021).
23 Reconciling mitigation and inequality reduction objectives requires policies that take into account
24 both objectives at all stages of policy making (Markkanen and Anger-Kraavi 2019), including
25 focusing on the carbon intensity of lifestyles (Scherer et al. 2018), attention to sufficiency and equity
26 (Fischetti 2018) and targeting the consumption of the richest and highest emitting households (Otto et
27 al. 2019).
28
29 In modelled mitigation pathways, inequality in per capita emissions between regions are generally
30 reduced over time, and the reduction is generally more pronounced in lower temperature pathways
31 (Box 3.6 Figure 1). Already in 2030, if Nationally Determined Contributions from the Paris
32 Agreement are fully achieved, inequalities in per capita GHG emissions between countries would be
33 reduced (Benveniste et al. 2018).
34
35
36 Box 3.6 Figure 1: Difference in per capita emissions of Kyoto gases between the highest emitting and the
37 lowest emitting of the 10 regions, in 2030 and 2050, by temperature category of pathways
Do Not Cite, Quote or Distribute 3-101 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1
2 Through avoiding impacts of climate change, which fall more heavily on low-income countries,
3 communities and households and exacerbate poverty, mitigation reduces inequalities and poverty (see
4 section 3.6.4.2).
5
6 END BOX HERE
7
8 The remainder of this section covers specific domains of sustainable development: food (3.7.2), water
9 (3.7.3), energy (3.7.4), health (3.7.5), biodiversity (3.7.6) and multisector - Cities, infrastructure,
10 industry, production & consumption (3.7.7). These represent the areas with the strongest research
11 connecting mitigation to sustainable development. The links to individual SDGs are given within
12 these sections. Each domain covers the benefits of avoided climate impacts and the implications
13 (synergies and trade-offs) of mitigation efforts.
14
15 3.7.2 Food
16 The goal of SDG2 is to achieve “zero-hunger” by 2030. According to the UN (2015), over 25% of the
17 global population currently experience food insecurity and nearly 40% of these experience severe
18 food insecurity, a situation worsened by the covid pandemic (Paslakis et al. 2021).
19
20 3.7.2.1 Benefits of avoided climate impacts along mitigation pathways
21 Climate change will reduce crop yields, increase food insecurity, and negatively influence nutrition
22 and mortality (high confidence) (AR6 WGII Chapter 5). Climate mitigation will thus reduce these
23 impacts, and hence reduce food insecurity (high confidence). The yield reduction of global food
24 production will increase food insecurity and influence nutrition and mortality (Springmann et al.
25 2016a; Hasegawa et al. 2014). For instance, (Springmann et al. 2016a) estimate that climate change
26 could lead to 315,000-736,000 additional deaths by 2050, though these could mostly be averted by
27 stringent mitigation efforts. Reducing warming reduces the impacts of climate change, including
28 extreme climates, on food production and risk of hunger (Hasegawa et al. 2014, 2021b).
29
30 3.7.2.2 Implications of mitigation efforts along pathways
31 Recent studies explore the effect of climate change mitigation on agricultural markets and food
32 security (Hasegawa et al. 2018; Havlík et al. 2014; Fujimori et al. 2019; Doelman et al. 2019).
33 Mitigation policies aimed at achieving 1.5-2°C, if not managed properly, could negatively affect the
34 food security through changes in land and food prices (high confidence), leading to increases in the
35 population at risk of hunger by 80 to 280 million people compared to baseline scenarios. These
36 studies assume uniform carbon prices on AFOLU sectors (with some sectoral caps) and do not
37 account for climate impacts on food production.
38
39 Mitigating climate change while ensuring that food security is not adversely affected requires a range
40 of different strategies and interventions (high confidence). (Fujimori et al. 2018) explore possible
41 economic solutions to these unintended impacts of mitigation (e.g., agricultural subsidies, food aid,
42 and domestic reallocation of income) with an additional small (<0.1%) change in global GDP.
43 Targeted food-security support is needed to shield impoverished and vulnerable people from the risk
44 of hunger that could be caused by the economic effects of policies narrowly focussed on climate
45 objectives. Introducing more biofuels and careful selection of bioenergy feedstocks could also reduce
46 negative impacts (FAO WFP, WHO 2017). Reconciling bioenergy demands with food and
47 biodiversity, as well as competition for land and water, will require changes in food systems –
48 agricultural intensification, open trade, less consumption of animal-products and reduced food losses
49 – and advanced biotechnologies (Henry et al. 2018; Xu et al. 2019).
50
Do Not Cite, Quote or Distribute 3-102 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 There are many other synergistic measures for climate mitigation and food security. Agricultural
2 technological innovation can improve the efficiency of land use and food systems, thus reducing the
3 pressure on land from increasing food demand (Foley et al. 2011; Humpenöder et al. 2018; Popp et al.
4 2014; Obersteiner et al. 2016; Doelman et al. 2019). Furthermore, decreasing consumption of animal
5 products could contribute to SDG3.4 by reducing the risk of non-communicable diseases (Garnett
6 2016).
7
8 Taken together, climate changes will reduce crop yields, increase food insecurity and influence
9 nutrition and mortality (high confidence) (see 3.7.2.1). However, if measures are not properly
10 designed, mitigating climate change will also negatively impact on food consumption and security.
11 Additional solutions to negative impacts associated with climate mitigation on food production and
12 consumption include a transition to a sustainable agriculture and food system that is less resource
13 intensive, more resilient to a changing climate, and in line with biodiversity and social targets (Kayal
14 et al. 2019).
15
16 3.7.3 Water
17 Water is relevant to SDG 6 (clean water and sanitation), SDG 15 (ecosystem protection and water
18 systems), and SDG Targets 12.4 and 3.9 (water pollution and health). This section discusses water
19 quantity, water quality, and water-related extremes. See 3.7.5 for water-related health effects.
20
21 3.7.3.1 Benefits of avoided climate impacts along mitigation pathways
22 Global precipitation, evapotranspiration, runoff and water availability increase with warming
23 (Hanasaki et al. 2013; Greve et al. 2018) (see also WGII Chapter 4). Climate change also affects the
24 occurrence of and exposure to hydrological extremes (high confidence) (Arnell and Lloyd-Hughes
25 2014; Asadieh and Krakauer 2017; Dottori et al. 2018; IPCC 2019a; Naumann et al. 2018; Do et al.
26 2020) (see also WGII Chapter 4). Climate models project increases in precipitation intensity (high
27 confidence), local flooding (medium confidence), and drought risk (very high confidence) (Arnell and
28 Lloyd-Hughes 2014; Asadieh and Krakauer 2017; Dottori et al. 2018; IPCC 2019a) (see also WGII
29 Chapter 4).
30
31 The effect of climate change on water availability and hydrological extremes varies by region (high
32 confidence) due to differences in the spatial patterns of projected precipitation changes (Hanasaki et
33 al. 2013; Koutroulis et al. 2019; Schewe et al. 2014; Schlosser et al. 2014; Asadieh and Krakauer
34 2017; Dottori et al. 2018; Naumann et al. 2018) (see also WGII Chapter 4). Global exposure to water
35 stress is projected to increase with increased warming, but increases will not occur in all regions
36 (Arnell and Lloyd-Hughes 2014; Gosling and Arnell 2016; Hanasaki et al. 2013; IPCC 2019a;
37 Schewe et al. 2014).
38
39 Limiting warming could reduce water-related risks (high confidence) (O’Neill et al. 2017b; Byers et
40 al. 2018; Hurlbert et al. 2019) (see also WGII Chapter 4) and the population exposed to increased
41 water stress (Arnell and Lloyd-Hughes 2014; Gosling and Arnell 2016; Hanasaki et al. 2013; IPCC
42 2019a; Schewe et al. 2014).
43
44 The effect of climate change on water depends on the climate model, the hydrological model, and the
45 metric (high confidence) stress (Arnell and Lloyd-Hughes 2014; Gosling and Arnell 2016; Hanasaki
46 et al. 2013; IPCC 2019a; Schewe et al. 2014; Schlosser et al. 2014). However, the effect of
47 socioeconomic development could be larger than the effect of climate change (high confidence)
48 (Arnell and Lloyd-Hughes 2014; Schlosser et al. 2014; Graham et al. 2020).
49
Do Not Cite, Quote or Distribute 3-103 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 Climate change can also affect water quality (both thermal and chemical) (Liu et al. 2017), leading to
2 increases in stream temperature and nitrogen loading in rivers (Ballard et al. 2019).
3
4 3.7.3.2 Implications of mitigation efforts along pathways
5
6 The effect of mitigation on water demand depends on the mitigation technologies deployed (high
7 confidence) (Bonsch et al. 2016; Chaturvedi et al. 2013a,b; Jakob and Steckel 2016; Hejazi et al.
8 2014; Kyle et al. 2013; Fujimori et al. 2017; Maïzi et al. 2017; Mouratiadou et al. 2016; Parkinson et
9 al. 2019; Bijl et al. 2018; Cui et al. 2018; Graham et al. 2018; Hanasaki et al. 2013). Some mitigation
10 options could increase water consumption (volume removed and not returned) while decreasing
11 withdrawals (total volume of water removed, some of which may be returned) (Kyle et al. 2013;
12 Mouratiadou et al. 2016; Fricko et al. 2016; Parkinson et al. 2019). Bioenergy and BECCS can
13 increase water withdrawals and water consumption (high confidence) (Bonsch et al. 2016; Chaturvedi
14 et al. 2013a; Hejazi et al. 2014; Jakob and Steckel 2016; Kyle et al. 2013; Fujimori et al. 2017; Maïzi
15 et al. 2017; Mouratiadou et al. 2016; Parkinson et al. 2019; Yamagata et al. 2018; Séférian et al. 2018)
16 (see also WGII Chapter 4). DACCS (Fuhrman et al. 2020) and CCS (Kyle et al. 2013; Fujimori et al.
17 2017) could increase water demand; however, the implications of CCS depend on the cooling
18 technology and when capture occurs (Magneschi et al. 2017; Maïzi et al. 2017; Giannaris et al. 2020).
19 Demand-side mitigation (e.g., dietary change, reduced food waste, reduced energy demand) can
20 reduce water demand (Bajželj et al. 2014; Aleksandrowicz et al. 2016; Springmann et al. 2018; Green
21 et al. 2018). Introducing specific measures (e.g., environmental flow requirements, improved
22 efficiency, priority rules) can reduce water withdrawals (Bertram et al. 2018; Bijl et al. 2018;
23 Parkinson et al. 2019).
24
25 The effect of mitigation on water quality depends on the mitigation option, its implementation, and
26 the aspect of quality considered (high confidence) (McElwee et al. 2020; Smith et al. 2019; Sinha et
27 al. 2019; Ng et al. 2010; Fuhrman et al. 2020; Flörke et al. 2019; Karlsson et al. 2020).
28
29 3.7.4 Energy
30 Energy is relevant to SDG7 on sustainable and affordable energy access. Access to sufficient levels of
31 reliable, affordable and renewable energy is essential for sustainable development. Currently, over 1
32 billion people still lack access to electricity (Ribas et al. 2019).
33
34 3.7.4.1 Benefits of avoided climate impacts along mitigation pathways
35 Climate change alters the production of energy through changes in temperature (hydropower, fossil
36 fuel, nuclear, solar, bioenergy, transmission and pipelines), precipitation (hydropower, fossil fuel,
37 nuclear, bioenergy), windiness (wind, wave), and cloudiness (solar) (high confidence). Increases in
38 temperature reduce efficiencies of thermal power plants (e.g., fossil fuel and nuclear plants) with air-
39 cooled condensers by 0.4-0.7% oC increase in ambient temperature (Cronin et al. 2018a; Yalew, S. G.
40 et al. 2020; Simioni and Schaeffer 2019). Potentials and costs for renewable energy technologies are
41 also affected by climate change, though with considerable regional variation and uncertainty (Gernaat
42 et al. 2021). Biofuel yields could increase or decrease depending on the level of warming, changes in
43 precipitation, and the effect of CO2 fertilization (Calvin et al. 2013; Kyle et al. 2014; Gernaat et al.
44 2021). Coastal energy facilities could potentially be impacted by sea-level rise (Brown et al. 2014).
45
46 The energy sector uses large volumes of water (Fricko et al. 2016), making it highly vulnerable to
47 climate change (Tan and Zhi 2016) (high confidence). Thermoelectric and hydropower sources are the
48 most vulnerable to water stress (van Vliet et al. 2016). Restricted water supply to these power sources
49 can affect grid security and affordable energy access (Koch et al. 2014; Ranzani et al. 2018; Zhang et
50 al. 2018d).The hydropower facilities from high mountain areas of Central Europe, Iceland, Western
Do Not Cite, Quote or Distribute 3-104 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 US/Canada, and Latin America (Hock et al. 2019) as well as Africa and China (Bartos and Chester
2 2015; Eyer and Wichman 2018; Tarroja et al. 2016; Savelsberg et al. 2018; Ranzani et al. 2018;
3 Zhang et al. 2018d; Conway et al. 2017; Zhou et al. 2018; Gaupp et al. 2015; Wang et al. 2019; Byers
4 et al. 2018) have experienced changes in seasonality and availability.
5
6 3.7.4.2 Implications of mitigation efforts along pathways
7 Extending energy access to all in line with SDG7 is compatible with strong mitigation consistent with
8 the Paris agreement (high confidence). The Low Energy Demand (LED) scenario projects that these
9 twin goals can be achieved by relying heavily on energy efficiency and rapid social transformations
10 (Grubler et al. 2018). The IEA’s Sustainable Development Scenario (IEA 2020a) achieves
11 development outcomes but with higher average energy use, and bottom-up modelling suggests that
12 decent living standards could be provided to all in 2040-2050 with roughly 150 EJ, or 40% of current
13 final energy use (Kikstra et al. 2021b; Millward-Hopkins et al. 2020). The trade-offs between climate
14 mitigation and increasing energy consumption of the world’s poorest are negligible (Rao and Min
15 2018; Scherer et al. 2018).
16
17 The additional energy demand to meet the basic cooling requirement in Global South is estimated to
18 be much larger than the electricity needed to provide basic residential energy services universally via
19 clean and affordable energy, as defined by SDG7 (IEA 2019; Mastrucci et al. 2019) (high
20 confidence). If conventional air conditioning systems are widely deployed to provide cooling, energy
21 use could rise significantly (van Ruijven et al. 2019; Falchetta and Mistry 2021; Bezerra et al. 2021),
22 thus creating a positive feedback further increasing cooling demand. However, the overall emissions
23 are barely altered by the changing energy demand composition with reductions in heating demand
24 occurring simultaneously (Isaac and van Vuuren 2009; Labriet et al. 2015; McFarland et al. 2015;
25 Clarke et al. 2018). Some mitigation scenarios show price increases of clean cooking fuels, slowing
26 the transition to clean cooking fuels (SDG 7.1) and leaving a billion people in 2050 still reliant on
27 solid fuels in South Asia (Cameron et al. 2016).
28
29 In contrast, future energy infrastructure could improve reliability, thus lowering dependence on high-
30 carbon, high-air pollution backup diesel generators (Farquharson et al. 2018) that are often used to
31 cope with unreliable power in developing countries (Maruyama Rentschler et al. 2019). There can be
32 significant reliability issues where mini-grids are used to electrify rural areas (Numminen and Lund
33 2019). A stable, sustainable energy transition policy that considers national sustainable development
34 in the short- and long-term is critical in driving a transition to an energy future that addresses the
35 trilemma of energy security, equity, and sustainability (La Viña et al. 2018).
36
37 3.7.5 Health
38
39 SDG 3 aims to ensure healthy lives and promote well-being for all at all ages. Climate change is
40 increasingly causing injuries, illnesses, malnutrition, threats to mental health and well-being, and
41 deaths (see WGII Chapter 7). Mitigation policies and technologies to reduce GHG emissions are often
42 beneficial for human health on a shorter time scale than benefits in terms of slowing climate change
43 (Limaye et al. 2020). The financial value of health benefits from improved air quality alone is
44 projected to exceed the costs of meeting the goals of the Paris Agreement (Markandya et al. 2018).
45
46 3.7.5.1 Benefits of avoided climate impacts along mitigation pathways
47
48 The human health chapter of the WGII contribution to the AR6 concluded that climate change is
49 increasingly affecting a growing number of health outcomes, with negative net impacts at the global
50 scale and positive only in a few limited situations. There are few estimates of economic costs of
Do Not Cite, Quote or Distribute 3-105 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 increases in climate-sensitive health outcomes. In the U.S. in 2012, the financial burden in terms of
2 deaths, hospitalizations, and emergency department visits for ten climate-sensitive events across 11
3 states were estimated to be USD 10.0 (2.7 – 24.6) billion in 2018 dollars (Limaye et al. 2019).
4
5 3.7.5.2 Implications of mitigation efforts along pathways
6
7 Transitioning toward equitable, low-carbon societies has multiple co-benefits for health and wellbeing
8 (see WGII Chapter 7). Health benefits can be gained from improvements in air quality through
9 transitioning to renewable energy and active transport (e.g., walking and cycling); shifting to
10 affordable low-meat, plant-rich diets; and green buildings and nature-based solutions, such as green
11 and blue urban infrastructure, as shown in Figure 3.40 (Iacobucci 2016).
12
13
14 Figure 3.40: Diagram showing the co-benefits between health and mitigation (Iacobucci 2016)
15
16 The avoided health impacts associated with climate change mitigation can substantially offset
17 mitigation costs at the societal level (Chang et al. 2017; Ščasný et al. 2015; Schucht et al. 2015;
18 Markandya et al. 2018). Models of health co-benefits show that a 1.5 C pathway could result in 152
19 +/- 43 million fewer premature deaths worldwide between 2020 and 2100 in comparison to a
20 business-as-usual scenario, particularly due to reductions in exposure to PM2.5 (Shindell et al. 2018;
21 Rauner et al. 2020a; Rafaj et al. 2021). Some of the most substantial health, wellbeing, and equity
22 benefits associated with climate action derive from investing in basic infrastructure: sanitation, clean
23 drinking water, clean energy, affordable healthy diets, clean public transport, and improved air quality
24 from transformative solutions across economic sectors including agriculture, energy, transport and
25 buildings (Chang et al. 2017).
Do Not Cite, Quote or Distribute 3-106 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1
2 The health co-benefits of the NDCs for 2040 were compared for two scenarios, one consistent with
3 the goal of the Paris Agreement and the SDGs and the other also placing health as a central focus of
4 the policies (i.e., health in all climate policies scenario) (Hamilton et al. 2021), for Brazil, China,
5 Germany, India, Indonesia, Nigeria, South Africa, the UK, and the USA. Modelling of the energy,
6 food and agriculture, and transport sectors, and associated risk factors related to mortality, suggested
7 the sustainable pathways scenario could result in annual reductions of 1.18 million air pollution-
8 related deaths, 5.86 million diet-related deaths, and 1.15 million deaths due to physical inactivity.
9 Adopting the more ambitious health in all climate policies scenario could result in further reductions
10 of 462,000 annual deaths attributable to air pollution, 572,000 annual deaths attributable to diet, and
11 943,000 annual deaths attributable to physical inactivity. These benefits were attributable to the
12 mitigation of direct greenhouse gas emissions and the commensurate actions that reduce exposure to
13 harmful pollutants, as well as improved diets and safe physical activity.
14
15 Cost-benefit analyses for climate mitigation in urban settings that do not account for health may
16 underestimate the potential cost savings and benefits (Hess et al. 2020). The net health benefits of
17 controlling air pollution as part of climate mitigation efforts could reach trillions of dollars annually,
18 depending on the air quality policies adopted globally (Markandya et al. 2018; Scovronick et al.
19 2019b). Air pollution reductions resulting from meeting the Paris Agreement targets were estimated to
20 provide health co-benefits-to-mitigation ratios of between 1.4 and 2.5 (Markandya et al. 2018). In
21 Asia, the benefit of air pollution reduction through mitigation measures were estimated to reduce
22 premature mortality by 0.79 million, with an associated health benefit of USD2.8 trillion versus
23 mitigation costs of USD840 billion, equating to 6% and 2% of GDP, respectively (Xie et al. 2018).
24 Similarly, stabilizing radiative forcing to 3.4 W/m2 in South Korea could cost USD1.3-8.5 billion in
25 2050 and could lead to a USD23.5 billion cost reduction from the combined benefits of avoided
26 premature mortality, health expenditures, and lost work hours (Kim et al. 2020). The health co-
27 benefits related to physical exercise and reduced air pollution largely offset the costs of implementing
28 low CO2 emitting urban mobility strategies in three Austrian cities (Wolkinger et al. 2018).
29
30 Just in the United States of America, over the next 50 years, a 2°C pathway could prevent roughly 4.5
31 million premature deaths, about 3.5 million hospitalizations and emergency room visits, and
32 approximately 300 million lost workdays (Shindell 2020). The estimated yearly benefits of USD700
33 billion were more than the estimated cost of the energy transition.
34
35 3.7.6 Biodiversity (land and water)
36
37 Biodiversity covers Life Below Water (SDG 14) and Life On Land (SDG 15). Ecosystem services are
38 relevant to the goals of Zero Hunger (SDG 2), Good Health and Well Being (SDG 3), Clean Water
39 and Sanitation (SDG 6) and Responsible Consumption and Production (SDG 12), as well as being
40 essential to human existence (Díaz et al. 2019).
41
42 3.7.6.1 Benefits of avoided climate impacts along mitigation pathways
43
44 3.7.6.1.1 Terrestrial and freshwater aquatic ecosystems
45 Climate change is a major driver of species extinction and terrestrial and freshwater ecosystems
46 destruction (see WGII Chapter 2) (high confidence). Analysis shows that approximately half of all
47 species with long-term records have shifted their ranges in elevation and about two thirds have
48 advanced their timing of spring events (Parmesan and Hanley 2015). Under 3.2 °C warming, 49% of
49 insects, 44% of plants and 26% of vertebrates are projected to be at risk of extinction. At 2°C, this
50 falls to 18% of insects, 16% of plants and 8% of vertebrates and at 1.5°C, to 6% of insects, 8% of
Do Not Cite, Quote or Distribute 3-107 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 plants and 4% of vertebrates (Warren et al. 2018). Incidents of migration of invasive species,
2 including pests and diseases, are also attributable to climate change, with negative impacts on food
3 security and vector-borne diseases. Moreover, if climate change reduces crop yields, cropland may
4 expand – a primary driver of biodiversity loss – in order to meet food demand (Molotoks et al. 2020).
5 Land restoration and halting land degradation under all mitigation scenarios has the potential for
6 synergy between mitigation and adaptation.
7
8 3.7.6.1.2 Marine and coastal ecosystems
9 Marine ecosystems are being affected by climate change and growing non-climate pressures including
10 temperature change, acidification, land-sourced pollution, sedimentation, resource extraction and
11 habitat destruction (high confidence) (IPCC 2019b; Bindoff et al. 2019). The impacts of climate
12 drivers and their combinations vary across taxa (WGII Chapter 3). The danger or warming and
13 acidification to coral reefs, rocky shores and kelp forests is well established (high confidence) (WGII
14 Chapter 3). Migration towards optimal thermal and chemical conditions (Burrows et al. 2019)
15 contributes to large scale redistribution of fish and invertebrate populations, and major impacts on
16 global marine biomass production and maximum sustainable yield (Bindoff et al. 2019).
17
18 3.7.6.2 Implications of mitigation efforts along pathways
19 Mitigation measures have the potential to reduce the progress of negative impacts on ecosystems,
20 although it is unlikely that all impacts can be mitigated (high confidence) (Ohashi et al. 2019). The
21 specifics of mitigation achievement are crucial, since large-scale deployment of some climate
22 mitigation and land-based CDR measures could have deleterious impacts on biodiversity (Santangeli
23 et al. 2016; Hof et al. 2018).
24 Climate change mitigation actions to reduce or slow negative impacts on ecosystems are likely to
25 support the achievement of SDGs 2, 3, 6, 12, 14 and 15. Some studies show that stringent and
26 constant GHG mitigation practices bring a net benefit to global biodiversity even if land-based
27 mitigation measures are also adopted (Ohashi et al. 2019), as opposed to delayed action which would
28 require much more widespread use of BECCS. Scenarios based on demand reductions of energy and
29 land-based production are expected to avoid many such consequences, due to their minimized reliance
30 on BECCS (Grubler et al. 2018; Conijn et al. 2018; Bowles et al. 2019; Soergel et al. 2021a).
31 Stringent mitigation that includes reductions in demand for animal-based foods and food-waste could
32 also relieve pressures on land-use and biodiversity (high confidence), both directly by reducing
33 agricultural land requirements (Leclère et al. 2020) and indirectly by reducing the need for land-based
34 CDR (van Vuuren et al. 2018).
35
36 As environmental conservation and sustainable use of the earth’s terrestrial species and ecosystems
37 are strongly related, recent studies have evaluated interconnections among key aspects of land and
38 show pathway to the global sustainable future of land (Erb et al. 2016; Humpenöder et al. 2018; Popp
39 et al. 2014; Obersteiner et al. 2016). Most studies agree that many biophysical options exist to achieve
40 global climate mitigation and sustainable land-use in future. Conserving local biodiversity requires
41 careful policy design in conjunction with land-use regulations and societal transformation in order to
42 minimize the conversion of natural habitats.
43
44 3.7.7 Cities and infrastructure
45 This subsection focuses upon SDG9, Industry, Innovation and Infrastructure and SDG11, Sustainable
46 Cities and Communities.
47
48 3.7.7.1 Benefits of avoided climate impacts along mitigation pathways
49 By 2100, urban population will be almost double and more urban areas will be built (Jiang and
50 O’Neill 2017), although Covid-19 may modify these trends (Kii 2021). Urbanization will amplify
Do Not Cite, Quote or Distribute 3-108 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 projected air temperature changes in cities, including amplifying heat waves (see WGI Chapter 10,
2 Box 10.3). Benefits of climate mitigation in urban areas include reducing heat, air pollution and
3 flooding. Industrial infrastructure and production-consumption supply networks also benefit from
4 avoided impacts.
5
6 3.7.7.2 Implications of mitigation efforts along pathways
7 Many co-benefits to urban mitigation actions (see Chapter 8, Section 8.2.1) that improve the
8 liveability of cities and contribute to achieving SDG11. In particular, compact, urban form, efficient
9 technologies and infrastructure can play a valuable role in mitigation by reducing energy demand
10 (Güneralp et al. 2017; Creutzig et al. 2016), thus averting carbon lock-in, while reducing land sprawl
11 and hence increasing carbon storage and biodiversity (D’Amour et al. 2017). Benefits of mitigation
12 include air quality improvements from decreased traffic and congestion when private vehicles are
13 displaced by other modes; health benefits from increases in active travel; and lowered urban heat
14 island effects from green-blue infrastructures (section 8.2.1).
15
16 However, increasing urban density or enlarging urban green spaces can increase property prices and
17 reduce affordability (Section 8.2.1). Raising living conditions for slum dwellers & people living in
18 informal settlements will require significant materials and energy; however, regeneration can be
19 conducted in ways that avoid carbon-intense infrastructure lock-in (see Chapter 8 & 9). Cities affect
20 other regions through supply chains (Marinova et al. 2020).
21
22 Sustainable production, consumption and management of natural resources are consistent with, and
23 necessary for, mitigation (Chapters 5 & 11). Demand-side measures can lower requirements for
24 upstream material and energy use (Chapter 5). In terms of industrial production, transformational
25 changes across sectors will be necessary for mitigation (Chapter 11, sections 11.3 and 11.4).
26
27 Addressing multiple SDG arenas requires new systemic thinking in the areas of governance and
28 policy, such as those proposed by (Sachs et al. 2019).
29
30 3.8 Feasibility of socio/techno/economic transitions
31 The objective of this section is to discuss concepts of feasibility in the context of the low carbon
32 transition and pathways. We aim to identify drivers of low carbon scenarios feasibility and to
33 highlight enabling conditions which cam ameliorate feasibility concerns.
34
35 3.8.1 Feasibility frameworks for the low carbon transition and scenarios
36 Effectively responding to climate change and achieving sustainable development requires overcoming
37 a series of challenges to transition away from fossil-based economies. Feasibility can be defined in
38 many ways (see also Chapter 1). The political science literature (Majone 1975a,b; Gilabert and
39 Lawford-Smith 2012) distinguishes the feasibility of ‘what’ (i.e. emission reduction strategies),’when
40 and where” (i.e. in the year 2050, globally) and “whom” (i.e. cities). It distinguishes desirability from
41 political feasibility (von Stechow et al. 2015): the former represents a normative assessments of the
42 compatibility with societal goals (i.e. SDGs) while the latter evaluate the plausibility of what can be
43 attained given the prevailing context of transformation (Nielsen et al. 2020). Feasibility concerns are
44 context and time dependent and malleable: enabling conditions can help overcome them. For
45 example, public support for carbon taxes has been hard to secure but appropriate policy design and
46 household rebates can help dissipate opposition (Carattini et al. 2019; Murray and Rivers 2015).
47
48 Regarding scenarios, the feasibility ‘what’ question is the one most commonly dealt with in the
49 literature, though most of the studies have focused on expanding low carbon system, and yet political
Do Not Cite, Quote or Distribute 3-109 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 constraints might arise mostly from phasing out fossil fuel-based ones (Spencer et al. 2018; Fattouh et
2 al. 2019). The ‘when and where’ dimension can also be related to the scenario assessment, but only
3 insofar the models generating them can differentiate time and geographical contextual factors.
4 Distinguishing mitigation potential by regional institutional capacity has a significant influence on the
5 costs of stabilizing climate (Iyer et al. 2015c). The ‘whom’ question is the most difficult to capture by
6 scenarios, given the multitude of actors involved as well as their complex interactions. The focus of
7 socio-technical transition sciences on the co-evolutionary processes can shed light on the dynamics of
8 feasibility (Nielsen et al. 2020).
9
10 The when-where-whom distinction allows depicting a feasibility frontier beyond which
11 implementation challenges prevent mitigation action (Jewell and Cherp 2020). Even if the current
12 feasibility frontier appears restraining in some jurisdictions, it is context-dependent and dynamic as
13 innovation proceeds and institutional capacity builds up (Nielsen et al. 2020). The question is whether
14 the feasibility frontier can move faster than the pace at which the carbon budget is being exhausted.
15 Jewell et al. (2019) show that the emission savings from the pledges of premature retirement of coal
16 plants is 150 times less than globally committed emissions from existing coal power plants. The
17 pledges come from countries with high institutional capacity and relatively low shares of coal in
18 electricity. Other factors currently limiting the capacity to steer transitions at the necessary speed
19 include the electoral market orientation of politicians (Willis 2017), the status quo orientation of
20 senior public officials (Geden 2016), path dependencies created by 'instrument constituencies' (Béland
21 and Howlett 2016), or the benefits of deliberate inconsistencies between talk, decisions and actions in
22 climate policy (Rickards et al. 2014). All in all, a number of different delay mechanisms in both
23 science and policy have been identified to potentially impede climate goal achievement (Karlsson and
24 Gilek 2020) - see also Chapter 13.
25
26 In addition to its contextual and dynamic nature, feasibility is a multi-dimensional concept. The IPCC
27 1.5°C special report distinguishes 6 dimensions of feasibility: geophysical, environmental-ecological,
28 technological, economic, socio-cultural and institutional. At the individual option level, different
29 mitigation strategies face various barriers as well as enablers (see Chapter 6 for the option-level
30 assessment). However, a systemic transformation involves interconnections of a wide range of
31 indicators. Model-based assessments are meant to capture the integrative elements of the transition
32 and of associated feasibility challenges. However, the translation of model-generated pathways into
33 feasibility concerns (Rogelj et al. 2018b) has developed only recently. Furthermore, multiple forms of
34 knowledge can be mobilized to support strategic decision-making and complement scenario analysis
35 (Turnheim and Nykvist 2019). We discuss both approaches next.
36
37 3.8.2 Feasibility appraisal of low carbon scenarios
38 Evaluating the feasibility of low carbon pathways can take different forms. In the narrowest sense,
39 there is feasibility pertaining the reporting of model-generated scenarios: here an infeasible scenario is
40 one which cannot meet the constraints embedded implicitly or explicitly in the models which
41 attempted to generate it. Second, there is a feasibility that relates to specific elements or overall
42 structure characterizing the low carbon transition compared to some specified benchmark.
43
44 3.8.2.1 Model solvability
45 In order to be generated, scenarios must be coherent with the constraints and assumptions embedded
46 in the models (i.e., deployment potential of given technologies, physical and geological limits) and in
47 the scenario design (i.e., carbon budget). Sometimes, models cannot solve specific scenarios. This
48 provides a first, coarse indication of feasibility concerns. Specific vetting criteria can be imposed,
49 such as carbon price values above which scenarios should not be reported, as in Clarke et al. (2009).
50 However, model solvability raises issues of aggregation in model ensemble. Since model solving is
Do Not Cite, Quote or Distribute 3-110 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 not a random process, but a function of the characteristics of the models, analysing only reported
2 outcomes leads to statistical biases (Tavoni and Tol 2010).
3
4 Although model-feasibility differs distinctly from feasibility in the real world, it can indicate the
5 relative challenges of low carbon scenarios - primarily when performed in a model ensemble of
6 sufficient size. Riahi et al. (2015) interpreted infeasibility across a large number of models as an
7 indication of increased risk that the transformation may not be attainable due to technical or economic
8 concerns. All models involved in a model comparison of 1.5°C targets (Rogelj et al. 2018b) (Table
9 S1) were able to solve under favourable underlying socio-economic assumptions (SSP1), but none for
10 the more challenging SSP3. This interpretation of feasibility was used to highlight the importance of
11 socio-economic drivers for attaining climate stabilization. Gambhir et al. (2017) constrained the
12 models to historically observed rates of change and found that it would no longer allow to solve for
13 2°C, highlighting the need for rapid technological change.
14
15 3.8.2.2 Scenario feasibility
16 Evaluating the feasibility of scenarios involves several steps (see Figure 3.41). First, one need to
17 identify which dimensions of feasibility to focus on. Then, for each dimension, one needs to select
18 relevant indicators for which sufficient empirical basis exists and which are an output of models (or at
19 least of a sufficient number of them). Then, thresholds marking different levels of feasibility concerns
20 are defined based on available literature, expert elicitations and empirical analysis based on
21 appropriately chosen historical precedents. Finally, scenario feasibility scores are obtained for each
22 indicator, and where needed aggregated up in time or dimensions, as a way to provide an overall
23 appraisal of feasibility trade-offs, depending on the timing, disruptiveness and scale of transformation.
24
25
26 Figure 3.41: Steps involved in evaluating the feasibility of scenarios
27 Most of the existing literature has focused on the technological dimensions, given the technology
28 focus of models and the ease of comparison. The literature points to varied findings. Some suggest
29 that scenarios envision technological progress consistent with historical benchmarks (Loftus et al.
30 2015; Wilson et al. 2013). Others that scenarios exceed historically observed rates of low carbon
31 technology deployment and of energy demand transformation globally (Napp et al. 2017; van der
32 Zwaan et al. 2013; Semieniuk et al. 2021; Cherp et al. 2021), but not for all countries (Cherp et al.
33 2021). The reason for these discrepancies depends on the unit of analysis and the indicators used.
34 Comparing a different kind of historical indicators, (van Sluisveld et al. 2015) find that indicators that
35 look into the absolute change of energy systems remain within the range of historical growth frontiers
36 for the next decade, but increase to unprecedented levels before mid-century. Expert assessments
37 provide another way of benchmarking scenarios, though they have shown to be systematically biased
38 (Wiser et al. 2021) and to underperform empirical methods (Meng et al. 2021). (van Sluisveld et al.
39 2018a) find that scenarios and experts align for baseline scenarios but differ for low carbon ones.
Do Not Cite, Quote or Distribute 3-111 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 Scenarios rely more on conventional technologies based on existing infrastructure (such as nuclear
2 and CCS) than what forecasted by experts. Overall, the technology assessment of the feasibility space
3 highlights that Paris-compliant transformations would have few precedents, but not zero (Cherp et al.
4 2021).
5
6 Recent approaches have addressed multiple dimensions of feasibility, an important advancement since
7 social and institutional aspects are as if not more important than technology ones (Jewell and Cherp
8 2020). Feasibility corridors of scenarios based on their scale, rate of change and disruptiveness have
9 been identified (Kriegler et al. 2018b; Warszawski et al. 2021). The reality check shows that many
10 1.5°C compatible scenarios violate the feasibility corridors. The ones which didn’t are associated with
11 a greater coverage of the available mitigation levers (Warszawski et al. 2021).
12
13 Brutschin et al. (2021) proposed an operational framework covering all six dimensions of feasibility.
14 They developed a set of multi-dimensional metrics capturing the timing, disruptiveness and the scale
15 of the transformative change within each dimension (as in Kriegler et al. (2018b)). Thresholds of
16 feasibility risks of different intensity are obtained through the review of the relevant literature and
17 empirical analysis of historical data. Novel indicators include governance levels (Andrijevic et al.
18 2020a). The 17 bottom-up indicators are then aggregated up across time and dimension, as way to
19 highlight feasibility trade-offs. Aggregation is done via compensatory approaches such as the
20 geometric mean. This is employed, for instance, for the Human Development Index. A conceptual
21 example of this approach as applied to the IPCC AR6 scenarios database is shown in Figure 3.42 and
22 further described in the Annex.
23
24
25 Figure 3.42: Example of multi-dimensional feasibility analysis and indicators used in the IPCC AR6
26 scenarios. The approach defines relevant indicators characterizing the key dimensions of feasibility.
27 Indicators capture the timing, scale and disruptiveness challenges. Low, medium and high feasibility
28 concerns are defined based on historical trends and available literature. Details about indicator and
29 thresholds values can be found in Annex III
30 In Figure 3.43, we show the results of applying the methodology of Brutschin et al. (2021) to the AR6
31 scenarios database. The charts highlight the dynamic nature of feasibility risks, which are mostly
32 concentrated in the decades before mid-century except for geophysical risks driven by CO2 removals
Do Not Cite, Quote or Distribute 3-112 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 later in the century. Different dimensions pose differentiated challenges: for example, institutional
2 feasibility challenges appear to be the most relevant, in line with the qualitative literature. Thus,
3 feasibility concerns might be particularly relevant in countries with weaker institutional capacity. The
4 Figure also highlights the key role of policy and technology, as enabling factors. In particular (Panel
5 B), internationally coordinated and immediate emission reductions allow to smooth out feasibility
6 concerns and reduce long term challenges compared to delayed policy action, as a result of a more
7 gradual transition and lower requirements of CO2 removals. For the same climate objective, different
8 illustrative mitigation pathways entail somewhat different degrees and distributions of implementation
9 challenges (panel C).
10
Do Not Cite, Quote or Distribute 3-113 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 Figure 3.43: Feasibility characteristics of the Paris-consistent scenarios in AR6 scenarios database,
2 applying the methodology by : Feasibility corridors for the AR6 scenarios database, applying the
3 methodology by (Brutschin et al. 2021). Panel A: the fraction of scenarios falling within 3 categories of
4 feasibility concerns (Plausible, Best Case, Unprecedented), for different times (2030, 2050, 2100), different
5 climate categories consistent with the Paris Agreement and five dimensions. Panel B: composite feasibility
6 score (obtained by geometric mean of underlying indicators) over time for scenarios with immediate and
7 delayed global mitigation efforts, for different climate categories (C1-C2-C3. Note: no C1 scenarios has
8 delayed participation). Panel C: Fraction of scenarios which in any point in time over the century exceed
9 the feasibility concerns, for C1 and C3 climate categories. Overlayed are the Illustrative Mitigation
10 Pathways (LP, SP, Ren: C1 category; Neg, GS: C3 category)
11 3.8.3 Feasibility in the light of socio-technical transitions
12 The limitations associated with quantitative low-carbon transition pathways stem from a predominant
13 reliance on techno-economic considerations with a simplified or non-existent representation of the
14 socio-political and institutional agreement. Accompanying the required deployment of low carbon
15 technologies will be the formation of new socio-technical systems (Bergek et al. 2008). With a socio-
16 technical system being defined as a cluster of elements comprising of technology, regulation, user
17 practices and markets, cultural meaning, infrastructure, maintenance networks, and supply networks
18 (Geels and Geels 2005; Hofman et al. 2004); the interrelationship between technological systems and
19 social systems must be comprehensively understood. It is of vital importance that the pro-cess of
20 technical change must considered in its institutional and social context so as to ascertain potential
21 transition barriers which in turn provide an indication of pathway feasibility. In order to address the
22 multitudinous challenges associated with low-carbon transition feasibility and governance, it has been
23 opined that the robustness of evaluating pathways may be improved by the bridging of differing
24 quantitative-qualitative analytical approaches (Haxeltine et al. 2008; Foxon et al. 2010; Hughes 2013;
25 Wangel et al. 2013; Li et al. 2015; Turnheim et al. 2015; Geels et al. 2016a,b, 2018; Moallemi et al.
26 2017; De Cian et al. 2018; Li and Strachan 2019). The rationale for such analytical bridging is to
27 rectify the issue that in isolation each disciplinary approach only can generate a fragmented
28 comprehension of the transition pathway with the consequence being an incomplete identification of
29 associated challenges in terms of feasibility. Concerning low-carbon transition pathways generated by
30 IAMs, it has been argued that a comprehensive analysis should include social scientific enquiry
31 (Geels et al. 2016a, 2018; van Sluisveld et al. 2018b). The normative analysis of IAM pathways
32 assists in the generation of a vision or the formulation of a general plan with this being complemented
33 by socio-technical transition theory (Geels et al. 2016a). Such an approach thereby allowing for the
34 socio-political feasibility and the social acceptance and legitimacy of low-carbon options to be
35 considered. Combining computer models and the multi-level perspective can help identify ‘transition
36 bottlenecks’ (Geels et al. 2018). Similarly, increased resolution of integrated assessment models’
37 actors has led to more realistic narratives of transition in terms of granularity and behaviour
38 (McCollum et al. 2017; van Sluisveld et al. 2018b). Increased data availability of actual behaviour
39 from smart technology lowers the barriers to representing behavioural change in computer
40 simulations, and thus better represent crucial demand side transformations (Creutzig et al. 2018).
41 Increasing the model resolution is a meaningful way forward. However, integrating a much broader
42 combination of real-life aspects and dynamics into models could lead to an increased complexity that
43 could restrict them to smaller fields of applications (De Cian et al. 2018).
44
45 Other elements of feasibility relate to social justice, which could be essential to enhance the political
46 and public acceptability of the low carbon transition. Reviewing the literature, one study finds that
47 employing social justice as an orienting principle can increase the political feasibility of low carbon
48 policies (Patterson et al. 2018). Three elements are identified as key: i) protecting vulnerable people
49 from climate change impacts, ii), protecting people from disruptions of transformation, iii), enhancing
50 the process of envisioning and implementing an equitable post-carbon society.
51
Do Not Cite, Quote or Distribute 3-114 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 3.8.4 Enabling factors
2 There is strong agreement that the climate policy institutional framework as well as technological
3 progress have a profound impact on the attainability of low carbon pathways. Delaying international
4 cooperation reduces the available carbon budget and locks into carbon intensive infrastructure
5 exacerbating implementation challenges (Clarke et al. 2009; Bosetti et al. 2009; Krey and Riahi 2009;
6 Boucher et al. 2009; Keppo and Rao 2007; van Vliet et al. 2009; Knopf et al. 2011; Luderer et al.
7 2013; Jakob et al. 2012; Aboumahboub et al. 2014; Bertram et al. 2021; Popp et al. 2014; Rogelj et al.
8 2013a; Riahi et al. 2015; Kriegler et al. 2014a; Gambhir et al. 2017). Similarly, technological
9 availability influences the feasibility of climate stabilization, though differently for different
10 technologies (Iyer et al. 2015a; Kriegler et al. 2014a; Riahi et al. 2015).
11
12 One of the most relevant factors affecting mitigation pathways and their feasibility is the rate and kind
13 of socio-economic development. For example, certain socio-economic trends and assumptions
14 about policy effectiveness preclude achieving stringent mitigation futures (Rogelj et al. 2018b).
15 The risk of failure increases markedly in high growth, unequal and/or energy-intensive worlds -
16 such as those characterized by the shared socio-economic pathways SSP3, SSP4 and SSP5. On the
17 other hand, socio-economic development conducive to mitigation relieves the energy sector
18 transformation from relying on large scale technology development: for example, the amount of
19 biomass with CCS in SSP1 is one third of that in SSP5. The reason why socio-economic trends matter
20 so much is that they both affect the CO2 emissions in counterfactual scenarios as well as the
21 mitigation capacity (Riahi et al. 2017; Rogelj et al. 2018b). Economic growth assumptions are the
22 most important determinant of scenario emissions (Marangoni et al. 2017). De- and post-growth
23 scenarios have been suggested as valuable alternatives to be considered (Hickel et al. 2021; Keyßer
24 and Lenzen 2021), though substantial challenges remain regarding political feasibility (Keyßer and
25 Lenzen 2021).
26
27 The type of policy instrument assumed to drive the decarbonization process also play a vital role for
28 determining feasibility. The majority of scenarios exploring climate stabilization pathways in the past
29 have focused on uniform carbon pricing as the most efficient instrument to regulate emissions.
30 However, carbon taxation raises political challenges (Beiser-McGrath and Bernauer 2019), see also
31 Chapter 13 and 14. Carbon pricing will transfer economic surplus from consumers and producers to
32 the government. Losses for producers will be highly concentrated in those industries possessing fixed
33 or durable assets with “high asset specificity” (Murphy 2002; Dolphin et al. 2020). These sectors have
34 opposed climate jurisdictions (Jenkins 2014). Citizens are sensitive to rising energy prices, though
35 revenue recycling can be used to increase support (Carattini et al. 2019). A recent model comparison
36 project confirms findings from the extant literature: using revenues to reduce pre-existing capital or,
37 to a lesser extent, labour taxes, reduces policy costs and eases distributional concerns (Mcfarland et al.
38 2018; Barron et al. 2018).
39
40 Nonetheless, winning support will require a mix of policies which go beyond carbon pricing, and
41 include subsidies, mandates and feebates (Rozenberg et al. 2018; Jenkins 2014). More recent
42 scenarios take into account a more comprehensive range of policies and regional heterogeneity in the
43 near to medium term (Roelfsema et al. 2020). Regulatory policies complementing carbon prices could
44 reduce the implementation challenges by increasing short term emission reduction, though they could
45 eventually reduce economic efficiency (Bertram et al. 2015b; Kriegler et al. 2018a). Innovation
46 policies such as subsidies to R&D have been shown to be desirable due to innovation market failures,
47 and also address the dynamic nature of political feasibility (Bosetti et al. 2011).
48
Do Not Cite, Quote or Distribute 3-115 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 3.9 Methods of assessment and gaps in knowledge and data
2 3.9.1 AR6 mitigation pathways
3 The analysis in this chapter relies on the available literature as well as an assessment of the scenarios
4 contained in the AR6 scenarios database. Scenarios were submitted by research and other institutions
5 following an open call (see Annex III Part II). The scenarios included in the AR6 scenarios database
6 are an unstructured ensemble, as they are from multiple underlying studies and depend on which
7 institutions chose to submit scenarios to the database. As noted in Section 3.2, they do not represent
8 the full scenario literature or the complete set of possible scenarios. For example, scenarios that
9 include climate change impacts or economic degrowth are not fully represented, as these scenarios,
10 with a few exceptions, were not submitted to the database. Additionally, sensitivity studies, which
11 could help elucidate model behaviour and drivers of change, are mostly absent from the database -
12 though examples exist in the literature (Marangoni et al. 2017).
13
14 The AR6 scenarios database contains 3131 scenarios of which 2425 with global scope were
15 considered by this chapter, generated by almost 100 different model versions, from more than 50
16 model families. Of the 1686 vetted scenarios, 1202 provided sufficient information for a climate
17 categorization. Around 46% of the pathways are consistent with an end of century temperature of at
18 least likely limiting warming to below 2°C. There are many ways of constructing scenarios that limit
19 warming to a particular level and the choice of scenario construction has implications for the timing
20 of both net zero CO2 and GHG emissions and the deployment of CDR (Emmerling et al. 2019; Rogelj
21 et al. 2019b; Johansson et al. 2020). The AR6 scenarios database includes scenarios where
22 temperature is temporarily exceeded (40% of all scenarios in the database have median temperature in
23 2100 that is 0.1C lower than median peak temperature). Climate stabilization scenarios are typically
24 implemented by assuming a carbon price rising at a particular rate per year, though that rate varies
25 across model, scenario, and time period. Standard scenarios assume a global single carbon price to
26 minimize policy costs. Cost minimizing pathways can be reconciled with equity considerations
27 through posterior international transfers. Many scenarios extrapolate current policies and include non-
28 market, regulatory instruments such as technology mandates.
29
30 Scenarios are not independent of each other and not representative of all possible outcomes, nor of the
31 underlying scenario generation process; thus, the statistical power of the database is limited.
32 Dependencies in the data generation process originate from various sources. Certain model groups,
33 and types, are over-represented. For example, 8 model teams contributed with 90% of scenarios.
34 Second, not all models can generate all scenarios, and these differences are not random, thereby
35 creating selection bias (Tavoni and Tol 2010). Third, there are strong model dependencies: the
36 modelling scientific community shares code and data, and several IAMs are open source.
37
38 3.9.2 Models assessed in this chapter
39 The models assessed in this chapter differ in their sectoral coverage and the level of complexity in
40 each sector. Models tend to have more detail in their representation of energy supply and
41 transportation, than they do for industry (Section 3.4; Annex III). Some models include detailed land
42 use models, while others exclude land models entirely and use supply curves to represent bioenergy
43 potential (Bauer et al. 2018a). IAMs do not include all mitigation options available in the literature
44 (Rogelj et al. 2018b; Smith et al. 2019). For example, most IAM pathways exclude many granular
45 demand-side mitigation options and land-based mitigation options found in more detailed sectoral
46 models; additionally, only a few pathways include CDR options beyond afforestation/reforestation
47 and BECCS. Section 3.4 and Chapter 12 include some results and comparisons to non-IAM models
48 (e.g., bottom-up studies and detailed sectoral models). These sectoral studies often include a more
49 complete set of mitigation options but exclude feedbacks and linkages across sectors which may alter
Do Not Cite, Quote or Distribute 3-116 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 the mitigation potential of a given sector. There is an increasing focus in IAM studies on SDGs
2 (Section 3.7), with some studies reporting the implications of mitigation pathways on SDGs (e.g.,
3 Bennich et al. (2020)) and others using achieving SDGs as a constraint on the scenario itself (van
4 Vuuren et al. 2015; Soergel et al. 2021a). However, IAMs are still limited in the SDGs they represent,
5 often focusing on energy, water, air pollution and land. On the economic side, the majority of the
6 models report information on marginal costs (i.e., carbon price). Only a subset provides full economic
7 implications measured by either economic activity or welfare. Also often missing, is detail about
8 economic inequality within countries or large aggregate regions.
9
10 For further details about the models and scenarios, see Annex III.
11
12
13 Frequently Asked Questions (FAQs)
14 FAQ 3.1 – Is it possible to stabilize warming without net negative CO2 and GHG emissions?
15 Yes. Achieving net zero CO2 emissions and sustaining them into the future is sufficient to stabilize
16 the CO2-induced warming signal which scales with the cumulative net amount of CO2 emissions. At
17 the same time, the warming signal of non-CO2 GHGs can be stabilized or reduced by declining
18 emissions that lead to stable or slightly declining concentrations in the atmosphere. For short-lived
19 GHGs with atmospheric lifetimes of less than 20 years, this is achieved when residual emissions are
20 reduced to levels that are lower than the natural removal of these gases in the atmosphere. Taken
21 together, mitigation pathways that bring CO2 emissions to net zero and sustain it, while strongly
22 reducing non-CO2 GHGs to levels that stabilize or decline their aggregate warming contribution, will
23 stabilize warming without using net negative CO2 emissions and with positive overall GHG emissions
24 when aggregated using GWP100. A considerable fraction of pathways limiting warming to 1.5°C
25 with no or limited overshoot and likely limiting warming to 2°C, respectively, do not or only
26 marginally (<10 GtCO2 cumulative until 2100) deploy net negative CO2 emissions (26% and 46%,
27 respectively) and do not reach net zero GHG emissions by the end of the century (48% and 70%,
28 respectively). This is no longer the case in pathways that return warming to 1.5°C after a high
29 overshoot (typically > 0.1°C). All of these pathways deploy net negative emissions on the order of
30 330 (26-623 GtCO2) (median and 5th-95th percentile) and 87% achieve net negative GHGs emissions
31 in AR6 GWP-100 before the end of the century. Hence, global net negative CO2 emissions, and net
32 zero or net negative GHG emissions, are only needed to decline, not to stabilize global warming. The
33 deployment of carbon dioxide removal (CDR) is distinct from the deployment of net negative CO2
34 emissions, because it is also used to neutralize residual CO2 emissions to achieve and sustain net zero
35 CO2 emissions. CDR deployment can be considerable in pathways without net negative emissions and
36 all pathways limiting warming to 1.5°C use it to some extent.
37
38 FAQ 3.2 – How can net zero emissions be achieved and what are the implications of net zero
39 emissions for the climate?
40 Halting global warming in the long term requires, at a minimum, that no additional CO2 emissions
41 from human activities are added to the atmosphere (i.e., CO2 emissions must reach “net” zero). Given
42 that CO2 emissions constitute the dominant human influence on global climate, global net zero CO2
43 emissions are a prerequisite for stabilizing warming at any level. However, CO2 is not the only
44 greenhouse gas that contributes to global warming and reducing emissions of other greenhouse gases
45 alongside CO2 towards net zero emissions of all GHGs would lower the level at which global
46 temperature would peak. The temperature implications of net zero GHG emissions depend on the
47 bundle of gases that is being considered, and the emissions metric used to calculate aggregated GHG
48 emissions and removals. If reached and sustained, global net zero GHG emissions using the 100-year
49 Global Warming Potential (GWP-100) will lead to gradually declining global temperature.
Do Not Cite, Quote or Distribute 3-117 Total pages: 156
Final Government Distribution Chapter 3 IPCC AR6 WGIII
1 Not all emissions can be avoided. Achieving net zero CO2 emissions globally therefore requires deep
2 emissions cuts across all sectors and regions, along with active removal of CO2 from the atmosphere
3 to balance remaining emissions that may be too difficult, too costly or impossible to abate at that time.
4 Achieving global net zero GHG emissions would require, in addition, deep reductions of non-CO2
5 emissions and additional CO2 removals to balance remaining non-CO2 emissions.
6 Not all regions and sectors must reach net zero CO2 or GHG emissions individually to achieve global
7 net zero CO2 or GHG emissions, respectively; instead, positive emissions in one sector or region can
8 be compensated by net negative emissions from another sector or region. The time each sector or
9 region reaches net zero CO2 or GHG emissions depends on the mitigation options available, the cost
10 of those options, and the policies implemented (including any consideration of equity or fairness).
11 Most modelled pathways that likely limit warming to 2°C above pre-industrial levels and below use
12 land-based CO2 removal such as afforestation/reforestation and BECCS to achieve net zero CO2 and
13 net zero GHG emissions even while some CO2 and non-CO2 emissions continue to occur. Pathways
14 with more demand-side interventions that limit the amount of energy we use, or where the diet that we
15 consume is changed, can achieve net zero CO2, or net zero GHG emissions with less carbon dioxide
16 removal. All available studies require at least some kind of carbon dioxide removal to reach net zero;
17 that is, there are no studies where absolute zero GHG or even CO2 emissions are reached by deep
18 emissions reductions alone.
19 Total GHG emissions are greater than emissions of CO2 only; reaching net zero CO2 emissions
20 therefore occurs earlier, by up to several decades, than net zero GHG emissions in all modelled
21 pathways. In most modelled pathways that likely limit warming to 2°C above pre-industrial levels and
22 below in the most cost-effective way, the AFOLU and energy supply sectors reach net zero CO2
23 emissions several decades earlier than other sectors; however, many pathways show much reduced,
24 but still positive net GHG emissions in the AFOLU sector in 2100.
25
26 FAQ 3.3 – How plausible are high emissions scenarios, and how do they inform policy?
27 IAMs are used to develop a wide range of scenarios describing future trajectories for greenhouse gas
28 emissions based on a wide set of assumptions regarding socio-economic development, technological
29 changes, political development and climate policy. Typically, the IAM-based scenarios can be divided
30 into a) reference scenarios (describing possible trajectories in the absence of new stringent climate
31 policies) and b) mitigation scenarios (describing the impact of various climate policy assumptions).
32 Reference scenarios typically result in high emissions and, subsequently, high levels of climate
33 change (in the order of 2.5-4 °C during the 21st century). The purpose of such reference scenarios is
34 to explore the consequences of climate change and act as a reference for mitigation scenarios. The
35 possible emission levels for reference scenarios diverge from stabilising and even slowly declining
36 emissions (e.g., for current policy scenarios or SSP1) to very high emission levels (e.g., SSP5 and
37 RCP8.5). The latter leads to nearly 5 °C of warming by the end of the century for medium climate
38 sensitivity. Hausfather and Peters (2020) pointed out that since 2011, the rapid development of
39 renewable energy technologies and emerging climate policy have made it considerably less likely that
40 emissions could end up as high as RCP8.5. This means that reaching emissions levels as high as
41 RCP8.5 has become less likely. Still, high emissions cannot be ruled out for many reasons, including
42 political factors and, for instance, higher than anticipated population and economic growth. Climate
43 projections of RCP8.5 can also result from strong feedbacks of climate change on (natural) emission
44 sources and high climate sensitivity (see WGI Chapter 7). Therefore, their median climate impacts
45 might also materialise while following a lower emission path (e.g., Hausfather and Betts (2020)). All-
46 in-all, this means that high-end scenarios have become considerably less likely since AR5 but cannot
47 be ruled out. High-end scenarios (like RCP8.5) can be very useful to explore high-end risks of climate
48 change but are not typical ‘business-as-usual” projections and should therefore not be presented as
49 such.
50
Do Not Cite, Quote or Distribute 3-118 Total pages: 156