Climate Change 2021

The Physical Science Basis

Working Group I Contribution to the

WGI Sixth Assessment Report of the
Intergovernmental Panel on Climate Change
 Climate Change 2021
The Physical Science Basis

Working Group I Contribution to

the Sixth Assessment Report of the
Intergovernmental Panel on Climate Change

Edited by

Valérie Masson-Delmotte Panmao Zhai

Co-Chair Working Group I Co-Chair Working Group I

Anna Pirani Sarah L. Connors Clotilde Péan

Head of TSU Head of Science Team Head of Operations

Yang Chen Leah Goldfarb Melissa I. Gomis

Senior Science officer Senior Science officer Senior Science officer

J.B. Robin Matthews Sophie Berger Mengtian Huang

Senior Science officer Science Officer Science Officer

Ozge Yelekçi Rong Yu Baiquan Zhou

Science Officer Science Officer Science Officer

Elisabeth Lonnoy Thomas K. Maycock Tim Waterfield

Project Assistant Science Editor IT Officer

Katherine Leitzell Nada Caud

Communication Manager Outreach Manager

Working Group I Technical Support Unit

© 2021 Intergovernmental Panel on Climate Change.
Electronic copies of this report are available from the IPCC website www.ipcc.ch

Front cover artwork: Changing by Alisa Singer, www.environmentalgraphiti.org © 2021 Alisa Singer.

Use the following reference to cite the entire volume:

IPCC, 2021: Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on
Climate Change [Masson-Delmotte, V., P. Zhai, A. Pirani, S.L. Connors, C. Péan, S. Berger, N. Caud, Y. Chen, L. Goldfarb, M.I. Gomis, M. Huang, K. Leitzell, E. Lonnoy, J.B.R.
Matthews, T.K. Maycock, T. Waterfield, O. Yelekçi, R. Yu, and B. Zhou (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 2391 pp.
doi:10.1017/9781009157896.
Foreword, Preface
and Dedication
 Foreword

Foreword

Foreword
It is unequivocal that human activities have heated our climate. The journey of this report reflects extraordinary efforts by all
Recent changes are rapid, intensifying, and unprecedented over contributors under exceptional circumstances. The COVID-19
centuries to thousands of years. With each additional increment of pandemic, which started during the review of the Second Order
warming, these changes will become larger, resulting in long-lasting, Draft, disrupted the process. However, the early response of Working
irreversible implications, in particular for sea level rise. United Nations Group I, through the hard work of the authors, review editors, chapter
Secretary-General António Guterres has stated that ‘the evidence is scientists, the Technical Support Unit, and Bureau members, made
irrefutable’ and ‘we see the warning signs in every continent and region’. it possible to deliver a report that meets the most stringent science
quality standards, thanks to the in-depth, broad review process, with
The Working Group I contribution to the Intergovernmental Panel an exhaustive, robust, rigorous and transparent assessment of the
on Climate Change (IPCC) Sixth Assessment Report provides a reality latest state of climate science knowledge. We commend the additional
check on climate change. We now have a much clearer picture of efforts taken, including by all reviewers from the research community
the past, present and possible future climates, and this information and government representatives who adapted to a new way of working
is essential for understanding where we are headed, what can be to complete this vital report. Our particular appreciation goes to the
done, and the multiple facets of a changing climate to prepare for, Working Group I Co-Chairs Valérie Masson-Delmotte and Panmao Zhai
in every region. Unless deep reductions in greenhouse gas emissions for their leadership throughout the process. Their extraordinary efforts
occur in the coming decades, global warming of 1.5°C and 2°C above ensured that this robust and clear report is available to the world today.
pre-industrial levels will be exceeded during the 21st century.
This Working Group I contribution to the IPCC’s Sixth Assessment
The new Working Group I Report structure integrates multiple lines Report covers important new advances in climate science that
of scientific evidence within each chapter and provides robust provide an invaluable input into climate negotiations and decision-
knowledge relevant for policymaking through the end-to-end making, with an emphasis on key conditions that, from a physical
assessment of key topics. This more holistic approach has resulted in science basis, are needed to limit global warming and inform risk
a report that provides a better understanding of the climate system, assessment and regional adaptation. The report was welcomed at
for both past and future changes, with a new emphasis on climate the 26th Conference of the Parties to the United Nations Framework
information for regions, which is critical for informing adaptation and Convention on Climate Change. It will also inform the 2023 global
risk management strategies. Today, many decisions remain grounded stocktake. Unless there are deep reductions in global greenhouse gas
in the experience of past climate variability. This report provides emissions, the goal of limiting warming well below 2°C and close to
a solid basis for taking into account future changes that need to be 1.5°C will be out of reach.
considered in today’s decisions, with increased relevance for climate
services, to enhance adaptation and resilience to climate change and The science is unequivocal, the changes are unprecedented, and there
curb greenhouse gas emissions. is no more time for delay.

Petteri Taalas Inger Andersen

Secretary-General Executive Director
World Meteorological Organization United Nations Environment Programme

v
 Preface

Preface

The Working Group I (WGI) contribution to the Sixth Assessment without being policy prescriptive and to facilitate the integration
Report (AR6) of the Intergovernmental Panel on Climate Change of the WGI key findings with the other AR6 Working Group reports.
(IPCC) focuses on a full and comprehensive assessment of the physical
science basis of climate change, based on evidence from more than
14,000 scientific publications available by 31 January 2021. Structure of the Report

Preface
This Report reflects recent climate science advances resulting from This Report consists of thirteen thematic chapters with their
progress in, and the integration of, multiple lines of evidence, supporting supplementary material, ten annexes (including the report
including: in situ and remote observations; paleoclimate information; Glossary, which is developed in coordination with Working Groups
understanding of climate drivers and physical, chemical and biological II and III where relevant), an integrative Technical Summary, and
processes and feedbacks; and global and regional climate modelling; a Summary for Policymakers. An innovation in this Working Group I
as well as advances in methods of analyses and insights from the assessment is the online Interactive Atlas (https://interactive-atlas.
growing field of climate services. ipcc.ch), a novel tool for flexible spatial and temporal analyses of
observed and projected climate change information, which enhances
The AR6 WGI Report builds on the WGI contribution to the IPCC’s accessibility for stakeholders and users to the data assessed in
Fifth Assessment Report (AR5) in 2013, and the AR6 Special Reports1 the Report.
released in 2018 and 2019.
The Summary for Policymakers and Technical Summary include the line
The report considers the current state of the climate in the long-term of sight, indicated in curly brackets, to the chapters and the specific
context, the understanding of human influence, the state of knowledge sections therein where the detailed assessment can be found. In this
about possible climate futures, climate information relevant for way, these summary components of the Report provide a roadmap to
climate-related risk assessment and regional adaptation, and the the contents of the entire report and a synthesis of the major findings
physical science basis on limiting human-induced climate change. that is traceable to the underlying literature and assessment.

The introduction (Chapter 1) of the Report frames the WGI

Scope of the Report assessment within the broader AR6 and global climate policy
context and introduces the key concepts, lines of evidence, and
As part of the scoping process and approval of the outline of the IPCC major developments. The remainder of the Report is structured in
Sixth Assessment Report, the WGI contribution evolved from the three segments. The first segment focuses on large-scale climate
structure of past WGI assessment reports, reflecting developments change (Chapters 2–4). These chapters provide an updated and
in climate science and the assessment of global and regional climate comprehensive assessment of knowledge about the current state
information that is relevant to inform decision-making. of the climate system, human influence and projections of future
change on key large-scale indicators of the climate system. Chapter 2
Some key new topics assessed by WGI include the global response assesses observed large-scale changes in the climate system in the
to new illustrative emissions scenarios, physical climate storylines, long-term paleoclimate context. Chapter 3 updates the assessment
low-likelihood, high-impact outcomes, and physical climate of human influence on the climate system, considering natural
conditions that affect society and/or ecosystems (defined as climatic variability, model performance and detection and attribution.
impact-drivers). Chapter 4 covers global climate projections spanning time horizons
from the near term (2021–2040) to the mid-term (2041–2060), long
The integration of multiple lines of evidence strengthens term (2081–2100) and beyond.
the understanding of past, current and possible future changes
in the climate system and the distillation of regional climate change The second segment of the report is dedicated to climate system
information at regional scale. The new structure of the WGI Report is components and processes that play key roles in global and regional
designed to facilitate this integration for end-to-end assessment of climate (Chapters 5–9), including carbon and other biogeochemical
key topics and to enhance the visibility of knowledge developments cycles, energy, and water; short-lived climate forcers (SLCFs) and
for global and regional climate change. This includes climate their link to air quality; and changes in the ocean, cryosphere
information that is relevant for risk assessment, regional adaptation and sea level. Chapter 5 addresses the assessment of the global
and for mitigation. It is also designed to inform decision-making biogeochemical budgets for carbon dioxide, methane, and nitrous

1 Global warming of 1.5°C: an IPCC special report on the impacts of global warming of 1.5°C above pre-industrial levels and related global greenhouse gas emission pathways, in the context
of strengthening the global response to the threat of climate change, sustainable development, and efforts to eradicate poverty (SR1.5); Climate Change and Land: an IPCC special report on
climate change, desertification, land degradation, sustainable land management, food security, and greenhouse gas fluxes in terrestrial ecosystems (SRCCL); IPCC Special Report on the Ocean and
Cryosphere in a Changing Climate (SROCC).

vii
 Preface

oxide and the assessment of carbon and other biogeochemical technical aspects of the assessment, such as descriptions of datasets,
feedbacks. Chapter 6 assesses changes in the emissions and models or methodologies supporting chapter analyses.
abundances of individual SLCFs, how these changes affect
Earth’s energy balance through radiative forcing and feedback in
the climate system, the implications of changing climate on air The Process
quality, and the implications of SLCF mitigation for climate and
for air quality. Chapter 7 addresses Earth’s energy budget through This WGI Assessment Report represents the combined efforts of
advances in observations, understanding and quantification of hundreds of leading experts in the field of climate science and has
effective radiative forcing, and the assessment of feedbacks and been prepared in accordance with principles and procedures of
climate sensitivity. Chapter 8 assesses observed and projected the IPCC.
Preface

changes in the global water cycle, the physical understanding of

the complexity of its response to multiple drivers, and implications A scoping meeting for the Sixth Assessment Report was held in May
for water availability. The physical processes underlying global 2017, and the outlines for the contributions of the three Working
and regional changes in the ocean, cryosphere and sea level and Groups were approved at the 46th Session of the Panel in September
the understanding of observed and projected future changes since 2017. Governments and IPCC observer organizations nominated
the AR5 and Special Report on the Ocean and Cryosphere in a experts for the author team. The team of 198 Coordinating Lead
Changing Climate are assessed in Chapter 9. Authors and Lead Authors plus 36 Review Editors selected by the
WGI Bureau was accepted at the 55th Session of the IPCC Bureau
The final four chapters of the Report (Chapters 10–12 and the Atlas) in January 2018. During the report preparation, a few changes have
are dedicated to the assessment and distillation of regional climate taken place to address core expertise gaps and to replace authors
information from multiple lines of evidence at sub-continental to who were not available anymore. In addition, 615 Contributing
local scales (including urban climate), building on information from Authors provided information to the author teams at their request.
previous chapters on large-scale climate and process understanding,
with a focus on recent and projected regional changes in mean Drafts prepared by the authors were subject to two rounds of
climate, extremes, and climatic impact-drivers. Chapter 10 assesses formal review and revision followed by a final round of government
the foundations of how regional climate information is distilled from comments on the Summary for Policymakers. A total of 78,007
multiple lines of evidence and the interplay between anthropogenic written review comments were submitted by 1891 individual expert
causes and internal variability at regional scales. Chapter 11 addresses reviewers and 47 governments. The Review Editors for each chapter
changes in weather and climate extremes on global and regional monitored the review process to ensure that all review comments
scales, including observed changes and their attribution, as well as received appropriate consideration. All review comments and
projected changes. Chapter 12 assesses the climatic conditions that responses are available online on the report website.
may lead to impacts and risks across the world’s regions. Changes
in mean climate at regional scales, in particular observed trends and Three in-person Lead Author Meetings took place to enhance progress
their attribution and projected future changes in temperature and coordination in the assessment process. These were particularly
and precipitation, are assessed in the Atlas chapter. critical for the intense cross-chapter coordination implied by the new
report outline.
The Interactive Atlas allows flexible spatial and temporal analyses
of climate variables, extreme indices and climatic impact-drivers, We strived to foster an inclusive environment that achieves a rigorous
including datasets underpinning assessment findings across report and transparent assessment process. These efforts included greater
chapters and syntheses of regional changes. attention to addressing implicit biases and technical considerations
to enhance the participation in the process and increasing the
Specific regional case studies are considered in different chapters, accessibility of the assessed information.
including rainfall changes in the Sahel and Western Africa,
South-Western North America, and South-Eastern South America; For the first time in the IPCC, WGI recommended the implementation
climate information relevant for water resources in small islands, of FAIR (findable, accessible, interoperable, reusable) data principles
for Cape Town drought, for the Indian summer monsoon, for in the assessment to document and curate the data assessed and
Mediterranean summer warming, and for the Hindu Kush Himalaya included in report figures. The motivation was to increase transparency
region; urban climate processes and trends; and the influence of the and accessibility of the assessment, support implementation of the
Arctic on mid-latitude climate. IPCC Error Protocol, and provide for the long-term curation of
the assessed digital information. This process was supported by our
All chapters contain Frequently Asked Questions, which are grounded close collaboration with the IPCC Task Group on Data Support for
in the assessment and written in language that is more accessible Climate Change Assessments (TG-Data) established in March 2018.
to a broad readership and can serve as a resource for teaching
and outreach. The COVID-19 pandemic was declared by the World Health
Organization at the start the Second Order Draft review. Following
Each chapter is accompanied by Supplementary Material, which an extensive consultation process with IPCC member countries,
is available online and provides traceability and transparency for authors and the scientific community, including editors of relevant

viii
 Preface

science journals, the timeline of the report was extended by four The Summary for Policymakers was approved line-by-line during the
months to balance the delays and challenges faced by authors first-ever virtual IPCC approval session, the 14th Session of IPCC
and the scientific community with maintaining momentum and Working Group I from 26 July – 06 August 2021, and the underlying
timeliness in the assessment process. The review process is critical Report was accepted during the 54th Session of the IPCC on the
for the rigor, objectivity and comprehensiveness of the assessment. 6th August 2021.
The adjustments to the timeline facilitated broad participation from
scientists and governments in the review process. Stringent scientific
rigor and quality of the assessment were maintained despite Acknowledgements
the pandemic.
We are very grateful for the exceptional rigor and dedication of the

Preface
The fourth Lead Author Meeting, due to be held in June 2020, volunteer Coordinating Lead Authors and Lead Authors throughout
was replaced by extensive virtual activities to address the Second this process, who delivered the most comprehensive assessment
Order Draft review comments and topics that cut across multiple ever of our physical understanding of climate change. We thank the
chapters. A final virtual Lead Author Meeting was held in February Review Editors for working alongside the author teams to ensure
2021 to finalize the report. Drafting meetings for the Summary for that the chapters are fully reflective of the input provided through
Policymakers also took place through virtual meetings. the review process. We express our sincere appreciation to all the
government and expert reviewers, including several group reviews
Addressing the implications of the COVID-19 pandemic on the from early-career scientists. We thank the many Contributing Authors
assessment process required innovation to facilitate international who provided input and important support to the authors.
virtual collaboration, including extra support and training for
participants and facilitators, support for participants with internet A special thanks goes to the Chapter Scientists of this report who
connectivity challenges, additional advance preparation, shorter went above and beyond what was expected of them: Kari Alterskjaer,
and more focused meetings with clear agendas and objectives and Lisa Bock, Katherine Dooley, Gregory Garner, Mathias Hauser, Tim
duplicated to account for time zones of participants, high levels Hermans, Lijuan Hua, Carley Iles, Maialen Iturbide, Laurice Preciado
of transparency – including the provision of written summaries Jamero, Martin Jury, Megan Kirchmeier Young, Chaincy Kuo, Hui-Wen
of meetings and decisions, and allotting time for asynchronous Lai, Alice Lebehot, Elizaveta Malinina Rieger, Sebastian Milinkski,
contributions to the discussion and decision-making process. Therese Myslinski, Tamzin Palmer, Browdie Pearson, Stephane Senesi,
Jérôme Servonnat, Chris Smith, David Smyth, Sabin Thazhe Purayil,
Documenting and understanding barriers to participation due to Emilie Vanvyve, Tania Villasenor Jorquera, Hui Wan and Kyung-
an increased reliance on online activities and the use of inclusive Sook Yu. Chapter scientists were recruited by and reported directly
practices required priority attention in these novel conditions. We to the Coordinating Lead Author(s) and provided technical support
gained experience in applying methods to facilitate participatory and to the chapters, including reference checking and compilation,
inclusive processes in the assessment and recognized the necessity figure drafting, traceability checking, identification of overlaps or
of fostering these approaches over the course of the assessment inconsistencies across chapters, and technical editing.
process, both during and in-between meetings, to build a stronger
community of practice within this unique international context. This We thank the Vice Chairs of the WGI Bureau for their dedication,
will be an important legacy for future assessment cycles. guidance and wisdom throughout the preparation of the Report and
their support for cross-Working Group coordination: Edvin Aldrian,
The preparation of the WGI AR6 report was also informed by Fatima Driouech, Gregory Flato, Jan Fuglestvedt, Muhammad I. Tariq,
recommendations from several IPCC expert meetings. The first Carolina Vera and Noureddine Yassaa.
meeting focused on assessing climate information for regions
(in 2018, co-organized by WGI and WGII, and hosted at ICTP, Trieste, We gratefully acknowledge the support from the host countries
Italy), which provided a scoping of the Interactive Atlas. A second and institutions of the WGI Lead Author Meetings (LAMs): China
meeting focused on short-lived climate forcers (in 2018, co-organized Meteorological Administration (CMA), China, for hosting the first
by TFI and WGI, and hosted by the World Meteorological Organization LAM; Environment Canada, Canada, for hosting the second LAM; and
in Geneva, Switzerland), which identified science advances in the Météo France, France, for hosting the third LAM. We also thank the
understanding of emissions and climate effects of SLCFs and needs Ministerio de Ciencia, Tecnologia, Conocimiento e Innovación and
for improvements in emission inventories and methodologies. A the Ministerio del Medio Ambiente, Chile, for offering to host the
third meeting on mitigation, sustainability, and climate stabilization fourth LAM meeting that we were unable to hold in person due to
scenarios (in 2019, organized by WGIII in Addis Ababa, Ethiopia), led the COVID-19 pandemic.
to cross-WG coordination related to scenarios. Recommendations
for clarity and readability from the 2016 IPCC Expert Meeting The support provided by governments and institutions, as well
on Communication (organized by the IPCC Secretariat in Oslo, as through contributions to the IPCC Trust Fund, is thankfully
Norway) were taken into account in developing technical guidance, acknowledged as it enabled the participation of the author
training and resources provided to authors and in particular for the teams in the preparation of the Report. The efficient operation of
preparation of the text and figures of the Frequently Asked Questions the WGI Technical Support Unit (TSU) was made possible by the
and the Summary for Policymakers. generous financial support provided by the government of France

ix
 Preface

and administrative and information technology support from the Lynn, Judith Ewa, Mxolisi Shongwe, Jennifer Lew Schneider, Jesbin
Université Paris Saclay (France), Institut Pierre Simon Laplace (IPSL) Baidya, Werani Zabula, Nina Peeva, Melissa Walsh, Joelle Fernandez,
and the Laboratoire des Sciences du Climat et de l’Environnement Laura Biagioni and Oksana Ekzarkho for their guidance and support
(LSCE). We thank the Norwegian Environment Agency for supporting to implement the many facets of the IPCC process. We were grateful
the preparation of the graphics for the Summary for Policymakers. that we could work together extensively on communication activities.
We thank the United Nations Environment Programme Library for We thank Sue Escott for her tireless work to strengthen how we
providing a service for authors to access literature. communicate the outcomes of the assessment.

The approval of the WGI Summary for Policymakers took place in We would like to acknowledge the support of the SHIFT Collaborative
an unprecedented context, with travel restrictions caused by the team – Stacy Barter and Michelle Colussi – and the generous
Preface

COVID-19 pandemic rendering an in-person IPCC Plenary session support of the Canadian government for training and tools on
impossible. We thank the support and advice of the IPCC Executive inclusive practices in a consensus-based decision-making context,
Committee and the tireless work of the ad-hoc task group that was which we have been able to use for a more inclusive assessment
established to advise us as Co-Chairs in preparing for the approval process, including when we moved to purely online approaches. We
session. The task group was led by IPCC Vice-Chair Ko Barrett and appreciated the presence of Jessica O’Reilly and Mark Vardy, who
included WGI Vice-Chairs Fatima Driouech, Greg Flato and Edvin have been with us throughout, working on an ethnographic study of
Aldrian; Anna Pirani and Sarah Connors of the WGI TSU; and Ermira how authors undertake the IPCC assessment, and we look forward to
Fida of the IPCC Secretariat. The task group prepared guidance for the insights from their research.
participants on the modalities of the session and a carefully structured
meeting schedule to implement a virtual approval process. A core outcome of the report has been the development of the WGI
Interactive Atlas produced by the Atlas chapter team. The Atlas is
The approval took place virtually for the first time and involved more dedicated to the memory of Gemma Teresa Narisma, who co-led this
than 186 hours of online meetings. We thank all participants for the innovative chapter with her extensive experience in regional climate
remarkable collaborative spirit and rigorous work undertaken during research and outstanding leadership. The development and technical
the session. IPCC Vice Chairs Ko Barrett, Thelma Krug and Youba implementation have been supported with in-kind contribution
Sokona brought their unwavering support to facilitate discussions from the Spanish government through the Spanish Research
amongst authors and delegations and provided core support for the Council (CSIC) Instituto de Física de Cantabria, in partnership with
success of the approval process. We are also grateful to the Vice Chairs Predictia Intelligent Data Solutions. Funding from the Spanish
of WGI, as well as Mark Howden and Andy Reisinger, Vice Chairs of Research & Development program is acknowledged (ref. PID2019-
WGII and WGIII respectively, and Jim Skea, Co-Chair of WGIII, for 111481RB-I00). We thank the modelling centres and institutions
their support to facilitate the discussions. The WGI TSU was joined that produce and make available the datasets used in this work. The
by members of the WGII and WGIII TSUs, as well as past interns and Interactive Atlas was first made available along with the rest of the
chapter scientists to staff this Herculean coordination effort. Report on 9 August 2021 and was visited by more than half a million
users worldwide during the first month.
Our warmest thanks go to the collegial and collaborative support
provided by Melinda Tignor, Elvira Poloczanska, Katja Mintenbeck, The WGI TSU has initiated the process to archive the data and
Bard Rama, Almut Niebuhr, Vincent Möller, Sina Löschke, Komila code from the report, building on the guidance and support from
Nabiyeva, Andrés Alegría, Stefanie Langsdorf, Andrew Okem, a large group of contributors. We are indebted to the members
Marlies Craig, Anka Mühle, Philisiwe Manqele, Stefan Weisfeld, of the IPCC Task Group on Data Support for Climate Change
Jussi Savolainen and Mallou from the Working Group II Technical Assessments (TG-Data) for their oversight, expert guidance and
Support Unit; Roger Fradera, Raphael Slade, Alaa Al Khourdajie, Minal constant encouragement, including the Co-chairs of the Task Group,
Pathak, Sigourney Luz, Malek Belkacemi, David McCollum, Renée van David Huard and Sebastian Vicuna, and members representing the
Diemen, Shreya Some, Purvi Vyas, Juliette Malley and Géninha Lisboa WGI science community, Michio Kawamiya, Silvina Solman, José
from the Working Group III Technical Support Unit; and Noémie Manuel Guttierez and Nana Ama Browne Klutse. For the preparation
Le Prince-Ringuet from the Synthesis Report Technical Support Unit. of the figure data and code for archival, we especially thank to our
dedicated contractor, Lina Sitz.
We are grateful for the close collaboration with authors and Bureau
members from Working Group II and III, including as contributing The IPCC Data Distribution Centre (DDC) has been indispensable
authors in many parts of the report. We thank the Co-Chairs Debra for this effort. For the archival of figure data, we are indebted to
Roberts, Hans-Otto Portner, Jim Skea and Priyadarshi R. Shukla for Charlotte Pascoe, Kate Winfield, and Martin Jukes from the UK
the collegial teamwork across Working Groups that has characterized Centre for Environmental Data Analysis (CEDA). For the archival of
the AR6. We also thank Eduardo Calvo Buendía and Kyoto Tanabe, the climate model data used as input to the report and intermediate
Co-Chairs of the Task Force on Greenhouse Gas Inventories, for their assessed datasets, we gratefully acknowledge Martina Stockhause
support and collaboration. of the German Climate Computing Centre (DKRZ). For the transfer
of metadata on archived data/code into the IPCC data catalogue, we
We thank Hoesung Lee, Chair of the IPCC, Abdalah Mokssit, Secretary thank MetadataWorks. Finally, we gratefully acknowledge funding
of the IPCC, and the staff of the IPCC Secretariat: Ermira Fida, Jonathan

x
 Preface

support from the Governments of the United Kingdom and Germany, We thank our past WGI TSU team members: Wilfran Moufouma-Okia,
without which data archival at the DDC would not have been possible. Roz Pidcock, and Rodrigro Manzanas. We also thank the contributions
of Margot Eyraud, Evéa Piedagnel, Mathilde Mousson and Felix
A special thanks goes to the visual design team of the Summary Chavelli, who joined the TSU as interns.
for Policymakers: Tom Johansen and Angela Morelli of Information
Design Lab and Jordan Harold and Irene Lorenzoni of the Tyndall We wish to express our sincere recognition to all those who
Centre for Climate Change Research, as well as to Nigel Hawtin for contributed to the WGI assessment given the implications of
graphical design support of the Report. We would like to thank Alisa undertaking this during the COVID-19 pandemic, all of whom have
Singer for creating the “Changing” artwork inspired by one of the worked from home under such challenging conditions.
scientific figures for the front cover of the Report.

Preface
Finally, on behalf of all the participants to this unprecedented
Our particular appreciation goes to the WGI TSU, whose tireless experience, we would like to thank colleagues, friends, and
dedication, professionalism and enthusiasm underpinned the families who have also been part of this intense journey for their
production of this Report. The Report could not have been prepared understanding and support.
without the commitment of members of the TSU, all new to the IPCC,
who rose to the unprecedented Sixth Assessment Report challenge This report shows that how much climate change we experience in
and were pivotal in all aspects of the preparation of the Report: Anna the future depends on our decisions now, and what to prepare for.
Pirani, Clotilde Péan, Sarah Connors, Yang Chen, Robin Matthews, We wish that this report is widely used to provide evidence-based
Melissa Gomis, Sophie Berger, Leah Goldfarb, Rong Yu, Baiquan Zhou, knowledge to inform decision-making, for teaching and training, and
Ozge Yelekci, Nada Caud, Katherine Leitzell, Tom Maycock, Mengtian to enhance climate literacy worldwide.
Huang, Elisabeth Lonnoy, Tim Waterfield and Diego Cammarano.

Valérie Masson-Delmotte Panmao Zhai

IPCC Working Group I Co-Chair IPCC Working Group I Co-Chair

xi
 Dedication

Dedication

Sir John Hougthon

(30 December 1931 – 15 April 2020)

Dedication
Preface
The Working Group I Contribution to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC) Climate
Change 2021: The Physical Science Basis is dedicated to the memory of Sir John Hougthon, who was one of the key figures in the
creation of the IPCC in 1988, and served as Chair and Co-Chair of Working Group I for the IPCC’s first three assessment reports from
1988 to 2002.

Sir John’s work was a major factor in the award of the Nobel Peace Prize to the IPCC in 2007, shared with former U.S. Vice-President
Al Gore. He contributed to the development of climate science and building international cooperation based upon climate research.
Sir John played a key role in ensuring a robust science-policy interface, used in the IPCC process, but his role in international scientific
research extended beyond the IPCC, for instance in contributing to the establishment of the World Climate Research Programme,
which he chaired from 1982 to 1984.

Sir John was a brilliant communicator among scientific colleagues, policymakers and the public at large, explaining the fact and threat
of climate change with clarity and directness.

xiii
Contents

Front Matter Foreword ��������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������� v

Preface ��������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������� vii

Dedication ���������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������� xiii

SPM Summary for Policymakers ������������������������������������������������������������������������������������������������������������������������������������������������������������������� 3

TS Technical Summary ���������������������������������������������������������������������������������������������������������������������������������������������������������������������������������� 35

Chapters Chapter 1 Framing, Context, and Methods ���������������������������������������������������������������������������������������������������������������������������� 147

Chapter 2 Changing State of the Climate System ������������������������������������������������������������������������������������������������������������ 287

Chapter 3 Human Influence on the Climate System �������������������������������������������������������������������������������������������������������� 423

Chapter 4 Future Global Climate: Scenario-based Projections and Near-term Information ������������������������ 553

Chapter 5 Global Carbon and Other Biogeochemical Cycles and Feedbacks ������������������������������������������������������ 673

Chapter 6 Short-lived Climate Forcers �������������������������������������������������������������������������������������������������������������������������������������� 817

Chapter 7 The Earth’s Energy Budget, Climate Feedbacks and Climate Sensitivity ������������������������������������������ 923

Chapter 8 Water Cycle Changes ����������������������������������������������������������������������������������������������������������������������������������������������� 1055

Chapter 9 Ocean, Cryosphere and Sea Level Change ��������������������������������������������������������������������������������������������������� 1211

Chapter 10 Linking Global to Regional Climate Change ����������������������������������������������������������������������������������������������� 1363

Chapter 11 Weather and Climate Extreme Events in a Changing Climate ����������������������������������������������������������� 1513

Chapter 12 Climate Change Information for Regional Impact and for Risk Assessment ����������������������������� 1767

Atlas ����������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������� 1927

Annexes Annex I Observational Products ������������������������������������������������������������������������������������������������������������������������������������������� 2061

Annex II Models ������������������������������������������������������������������������������������������������������������������������������������������������������������������������������ 2087

Annex III Tables of Historical and Projected Well-mixed Greenhouse Gas Mixing Ratios
and Effective Radiative Forcing of All Climate Forcers ���������������������������������������������������������������������������� 2139

Annex IV Modes of Variability ���������������������������������������������������������������������������������������������������������������������������������������������������� 2153

Annex V Monsoons ������������������������������������������������������������������������������������������������������������������������������������������������������������������������ 2193

Annex VI Climatic Impact-driver and Extreme Indices �������������������������������������������������������������������������������������������������� 2205

Annex VII Glossary ���������������������������������������������������������������������������������������������������������������������������������������������������������������������������� 2215

Annex VIII Acronyms ������������������������������������������������������������������������������������������������������������������������������������������������������������������������ 2257

Annex IX Contributors to the IPCC WGI Sixth Assessment Report ������������������������������������������������������������������������ 2267

Annex X Expert Reviewers of the IPCC Sixth Assessment Report �������������������������������������������������������������������������� 2287

Index ����������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������� 2339

Summary for
Policymakers
 Summary for
Policymakers
Drafting Authors:
Richard P. Allan (United Kingdom), Paola A. Arias (Colombia), Sophie Berger (France/Belgium), Josep G.
Canadell (Australia), Christophe Cassou (France), Deliang Chen (Sweden), Annalisa Cherchi (Italy), Sarah

SPM
L. Connors (France/United Kingdom), Erika Coppola (Italy), Faye Abigail Cruz (Philippines), Aïda Diongue-
Niang (Senegal), Francisco J. Doblas-Reyes (Spain), Hervé Douville (France), Fatima Driouech (Morocco),
Tamsin L. Edwards (United Kingdom), François Engelbrecht (South Africa), Veronika Eyring (Germany),
Erich Fischer (Switzerland), Gregory M. Flato (Canada), Piers Forster (United Kingdom), Baylor Fox-
Kemper (United States of America), Jan S. Fuglestvedt (Norway), John C. Fyfe (Canada), Nathan P. Gillett
(Canada), Melissa I. Gomis (France/Switzerland), Sergey K. Gulev (Russian Federation), José Manuel
Gutiérrez (Spain), Rafiq Hamdi (Belgium), Jordan Harold (United Kingdom), Mathias Hauser (Switzerland),
Ed Hawkins (United Kingdom), Helene T. Hewitt (United Kingdom), Tom Gabriel Johansen (Norway),
Christopher Jones (United Kingdom), Richard G. Jones (United Kingdom), Darrell S. Kaufman (United
States of America), Zbigniew Klimont (Austria/Poland), Robert E. Kopp (United States of America), Charles
Koven (United States of America), Gerhard Krinner (France/Germany, France), June-Yi Lee (Republic of
Korea), Irene Lorenzoni (United Kingdom/Italy), Jochem Marotzke (Germany), Valérie Masson-Delmotte
(France), Thomas K. Maycock (United States of America), Malte Meinshausen (Australia/Germany), Pedro
M.S. Monteiro (South Africa), Angela Morelli (Norway/Italy), Vaishali Naik (United States of America),
Dirk Notz (Germany), Friederike Otto (United Kingdom/Germany), Matthew D. Palmer (United Kingdom),
Izidine Pinto (South Africa/Mozambique), Anna Pirani (Italy), Gian-Kasper Plattner (Switzerland),
Krishnan Raghavan (India), Roshanka Ranasinghe (The Netherlands/Sri Lanka, Australia), Joeri Rogelj
(United Kingdom/Belgium), Maisa Rojas (Chile), Alex C. Ruane (United States of America), Jean-Baptiste
Sallée (France), Bjørn H. Samset (Norway), Sonia I. Seneviratne (Switzerland), Jana Sillmann (Norway/
Germany), Anna A. Sörensson (Argentina), Tannecia S. Stephenson (Jamaica), Trude Storelvmo (Norway),
Sophie Szopa (France), Peter W. Thorne (Ireland/United Kingdom), Blair Trewin (Australia), Robert Vautard
(France), Carolina Vera (Argentina), Noureddine Yassaa (Algeria), Sönke Zaehle (Germany), Panmao Zhai
(China), Xuebin Zhang (Canada), Kirsten Zickfeld (Canada/Germany)

Contributing Authors:
Krishna M. AchutaRao (India), Bhupesh Adhikary (Nepal), Edvin Aldrian (Indonesia), Kyle Armour (United
States of America), Govindasamy Bala (India/United States of America), Rondrotiana Barimalala (South
Africa/Madagascar), Nicolas Bellouin (United Kingdom/France), William Collins (United Kingdom),
William D. Collins (United States of America), Susanna Corti (Italy), Peter M. Cox (United Kingdom),
Frank J. Dentener (EU/The Netherlands), Claudine Dereczynski (Brazil), Alejandro Di Luca (Australia,
Canada/Argentina), Alessandro Dosio (Italy), Leah Goldfarb (France/United States of America), Irina
V. Gorodetskaya (Portugal/Belgium, Russian Federation), Pandora Hope (Australia), Mark Howden
(Australia), A.K.M Saiful Islam (Bangladesh), Yu Kosaka (Japan), James Kossin (United States of America),
Svitlana Krakovska (Ukraine), Chao Li (China), Jian Li (China), Thorsten Mauritsen (Germany/Denmark),
Sebastian Milinski (Germany), Seung-Ki Min (Republic of Korea), Thanh Ngo Duc (Vietnam), Andy
Reisinger (New Zealand), Lucas Ruiz (Argentina), Shubha Sathyendranath (United Kingdom/Canada,
Overseas Citizen of India), Aimée B. A. Slangen (The Netherlands), Chris Smith (United Kingdom), Izuru
Takayabu (Japan), Muhammad Irfan Tariq (Pakistan), Anne-Marie Treguier (France), Bart van den Hurk
(The Netherlands), Karina von Schuckmann (France/Germany), Cunde Xiao (China)

This Summary for Policymakers should be cited as:

IPCC, 2021: Summary for Policymakers. In: Climate Change 2021: The Physical Science Basis. Contribution of Working Group I
to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change [Masson-Delmotte, V., P. Zhai, A. Pirani, S.L.
Connors, C. Péan, S. Berger, N. Caud, Y. Chen, L. Goldfarb, M.I. Gomis, M. Huang, K. Leitzell, E. Lonnoy, J.B.R. Matthews, T.K.
Maycock, T. Waterfield, O. Yelekçi, R. Yu, and B. Zhou (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New
York, NY, USA, pp. 3−32, doi:10.1017/9781009157896.001.

3
 Summary for Policymakers

Introduction
This Summary for Policymakers (SPM) presents key findings of the Working Group I (WGI) contribution to the Intergovernmental
Panel on Climate Change (IPCC) Sixth Assessment Report (AR6)1 on the physical science basis of climate change. The report builds
upon the 2013 Working Group I contribution to the IPCC’s Fifth Assessment Report (AR5) and the 2018–2019 IPCC Special Reports2
of the AR6 cycle and incorporates subsequent new evidence from climate science.3
SPM
This SPM provides a high-level summary of the understanding of the current state of the climate, including how it is changing and the
role of human influence, the state of knowledge about possible climate futures, climate information relevant to regions and sectors,
and limiting human-induced climate change.

Based on scientific understanding, key findings can be formulated as statements of fact or associated with an assessed level of
confidence indicated using the IPCC calibrated language.4

The scientific basis for each key finding is found in chapter sections of the main Report and in the integrated synthesis presented
in the Technical Summary (hereafter TS), and is indicated in curly brackets. The AR6 WGI Interactive Atlas facilitates exploration of
these key synthesis findings, and supporting climate change information, across the WGI reference regions.5

A. The Current State of the Climate

Since AR5, improvements in observationally based estimates and information from paleoclimate archives provide a comprehensive
view of each component of the climate system and its changes to date. New climate model simulations, new analyses, and methods
combining multiple lines of evidence lead to improved understanding of human influence on a wider range of climate variables,
including weather and climate extremes. The time periods considered throughout this section depend upon the availability of
observational products, paleoclimate archives and peer-reviewed studies.

A.1 It is unequivocal that human influence has warmed the atmosphere, ocean and land. Widespread and rapid
changes in the atmosphere, ocean, cryosphere and biosphere have occurred.
{2.2, 2.3, Cross-Chapter Box 2.3, 3.3, 3.4, 3.5, 3.6, 3.8, 5.2, 5.3, 6.4, 7.3, 8.3, 9.2, 9.3, 9.5, 9.6, Cross-Chapter
Box 9.1} (Figure SPM.1, Figure SPM.2)

A.1.1 Observed increases in well-mixed greenhouse gas (GHG) concentrations since around 1750 are unequivocally caused
by human activities. Since 2011 (measurements reported in AR5), concentrations have continued to increase in the
atmosphere, reaching annual averages of 410 parts per million (ppm) for carbon dioxide (CO2), 1866 parts per billion
(ppb) for methane (CH4), and 332 ppb for nitrous oxide (N2O) in 2019.6 Land and ocean have taken up a near-constant
proportion (globally about 56% per year) of CO2 emissions from human activities over the past six decades, with regional
differences (high confidence).7
{2.2, 5.2, 7.3, TS.2.2, Box TS.5}

1 Decision IPCC/XLVI-2.
2 The three Special Reports are: Global Warming of 1.5°C: An IPCC Special Report on the impacts of global warming of 1.5°C above pre-industrial levels and related global greenhouse
gas emission pathways, in the context of strengthening the global response to the threat of climate change, sustainable development, and efforts to eradicate poverty (SR1.5);
Climate Change and Land: An IPCC Special Report on climate change, desertification, land degradation, sustainable land management, food security, and greenhouse gas fluxes in
terrestrial ecosystems (SRCCL); IPCC Special Report on the Ocean and Cryosphere in a Changing Climate (SROCC).
3 The assessment covers scientific literature accepted for publication by 31 January 2021.
4 Each finding is grounded in an evaluation of underlying evidence and agreement. A level of confidence is expressed using five qualifiers: very low, low, medium, high and very high,
and typeset in italics, for example, medium confidence. The following terms have been used to indicate the assessed likelihood of an outcome or result: virtually certain 99–100%
probability; very likely 90–100%; likely 66–100%; about as likely as not 33–66%; unlikely 0–33%; very unlikely 0–10%; and exceptionally unlikely 0–1%. Additional terms
(extremely likely 95–100%; more likely than not >50–100%; and extremely unlikely 0–5%) are also used when appropriate. Assessed likelihood is typeset in italics, for example,
very likely. This is consistent with AR5. In this Report, unless stated otherwise, square brackets [x to y] are used to provide the assessed very likely range, or 90% interval.
5 The Interactive Atlas is available at https://interactive-atlas.ipcc.ch
6 Other GHG concentrations in 2019 were: perfluorocarbons (PFCs) – 109 parts per trillion (ppt) CF4 equivalent; sulphur hexafluoride (SF6) – 10 ppt; nitrogen trifluoride (NF3) – 2 ppt;
hydrofluorocarbons (HFCs) – 237 ppt HFC-134a equivalent; other Montreal Protocol gases (mainly chlorofluorocarbons (CFCs) and hydrochlorofluorocarbons (HCFCs)) – 1032 ppt
CFC-12 equivalent). Increases from 2011 are 19 ppm for CO2, 63 ppb for CH4 and 8 ppb for N2O.
7 Land and ocean are not substantial sinks for other GHGs.

4
 Summary for Policymakers

A.1.2 Each of the last four decades has been successively warmer than any decade that preceded it since 1850. Global
surface temperature8 in the first two decades of the 21st century (2001–2020) was 0.99 [0.84 to 1.10] °C higher than
1850–1900.9 Global surface temperature was 1.09 [0.95 to 1.20] °C higher in 2011–2020 than 1850–1900, with larger
increases over land (1.59 [1.34 to 1.83] °C) than over the ocean (0.88 [0.68 to 1.01] °C). The estimated increase in
global surface temperature since AR5 is principally due to further warming since 2003–2012 (+0.19 [0.16 to 0.22] °C).
Additionally, methodological advances and new datasets contributed approximately 0.1°C to the updated estimate of
warming in AR6.10 SPM
{2.3, Cross-Chapter Box 2.3} (Figure SPM.1)

A.1.3 The likely range of total human-caused global surface temperature increase from 1850–1900 to 2010–201911 is 0.8°C to
1.3°C, with a best estimate of 1.07°C. It is likely that well-mixed GHGs contributed a warming of 1.0°C to 2.0°C, other
human drivers (principally aerosols) contributed a cooling of 0.0°C to 0.8°C, natural drivers changed global surface
temperature by –0.1°C to +0.1°C, and internal variability changed it by –0.2°C to +0.2°C. It is very likely that well-mixed
GHGs were the main driver12 of tropospheric warming since 1979 and extremely likely that human-caused stratospheric
ozone depletion was the main driver of cooling of the lower stratosphere between 1979 and the mid-1990s.
{3.3, 6.4, 7.3, TS.2.3, Cross-Section Box TS.1} (Figure SPM.2)

A.1.4 Globally averaged precipitation over land has likely increased since 1950, with a faster rate of increase since the 1980s
(medium confidence). It is likely that human influence contributed to the pattern of observed precipitation changes
since the mid-20th century and extremely likely that human influence contributed to the pattern of observed changes
in near-surface ocean salinity. Mid-latitude storm tracks have likely shifted poleward in both hemispheres since the
1980s, with marked seasonality in trends (medium confidence). For the Southern Hemisphere, human influence very likely
contributed to the poleward shift of the closely related extratropical jet in austral summer.
{2.3, 3.3, 8.3, 9.2, TS.2.3, TS.2.4, Box TS.6}

A.1.5 Human influence is very likely the main driver of the global retreat of glaciers since the 1990s and the decrease in Arctic
sea ice area between 1979–1988 and 2010–2019 (decreases of about 40% in September and about 10% in March). There
has been no significant trend in Antarctic sea ice area from 1979 to 2020 due to regionally opposing trends and large
internal variability. Human influence very likely contributed to the decrease in Northern Hemisphere spring snow cover
since 1950. It is very likely that human influence has contributed to the observed surface melting of the Greenland Ice
Sheet over the past two decades, but there is only limited evidence, with medium agreement, of human influence on the
Antarctic Ice Sheet mass loss.
{2.3, 3.4, 8.3, 9.3, 9.5, TS.2.5}

A.1.6 It is virtually certain that the global upper ocean (0–700 m) has warmed since the 1970s and extremely likely that human
influence is the main driver. It is virtually certain that human-caused CO2 emissions are the main driver of current global
acidification of the surface open ocean. There is high confidence that oxygen levels have dropped in many upper ocean
regions since the mid-20th century and medium confidence that human influence contributed to this drop.
{2.3, 3.5, 3.6, 5.3, 9.2, TS.2.4}

A.1.7 Global mean sea level increased by 0.20 [0.15 to 0.25] m between 1901 and 2018. The average rate of sea level rise was
1.3 [0.6 to 2.1] mm yr–1 between 1901 and 1971, increasing to 1.9 [0.8 to 2.9] mm yr–1 between 1971 and 2006, and
further increasing to 3.7 [3.2 to 4.2] mm yr–1 between 2006 and 2018 (high confidence). Human influence was very likely
the main driver of these increases since at least 1971.
{2.3, 3.5, 9.6, Cross-Chapter Box 9.1, Box TS.4}

8 The term ‘global surface temperature’ is used in reference to both global mean surface temperature and global surface air temperature throughout this SPM. Changes in these
quantities are assessed with high confidence to differ by at most 10% from one another, but conflicting lines of evidence lead to low confidence in the sign (direction) of any
difference in long-term trend. {Cross-Section Box TS.1}
9 The period 1850–1900 represents the earliest period of sufficiently globally complete observations to estimate global surface temperature and, consistent with AR5 and SR1.5, is
used as an approximation for pre-industrial conditions.
10 Since AR5, methodological advances and new datasets have provided a more complete spatial representation of changes in surface temperature, including in the Arctic. These
and other improvements have also increased the estimate of global surface temperature change by approximately 0.1°C, but this increase does not represent additional physical
warming since AR5.
11 The period distinction with A.1.2 arises because the attribution studies consider this slightly earlier period. The observed warming to 2010–2019 is 1.06 [0.88 to 1.21] °C.
12 Throughout this SPM, ‘main driver’ means responsible for more than 50% of the change.

5
 Summary for Policymakers

A.1.8 Changes in the land biosphere since 1970 are consistent with global warming: climate zones have shifted poleward in
both hemispheres, and the growing season has on average lengthened by up to two days per decade since the 1950s
in the Northern Hemisphere extratropics (high confidence).
{2.3, TS.2.6}

SPM

Human influence has warmed the climate at a rate that is unprecedented

in at least the last 2000 years

Changes in global surface temperature relative to 1850–1900

(a) Change in global surface temperature (decadal average) (b) Change in global surface temperature (annual average) as observed and
as reconstructed (1–2000) and observed (1850–2020) simulated using human & natural and only natural factors (both 1850–2020)
ºC ºC
2.0 2.0
Warming is unprecedented
in more than 2000 years
1.5 1.5
Warmest multi-century observed
period in more than
100,000 years simulated
1.0 1.0 1.0 human &
observed natural

0.5 0.5

0.2 simulated
natural only
0.0 0.0 (solar &
volcanic)
reconstructed
–0.5 –0.5

–1
1 500 1000 1500 1850 2020 1850 1900 1950 2000 2020

Figure SPM.1 | History of global temperature change and causes of recent warming
Panel (a) Changes in global surface temperature reconstructed from paleoclimate archives (solid grey line, years 1–2000) and from direct
observations (solid black line, 1850–2020), both relative to 1850–1900 and decadally averaged. The vertical bar on the left shows the estimated temperature
(very likely range) during the warmest multi-century period in at least the last 100,000 years, which occurred around 6500 years ago during the current interglacial
period (Holocene). The Last Interglacial, around 125,000 years ago, is the next most recent candidate for a period of higher temperature. These past warm periods
were caused by slow (multi-millennial) orbital variations. The grey shading with white diagonal lines shows the very likely ranges for the temperature reconstructions.
Panel (b) Changes in global surface temperature over the past 170 years (black line) relative to 1850–1900 and annually averaged, compared to
Coupled Model Intercomparison Project Phase 6 (CMIP6) climate model simulations (see Box SPM.1) of the temperature response to both human and natural
drivers (brown) and to only natural drivers (solar and volcanic activity, green). Solid coloured lines show the multi-model average, and coloured shades show the
very likely range of simulations. (See Figure SPM.2 for the assessed contributions to warming).
{2.3.1; Cross-Chapter Box 2.3; 3.3; TS.2.2; Cross-Section Box TS.1, Figure 1a}

6
 Summary for Policymakers

Observed warming is driven by emissions from human activities, with

greenhouse gas warming partly masked by aerosol cooling

Observed warming Contributions to warming based on two complementary approaches

(a) Observed warming (b) Aggregated contributions to (c) Contributions to 2010–2019
SPM
2010–2019 relative to 2010–2019 warming relative to warming relative to 1850–1900,
1850–1900 1850–1900, assessed from assessed from radiative
ºC attribution studies ºC forcing studies ºC
2.0 2.0 2.0

1.5 1.5 1.5

1.0 1.0 1.0

0.5 0.5 0.5

0.0 0.0 0.0

–0.5 –0.5 –0.5

–1.0 –1.0 –1.0

and carbon monoxide

Volatile organic compounds
Total human influence

Well-mixed greenhouse gases

Other human drivers

Solar and volcanic drivers

Internal variability

Carbon dioxide

Methane

Nitrous oxide

Halogenated gases

Nitrogen oxides

Sulphur dioxide

Organic carbon

Ammonia

Black carbon

and irrigation
Land-use reflectance
Aviation contrails
Mainly contribute to Mainly contribute to
changes in changes in
non-CO₂ greenhouse gases anthropogenic aerosols

Figure SPM.2 | Assessed contributions to observed warming in 2010–2019 relative to 1850–1900

Panel (a) Observed global warming (increase in global surface temperature). Whiskers show the very likely range.
Panel (b) Evidence from attribution studies, which synthesize information from climate models and observations. The panel shows temperature
change attributed to: total human influence; changes in well-mixed greenhouse gas concentrations; other human drivers due to aerosols, ozone and land-use
change (land-use reflectance); solar and volcanic drivers; and internal climate variability. Whiskers show likely ranges.
Panel (c) Evidence from the assessment of radiative forcing and climate sensitivity. The panel shows temperature changes from individual components
of human influence: emissions of greenhouse gases, aerosols and their precursors; land-use changes (land-use reflectance and irrigation); and aviation contrails.
Whiskers show very likely ranges. Estimates account for both direct emissions into the atmosphere and their effect, if any, on other climate drivers. For aerosols,
both direct effects (through radiation) and indirect effects (through interactions with clouds) are considered.
{Cross-Chapter Box 2.3, 3.3.1, 6.4.2, 7.3}

7
 Summary for Policymakers

A.2 The scale of recent changes across the climate system as a whole – and the present state of many aspects of
the climate system – are unprecedented over many centuries to many thousands of years.
{2.2, 2.3, Cross-Chapter Box 2.1, 5.1} (Figure SPM.1)

A.2.1 In 2019, atmospheric CO2 concentrations were higher than at any time in at least 2 million years (high confidence), and
concentrations of CH4 and N2O were higher than at any time in at least 800,000 years (very high confidence). Since 1750,
SPM increases in CO2 (47%) and CH4 (156%) concentrations far exceed – and increases in N2O (23%) are similar to – the natural
multi-millennial changes between glacial and interglacial periods over at least the past 800,000 years (very high confidence).
{2.2, 5.1, TS.2.2}

A.2.2 Global surface temperature has increased faster since 1970 than in any other 50-year period over at least the last 2000
years (high confidence). Temperatures during the most recent decade (2011–2020) exceed those of the most recent
multi-century warm period, around 6500 years ago13 [0.2°C to 1°C relative to 1850–1900] (medium confidence). Prior
to that, the next most recent warm period was about 125,000 years ago, when the multi-century temperature [0.5°C to
1.5°C relative to 1850–1900] overlaps the observations of the most recent decade (medium confidence).
{2.3, Cross-Chapter Box 2.1, Cross-Section Box TS.1} (Figure SPM.1)

A.2.3 In 2011–2020, annual average Arctic sea ice area reached its lowest level since at least 1850 (high confidence). Late
summer Arctic sea ice area was smaller than at any time in at least the past 1000 years (medium confidence). The global
nature of glacier retreat since the 1950s, with almost all of the world’s glaciers retreating synchronously, is unprecedented
in at least the last 2000 years (medium confidence).
{2.3, TS.2.5}

A.2.4 Global mean sea level has risen faster since 1900 than over any preceding century in at least the last 3000 years (high
confidence). The global ocean has warmed faster over the past century than since the end of the last deglacial transition
(around 11,000 years ago) (medium confidence). A long-term increase in surface open ocean pH occurred over the past
50 million years (high confidence). However, surface open ocean pH as low as recent decades is unusual in the last
2 million years (medium confidence).
{2.3, TS.2.4, Box TS.4}

A.3 Human-induced climate change is already affecting many weather and climate extremes in every region
across the globe. Evidence of observed changes in extremes such as heatwaves, heavy precipitation, droughts,
and tropical cyclones, and, in particular, their attribution to human influence, has strengthened since AR5.
{2.3, 3.3, 8.2, 8.3, 8.4, 8.5, 8.6, Box 8.1, Box 8.2, Box 9.2, 10.6, 11.2, 11.3, 11.4, 11.6, 11.7, 11.8, 11.9, 12.3}
(Figure SPM.3)

A.3.1 It is virtually certain that hot extremes (including heatwaves) have become more frequent and more intense across most
land regions since the 1950s, while cold extremes (including cold waves) have become less frequent and less severe, with
high confidence that human-induced climate change is the main driver14 of these changes. Some recent hot extremes
observed over the past decade would have been extremely unlikely to occur without human influence on the climate
system. Marine heatwaves have approximately doubled in frequency since the 1980s (high confidence), and human
influence has very likely contributed to most of them since at least 2006.
{Box 9.2, 11.2, 11.3, 11.9, TS.2.4, TS.2.6, Box TS.10} (Figure SPM.3)

A.3.2 The frequency and intensity of heavy precipitation events have increased since the 1950s over most land area for which
observational data are sufficient for trend analysis (high confidence), and human-induced climate change is likely the
main driver. Human-induced climate change has contributed to increases in agricultural and ecological droughts15 in some
regions due to increased land evapotranspiration16 (medium confidence).
{8.2, 8.3, 11.4, 11.6, 11.9, TS.2.6, Box TS.10} (Figure SPM.3)

13 As stated in section B.1, even under the very low emissions scenario SSP1-1.9, temperatures are assessed to remain elevated above those of the most recent decade until at least
2100 and therefore warmer than the century-scale period 6500 years ago.
14 As indicated in footnote 12, throughout this SPM, ‘main driver’ means responsible for more than 50% of the change.
15 Agricultural and ecological drought (depending on the affected biome): a period with abnormal soil moisture deficit, which results from combined shortage of precipitation
and excess evapotranspiration, and during the growing season impinges on crop production or ecosystem function in general (see Annex VII: Glossary). Observed changes in
meteorological droughts (precipitation deficits) and hydrological droughts (streamflow deficits) are distinct from those in agricultural and ecological droughts and are addressed in
the underlying AR6 material (Chapter 11).
16 The combined processes through which water is transferred to the atmosphere from open water and ice surfaces, bare soils and vegetation that make up the Earth’s surface (Glossary).

8
 Summary for Policymakers

A.3.3 Decreases in global land monsoon precipitation17 from the 1950s to the 1980s are partly attributed to human-caused
Northern Hemisphere aerosol emissions, but increases since then have resulted from rising GHG concentrations and
decadal to multi-decadal internal variability (medium confidence). Over South Asia, East Asia and West Africa, increases
in monsoon precipitation due to warming from GHG emissions were counteracted by decreases in monsoon precipitation
due to cooling from human-caused aerosol emissions over the 20th century (high confidence). Increases in West African
monsoon precipitation since the 1980s are partly due to the growing influence of GHGs and reductions in the cooling
effect of human-caused aerosol emissions over Europe and North America (medium confidence). SPM
{2.3, 3.3, 8.2, 8.3, 8.4, 8.5, 8.6, Box 8.1, Box 8.2, 10.6, Box TS.13}

A.3.4 It is likely that the global proportion of major (Category 3–5) tropical cyclone occurrence has increased over the last four
decades, and it is very likely that the latitude where tropical cyclones in the western North Pacific reach their peak intensity
has shifted northward; these changes cannot be explained by internal variability alone (medium confidence). There is low
confidence in long-term (multi-decadal to centennial) trends in the frequency of all-category tropical cyclones. Event
attribution studies and physical understanding indicate that human-induced climate change increases heavy precipitation
associated with tropical cyclones (high confidence), but data limitations inhibit clear detection of past trends on the
global scale.
{8.2, 11.7, Box TS.10}

A.3.5 Human influence has likely increased the chance of compound extreme events18 since the 1950s. This includes increases in
the frequency of concurrent heatwaves and droughts on the global scale (high confidence), fire weather in some regions
of all inhabited continents (medium confidence), and compound flooding in some locations (medium confidence).
{11.6, 11.7, 11.8, 12.3, 12.4, TS.2.6, Table TS.5, Box TS.10}

17 The global monsoon is defined as the area in which the annual range (local summer minus local winter) of precipitation is greater than 2.5 mm day–1 (Glossary). Global land monsoon
precipitation refers to the mean precipitation over land areas within the global monsoon.
18 Compound extreme events are the combination of multiple drivers and/or hazards that contribute to societal or environmental risk (Glossary). Examples are concurrent heatwaves
and droughts, compound flooding (e.g., a storm surge in combination with extreme rainfall and/or river flow), compound fire weather conditions (i.e., a combination of hot, dry and
windy conditions), or concurrent extremes at different locations.

9
 Summary for Policymakers

Climate change is already affecting every inhabited region across the globe,
with human influence contributing to many observed changes in weather
and climate extremes
(a) Synthesis of assessment of observed change in hot extremes and
confidence in human contribution to the observed changes in the world’s regions
SPM Type of observed change
in hot extremes North
America GIC Europe
Increase (41) NWN NEN NEU RAR

WNA CNA ENA WCE EEU WSB ESB RFE Asia

Decrease (0)

NCA MED WCA ECA TIB EAS

Low agreement in the type of change (2) Small
Islands
Central SCA CAR SAH ARP SAS SEA
Limited data and/or literature (2) America PAC
NWS NSA WAF CAF NEAF
NAU
Confidence in human contribution Small
to the observed change SAM NES WSAF SEAF MDG Islands
CAU EAU
High
South SWS SES ESAF
Medium America Africa
Australasia
SAU NZ
Low due to limited agreement SSA
Low due to limited evidence
Type of observed change since the 1950s

(b) Synthesis of assessment of observed change in heavy precipitation and

confidence in human contribution to the observed changes in the world’s regions
Type of observed change
in heavy precipitation North
America GIC Europe
NWN NEN NEU RAR
Increase (19)
WNA CNA ENA WCE EEU WSB ESB RFE Asia
Decrease (0)
NCA MED WCA ECA TIB EAS
Small
Low agreement in the type of change (8) Islands
Central SCA CAR SAH ARP SAS SEA
Limited data and/or literature (18) America PAC
NWS NSA WAF CAF NEAF
NAU
Confidence in human contribution Small
SAM NES WSAF SEAF MDG Islands
to the observed change CAU EAU
High South SWS SES ESAF
Africa SAU
Medium America
Australasia NZ
Low due to limited agreement SSA
Low due to limited evidence
Type of observed change since the 1950s

(c) Synthesis of assessment of observed change in agricultural and ecological drought

and confidence in human contribution to the observed changes in the world’s regions
Type of observed change
in agricultural and ecological drought North
America GIC Europe
Increase (12) NWN NEN NEU RAR

WNA CNA ENA WCE EEU WSB ESB RFE Asia

Decrease (1)

NCA MED WCA ECA TIB EAS

Low agreement in the type of change (28) Small
Islands
Central SCA CAR SAH ARP SAS SEA
Limited data and/or literature (4) America PAC
NWS NSA WAF CAF NEAF
NAU
Confidence in human contribution Small
to the observed change SAM NES WSAF SEAF MDG Islands
CAU EAU
High
South SWS SES ESAF
Medium America Africa SAU NZ
Australasia
Low due to limited agreement SSA
Low due to limited evidence
Type of observed change since the 1950s

Each hexagon corresponds IPCC AR6 WGI reference regions: North America: NWN (North-Western North America, NEN (North-Eastern North America), WNA
to one of the IPCC AR6 (Western North America), CNA (Central North America), ENA (Eastern North America), Central America: NCA (Northern Central America),
WGI reference regions SCA (Southern Central America), CAR (Caribbean), South America: NWS (North-Western South America), NSA (Northern South America), NES
(North-Eastern South America), SAM (South American Monsoon), SWS (South-Western South America), SES (South-Eastern South America),
North-Western SSA (Southern South America), Europe: GIC (Greenland/Iceland), NEU (Northern Europe), WCE (Western and Central Europe), EEU (Eastern
NWN Europe), MED (Mediterranean), Africa: MED (Mediterranean), SAH (Sahara), WAF (Western Africa), CAF (Central Africa), NEAF (North Eastern
North America
Africa), SEAF (South Eastern Africa), WSAF (West Southern Africa), ESAF (East Southern Africa), MDG (Madagascar), Asia: RAR (Russian
Arctic), WSB (West Siberia), ESB (East Siberia), RFE (Russian Far East), WCA (West Central Asia), ECA (East Central Asia), TIB (Tibetan Plateau),
EAS (East Asia), ARP (Arabian Peninsula), SAS (South Asia), SEA (South East Asia), Australasia: NAU (Northern Australia), CAU (Central
Australia), EAU (Eastern Australia), SAU (Southern Australia), NZ (New Zealand), Small Islands: CAR (Caribbean), PAC (Pacific Small Islands)
10
 Summary for Policymakers

Figure SPM.3 | Synthesis of assessed observed and attributable regional changes

The IPCC AR6 WGI inhabited regions are displayed as hexagons with identical size in their approximate geographical location (see legend for regional acronyms).
All assessments are made for each region as a whole and for the 1950s to the present. Assessments made on different time scales or more local spatial scales might
differ from what is shown in the figure. The colours in each panel represent the four outcomes of the assessment on observed changes. Striped hexagons (white
and light-grey) are used where there is low agreement in the type of change for the region as a whole, and grey hexagons are used when there is limited data and/
or literature that prevents an assessment of the region as a whole. Other colours indicate at least medium confidence in the observed change. The confidence
level for the human influence on these observed changes is based on assessing trend detection and attribution and event attribution literature, and it is indicated
by the number of dots: three dots for high confidence, two dots for medium confidence and one dot for low confidence (single, filled dot: limited agreement; single, SPM
empty dot: limited evidence).
Panel (a) For hot extremes, the evidence is mostly drawn from changes in metrics based on daily maximum temperatures; regional studies using other indices
(heatwave duration, frequency and intensity) are used in addition. Red hexagons indicate regions where there is at least medium confidence in an observed increase
in hot extremes.
Panel (b) For heavy precipitation, the evidence is mostly drawn from changes in indices based on one-day or five-day precipitation amounts using global and
regional studies. Green hexagons indicate regions where there is at least medium confidence in an observed increase in heavy precipitation.
Panel (c) Agricultural and ecological droughts are assessed based on observed and simulated changes in total column soil moisture, complemented
by evidence on changes in surface soil moisture, water balance (precipitation minus evapotranspiration) and indices driven by precipitation and atmospheric
evaporative demand. Yellow hexagons indicate regions where there is at least medium confidence in an observed increase in this type of drought, and green
hexagons indicate regions where there is at least medium confidence in an observed decrease in agricultural and ecological drought.
For all regions, Table TS.5 shows a broader range of observed changes besides the ones shown in this figure. Note that Southern South America (SSA) is the only
region that does not display observed changes in the metrics shown in this figure, but is affected by observed increases in mean temperature, decreases in frost
and increases in marine heatwaves.
{11.9, Atlas 1.3.3, Figure Atlas.2, Table TS.5; Box TS.10, Figure 1}

A.4 Improved knowledge of climate processes, paleoclimate evidence and the response of the climate system to
increasing radiative forcing gives a best estimate of equilibrium climate sensitivity of 3°C, with a narrower
range compared to AR5.
{2.2, 7.3, 7.4, 7.5, Box 7.2, 9.4, 9.5, 9.6, Cross-Chapter Box 9.1}

A.4.1 Human-caused radiative forcing of 2.72 [1.96 to 3.48] W m–2 in 2019 relative to 1750 has warmed the climate system. This
warming is mainly due to increased GHG concentrations, partly reduced by cooling due to increased aerosol concentrations.
The radiative forcing has increased by 0.43 W m–2 (19%) relative to AR5, of which 0.34 W m–2 is due to the increase in GHG
concentrations since 2011. The remainder is due to improved scientific understanding and changes in the assessment of
aerosol forcing, which include decreases in concentration and improvement in its calculation (high confidence).
{2.2, 7.3, TS.2.2, TS.3.1}

A.4.2 Human-caused net positive radiative forcing causes an accumulation of additional energy (heating) in the climate system,
partly reduced by increased energy loss to space in response to surface warming. The observed average rate of heating of
the climate system increased from 0.50 [0.32 to 0.69] W m–2 for the period 1971–200619 to 0.79 [0.52 to 1.06] W m–2 for
the period 2006–201820 (high confidence). Ocean warming accounted for 91% of the heating in the climate system, with
land warming, ice loss and atmospheric warming accounting for about 5%, 3% and 1%, respectively (high confidence).
{7.2, Box 7.2, TS.3.1}

A.4.3 Heating of the climate system has caused global mean sea level rise through ice loss on land and thermal expansion
from ocean warming. Thermal expansion explained 50% of sea level rise during 1971–2018, while ice loss from glaciers
contributed 22%, ice sheets 20% and changes in land-water storage 8%. The rate of ice-sheet loss increased by a factor
of four between 1992–1999 and 2010–2019. Together, ice-sheet and glacier mass loss were the dominant contributors to
global mean sea level rise during 2006–2018 (high confidence).
{9.4, 9.5, 9.6, Cross-Chapter Box 9.1}

A.4.4 The equilibrium climate sensitivity is an important quantity used to estimate how the climate responds to radiative
forcing. Based on multiple lines of evidence,21 the very likely range of equilibrium climate sensitivity is between 2°C (high
confidence) and 5°C (medium confidence). The AR6 assessed best estimate is 3°C with a likely range of 2.5°C to 4°C
(high confidence), compared to 1.5°C to 4.5°C in AR5, which did not provide a best estimate.
{7.4, 7.5, TS.3.2}

19 Cumulative energy increase of 282 [177 to 387] ZJ over 1971–2006 (1 ZJ = 1021 joules).
20 Cumulative energy increase of 152 [100 to 205] ZJ over 2006–2018.
21 Understanding of climate processes, the instrumental record, paleoclimates and model-based emergent constraints (Glossary).

11
 Summary for Policymakers

B. Possible Climate Futures

A set of five new illustrative emissions scenarios is considered consistently across this Report to explore the climate response to
a broader range of greenhouse gas (GHG), land-use and air pollutant futures than assessed in AR5. This set of scenarios drives
climate model projections of changes in the climate system. These projections account for solar activity and background forcing
from volcanoes. Results over the 21st century are provided for the near term (2021–2040), mid-term (2041–2060) and long term
SPM (2081–2100) relative to 1850–1900, unless otherwise stated.

Box SPM.1 | Scenarios, Climate Models and Projections

Box SPM.1.1: This Report assesses the climate response to five illustrative scenarios that cover the range of possible future
development of anthropogenic drivers of climate change found in the literature. They start in 2015, and include scenarios22
with high and very high GHG emissions (SSP3-7.0 and SSP5-8.5) and CO2 emissions that roughly double from current
levels by 2100 and 2050, respectively, scenarios with intermediate GHG emissions (SSP2-4.5) and CO2 emissions remaining
around current levels until the middle of the century, and scenarios with very low and low GHG emissions and CO2 emissions
declining to net zero around or after 2050, followed by varying levels of net negative CO2 emissions23 (SSP1-1.9 and
SSP1-2.6), as illustrated in Figure SPM.4. Emissions vary between scenarios depending on socio-economic assumptions,
levels of climate change mitigation and, for aerosols and non-methane ozone precursors, air pollution controls. Alternative
assumptions may result in similar emissions and climate responses, but the socio-economic assumptions and the feasibility
or likelihood of individual scenarios are not part of the assessment.
{1.6, Cross-Chapter Box 1.4, TS.1.3} (Figure SPM.4)

Box SPM.1.2: This Report assesses results from climate models participating in the Coupled Model Intercomparison Project
Phase 6 (CMIP6) of the World Climate Research Programme. These models include new and better representations of
physical, chemical and biological processes, as well as higher resolution, compared to climate models considered in previous
IPCC assessment reports. This has improved the simulation of the recent mean state of most large-scale indicators of climate
change and many other aspects across the climate system. Some differences from observations remain, for example in
regional precipitation patterns. The CMIP6 historical simulations assessed in this Report have an ensemble mean global
surface temperature change within 0.2°C of the observations over most of the historical period, and observed warming is
within the very likely range of the CMIP6 ensemble. However, some CMIP6 models simulate a warming that is either above
or below the assessed very likely range of observed warming.
{1.5, Cross-Chapter Box 2.2, 3.3, 3.8, TS.1.2, Cross-Section Box TS.1} (Figure SPM.1b, Figure SPM.2)

Box SPM.1.3: The CMIP6 models considered in this Report have a wider range of climate sensitivity than in CMIP5 models
and the AR6 assessed very likely range, which is based on multiple lines of evidence. These CMIP6 models also show
a higher average climate sensitivity than CMIP5 and the AR6 assessed best estimate. The higher CMIP6 climate sensitivity
values compared to CMIP5 can be traced to an amplifying cloud feedback that is larger in CMIP6 by about 20%.
{Box 7.1, 7.3, 7.4, 7.5, TS.3.2}

Box SPM.1.4: For the first time in an IPCC report, assessed future changes in global surface temperature, ocean warming
and sea level are constructed by combining multi-model projections with observational constraints based on past simulated
warming, as well as the AR6 assessment of climate sensitivity. For other quantities, such robust methods do not yet exist
to constrain the projections. Nevertheless, robust projected geographical patterns of many variables can be identified at
a given level of global warming, common to all scenarios considered and independent of timing when the global warming
level is reached.
{1.6, 4.3, 4.6, Box 4.1, 7.5, 9.2, 9.6, Cross-Chapter Box 11.1, Cross-Section Box TS.1}

22 Throughout this Report, the five illustrative scenarios are referred to as SSPx-y, where ‘SSPx’ refers to the Shared Socio-economic Pathway or ‘SSP’ describing the socio-economic
trends underlying the scenario, and ‘y’ refers to the approximate level of radiative forcing (in watts per square metre, or W m–2) resulting from the scenario in the year 2100.
A detailed comparison to scenarios used in earlier IPCC reports is provided in Section TS.1.3, and Sections 1.6 and 4.6. The SSPs that underlie the specific forcing scenarios used to
drive climate models are not assessed by WGI. Rather, the SSPx-y labelling ensures traceability to the underlying literature in which specific forcing pathways are used as input to the
climate models. IPCC is neutral with regard to the assumptions underlying the SSPs, which do not cover all possible scenarios. Alternative scenarios may be considered or developed.
23 Net negative CO2 emissions are reached when anthropogenic removals of CO2 exceed anthropogenic emissions (Glossary).

12
 Summary for Policymakers

Box SPM.1 (continued)

Future emissions cause future additional warming, with total warming

dominated by past and future CO₂ emissions
(a) Future annual emissions of CO₂ (left) and of a subset of key non-CO₂ drivers (right), across five illustrative scenarios
SPM

Carbon dioxide (GtCO₂/yr) Selected contributors to non-CO₂ GHGs

Methane (MtCH₄/yr)
140
800 SSP3-7.0
SSP5-8.5 600
120 SSP5-8.5
400
SSP2-4.5
200
SSP1-2.6
100 0 SSP1-1.9
2015 2050 2100
SSP3-7.0
80 Nitrous oxide (MtN₂O/yr)
SSP3-7.0
20

60 SSP5-8.5
10 SSP2-4.5
SSP1-2.6
SSP1-1.9
40 0
2015 2050 2100

20 One air pollutant and contributor to aerosols

SSP2-4.5 Sulphur dioxide (MtSO₂/yr)
0 120
SSP1-2.6
80 SSP3-7.0
SSP1-1.9
–20
40 SSP2-4.5
2015 2050 2100 SSP5-8.5
SSP1-1.9
0 SSP1-2.6
2015 2050 2100

(b) Contribution to global surface temperature increase from different emissions, with a dominant role of CO₂ emissions
Change in global surface temperature in 2081–2100 relative to 1850–1900 (ºC)

SSP1-1.9 SSP1-2.6 SSP2-4.5 SSP3-7.0 SSP5-8.5

ºC ºC ºC ºC ºC
6 6 6 6 6

5 5 5 5 5

4 4 4 4 4

3 3 3 3 3

2 2 2 2 2

1 1 1 1 1

0 0 0 0 0

–1 –1 –1 –1 –1
Total CO₂ Non-CO₂ Aerosols Total CO₂ Non-CO₂ Aerosols Total CO₂ Non-CO₂ Aerosols Total CO₂ Non-CO₂ Aerosols Total CO₂ Non-CO₂ Aerosols
(observed) GHGs land use (observed) GHGs Land use (observed) GHGs Land use (observed) GHGs Land use (observed) GHGs Land use

Total warming (observed warming to date in darker shade), warming from CO₂, warming from non-CO₂ GHGs and cooling from changes in aerosols and land use

Figure SPM.4 | Future anthropogenic emissions of key drivers of climate change and warming contributions by groups of drivers for
the five illustrative scenarios used in this report
The five scenarios are SSP1-1.9, SSP1-2.6, SSP2-4.5, SSP3-7.0 and SSP5-8.5.
Panel (a) Annual anthropogenic (human-caused) emissions over the 2015–2100 period. Shown are emissions trajectories for carbon dioxide
(CO2) from all sectors (GtCO2/yr) (left graph) and for a subset of three key non-CO2 drivers considered in the scenarios: methane (CH4, MtCH4/yr, top-right
graph); nitrous oxide (N2O, MtN2O/yr, middle-right graph); and sulphur dioxide (SO2, MtSO2/yr, bottom-right graph, contributing to anthropogenic aerosols
in panel (b).

13
 Summary for Policymakers

Panel (b) Warming contributions by groups of anthropogenic drivers and by scenario are shown as the change in global surface
temperature (°C) in 2081–2100 relative to 1850–1900, with indication of the observed warming to date. Bars and whiskers represent median values
and the very likely range, respectively. Within each scenario bar plot, the bars represent: total global warming (°C; ‘total’ bar) (see Table SPM.1); warming
contributions (°C) from changes in CO2 (‘CO2’ bar) and from non-CO2 greenhouse gases (GHGs; ‘non-CO2 GHGs’ bar: comprising well-mixed greenhouse
gases and ozone); and net cooling from other anthropogenic drivers (‘aerosols and land use’ bar: anthropogenic aerosols, changes in reflectance due to
land-use and irrigation changes, and contrails from aviation) (see Figure SPM.2, panel c, for the warming contributions to date for individual drivers). The
SPM best estimate for observed warming in 2010–2019 relative to 1850–1900 (see Figure SPM.2, panel a) is indicated in the darker column in the ‘total’ bar.
Warming contributions in panel (b) are calculated as explained in Table SPM.1 for the total bar. For the other bars, the contribution by groups of drivers is
calculated with a physical climate emulator of global surface temperature that relies on climate sensitivity and radiative forcing assessments.
{Cross-Chapter Box 1.4; 4.6; Figure 4.35; 6.7; Figures 6.18, 6.22 and 6.24; 7.3; Cross-Chapter Box 7.1; Figure 7.7; Box TS.7; Figures TS.4 and TS.15}

B.1 Global surface temperature will continue to increase until at least mid-century under all emissions scenarios
considered. Global warming of 1.5°C and 2°C will be exceeded during the 21st century unless deep reductions
in CO2 and other greenhouse gas emissions occur in the coming decades.
{2.3, Cross-Chapter Box 2.3, Cross-Chapter Box 2.4, 4.3, 4.4, 4.5} (Figure SPM.1, Figure SPM.4, Figure SPM.8,
Table SPM.1, Box SPM.1)

B.1.1 Compared to 1850–1900, global surface temperature averaged over 2081–2100 is very likely to be higher by 1.0°C to
1.8°C under the very low GHG emissions scenario considered (SSP1-1.9), by 2.1°C to 3.5°C in the intermediate GHG
emissions scenario (SSP2-4.5) and by 3.3°C to 5.7°C under the very high GHG emissions scenario (SSP5-8.5).24 The last
time global surface temperature was sustained at or above 2.5°C higher than 1850–1900 was over 3 million years ago
(medium confidence).
{2.3, Cross-Chapter Box 2.4, 4.3, 4.5, Box TS.2, Box TS.4, Cross-Section Box TS.1} (Table SPM.1)

Table SPM.1 | Changes in global surface temperature, which are assessed based on multiple lines of evidence, for selected 20-year time
periods and the five illustrative emissions scenarios considered. Temperature differences relative to the average global surface temperature of the
period 1850–1900 are reported in °C. This includes the revised assessment of observed historical warming for the AR5 reference period 1986–2005, which
in AR6 is higher by 0.08 [–0.01 to +0.12] °C than in AR5 (see footnote 10). Changes relative to the recent reference period 1995–2014 may be calculated
approximately by subtracting 0.85°C, the best estimate of the observed warming from 1850–1900 to 1995–2014.
{Cross-Chapter Box 2.3, 4.3, 4.4, Cross-Section Box TS.1}

Near term, 2021–2040 Mid-term, 2041–2060 Long term, 2081–2100

Very likely Very likely Very likely
Scenario Best estimate (°C) Best estimate (°C) Best estimate (°C)
range (°C) range (°C) range (°C)
SSP1-1.9 1.5 1.2 to 1.7 1.6 1.2 to 2.0 1.4 1.0 to 1.8

SSP1-2.6 1.5 1.2 to 1.8 1.7 1.3 to 2.2 1.8 1.3 to 2.4

SSP2-4.5 1.5 1.2 to 1.8 2.0 1.6 to 2.5 2.7 2.1 to 3.5

SSP3-7.0 1.5 1.2 to 1.8 2.1 1.7 to 2.6 3.6 2.8 to 4.6

SSP5-8.5 1.6 1.3 to 1.9 2.4 1.9 to 3.0 4.4 3.3 to 5.7

B.1.2 Based on the assessment of multiple lines of evidence, global warming of 2°C, relative to 1850–1900, would be exceeded
during the 21st century under the high and very high GHG emissions scenarios considered in this report (SSP3-7.0 and
SSP5-8.5, respectively). Global warming of 2°C would extremely likely be exceeded in the intermediate GHG emissions
scenario (SSP2-4.5). Under the very low and low GHG emissions scenarios, global warming of 2°C is extremely unlikely
to be exceeded (SSP1-1.9) or unlikely to be exceeded (SSP1-2.6).25 Crossing the 2°C global warming level in the mid-
term period (2041–2060) is very likely to occur under the very high GHG emissions scenario (SSP5-8.5), likely to occur
under the high GHG emissions scenario (SSP3-7.0), and more likely than not to occur in the intermediate GHG emissions
scenario (SSP2-4.5).26
{4.3, Cross-Section Box TS.1} (Table SPM.1, Figure SPM.4, Box SPM.1)

24 Changes in global surface temperature are reported as running 20-year averages, unless stated otherwise.
25 SSP1-1.9 and SSP1-2.6 are scenarios that start in 2015 and have very low and low GHG emissions, respectively, and CO2 emissions declining to net zero around or after 2050,
followed by varying levels of net negative CO2 emissions.
26 Crossing is defined here as having the assessed global surface temperature change, averaged over a 20-year period, exceed a particular global warming level.

14
 Summary for Policymakers

B.1.3 Global warming of 1.5°C relative to 1850–1900 would be exceeded during the 21st century under the intermediate, high
and very high GHG emissions scenarios considered in this report (SSP2-4.5, SSP3-7.0 and SSP5-8.5, respectively). Under
the five illustrative scenarios, in the near term (2021–2040), the 1.5°C global warming level is very likely to be exceeded
under the very high GHG emissions scenario (SSP5-8.5), likely to be exceeded under the intermediate and high GHG
emissions scenarios (SSP2-4.5 and SSP3-7.0), more likely than not to be exceeded under the low GHG emissions scenario
(SSP1-2.6) and more likely than not to be reached under the very low GHG emissions scenario (SSP1-1.9).27 Furthermore, for
the very low GHG emissions scenario (SSP1-1.9), it is more likely than not that global surface temperature would decline SPM
back to below 1.5°C toward the end of the 21st century, with a temporary overshoot of no more than 0.1°C above 1.5°C
global warming.
{4.3, Cross-Section Box TS.1} (Table SPM.1, Figure SPM.4)

B.1.4 Global surface temperature in any single year can vary above or below the long-term human-induced trend, due to
substantial natural variability.28 The occurrence of individual years with global surface temperature change above a certain
level, for example 1.5°C or 2°C, relative to 1850–1900 does not imply that this global warming level has been reached.29
{Cross-Chapter Box 2.3, 4.3, 4.4, Box 4.1, Cross-Section Box TS.1} (Table SPM.1, Figure SPM.1, Figure SPM.8)

B.2 Many changes in the climate system become larger in direct relation to increasing global warming. They
include increases in the frequency and intensity of hot extremes, marine heatwaves, heavy precipitation,
and, in some regions, agricultural and ecological droughts; an increase in the proportion of intense tropical
cyclones; and reductions in Arctic sea ice, snow cover and permafrost.
{4.3, 4.5, 4.6, 7.4, 8.2, 8.4, Box 8.2, 9.3, 9.5, Box 9.2, 11.1, 11.2, 11.3, 11.4, 11.6, 11.7, 11.9, Cross-Chapter Box
11.1, 12.4, 12.5, Cross-Chapter Box 12.1, Atlas.4, Atlas.5, Atlas.6, Atlas.7, Atlas.8, Atlas.9, Atlas.10, Atlas.11}
(Figure SPM.5, Figure SPM.6, Figure SPM.8)

B.2.1 It is virtually certain that the land surface will continue to warm more than the ocean surface (likely 1.4 to 1.7 times more).
It is virtually certain that the Arctic will continue to warm more than global surface temperature, with high confidence
above two times the rate of global warming.
{2.3, 4.3, 4.5, 4.6, 7.4, 11.1, 11.3, 11.9, 12.4, 12.5, Cross-Chapter Box 12.1, Atlas.4, Atlas.5, Atlas.6, Atlas.7, Atlas.8, Atlas.9,
Atlas.10, Atlas.11, Cross-Section Box TS.1, TS.2.6} (Figure SPM.5)

B.2.2 With every additional increment of global warming, changes in extremes continue to become larger. For example, every
additional 0.5°C of global warming causes clearly discernible increases in the intensity and frequency of hot extremes,
including heatwaves (very likely), and heavy precipitation (high confidence), as well as agricultural and ecological
droughts30 in some regions (high confidence). Discernible changes in intensity and frequency of meteorological droughts,
with more regions showing increases than decreases, are seen in some regions for every additional 0.5°C of global
warming (medium confidence). Increases in frequency and intensity of hydrological droughts become larger with
increasing global warming in some regions (medium confidence). There will be an increasing occurrence of some extreme
events unprecedented in the observational record with additional global warming, even at 1.5°C of global warming.
Projected percentage changes in frequency are larger for rarer events (high confidence).
{8.2, 11.2, 11.3, 11.4, 11.6, 11.9, Cross-Chapter Box 11.1, Cross-Chapter Box 12.1, TS.2.6} (Figure SPM.5, Figure SPM.6)

B.2.3 Some mid-latitude and semi-arid regions, and the South American Monsoon region, are projected to see the highest
increase in the temperature of the hottest days, at about 1.5 to 2 times the rate of global warming (high confidence). The
Arctic is projected to experience the highest increase in the temperature of the coldest days, at about three times the rate
of global warming (high confidence). With additional global warming, the frequency of marine heatwaves will continue
to increase (high confidence), particularly in the tropical ocean and the Arctic (medium confidence).
{Box 9.2, 11.1, 11.3, 11.9, Cross-Chapter Box 11.1, Cross-Chapter Box 12.1, 12.4, TS.2.4, TS.2.6} (Figure SPM.6)

27 The AR6 assessment of when a given global warming level is first exceeded benefits from the consideration of the illustrative scenarios, the multiple lines of evidence entering the
assessment of future global surface temperature response to radiative forcing, and the improved estimate of historical warming. The AR6 assessment is thus not directly comparable to
the SR1.5 SPM, which reported likely reaching 1.5°C global warming between 2030 and 2052, from a simple linear extrapolation of warming rates of the recent past. When considering
scenarios similar to SSP1-1.9 instead of linear extrapolation, the SR1.5 estimate of when 1.5°C global warming is first exceeded is close to the best estimate reported here.
28 Natural variability refers to climatic fluctuations that occur without any human influence, that is, internal variability combined with the response to external natural factors such as
volcanic eruptions, changes in solar activity and, on longer time scales, orbital effects and plate tectonics (Glossary).
29 The internal variability in any single year is estimated to be about ±0.25°C (5–95% range, high confidence).
30 Projected changes in agricultural and ecological droughts are primarily assessed based on total column soil moisture. See footnote 15 for definition and relation to precipitation
and evapotranspiration.

15
 Summary for Policymakers

B.2.4 It is very likely that heavy precipitation events will intensify and become more frequent in most regions with additional global
warming. At the global scale, extreme daily precipitation events are projected to intensify by about 7% for each 1°C of global
warming (high confidence). The proportion of intense tropical cyclones (Category 4–5) and peak wind speeds of the most
intense tropical cyclones are projected to increase at the global scale with increasing global warming (high confidence).
{8.2, 11.4, 11.7, 11.9, Cross-Chapter Box 11.1, Box TS.6, TS.4.3.1} (Figure SPM.5, Figure SPM.6)

SPM B.2.5 Additional warming is projected to further amplify permafrost thawing and loss of seasonal snow cover, of land ice and of
Arctic sea ice (high confidence). The Arctic is likely to be practically sea ice-free in September31 at least once before 2050
under the five illustrative scenarios considered in this report, with more frequent occurrences for higher warming levels.
There is low confidence in the projected decrease of Antarctic sea ice.
{4.3, 4.5, 7.4, 8.2, 8.4, Box 8.2, 9.3, 9.5, 12.4, Cross-Chapter Box 12.1, Atlas.5, Atlas.6, Atlas.8, Atlas.9, Atlas.11, TS.2.5}
(Figure SPM.8)

With every increment of global warming, changes get larger

in regional mean temperature, precipitation and soil moisture

(a) Annual mean temperature change (°C)

at 1°C global warming
Observed change per 1°C global warming Simulated change at 1°C global warming

Warming at 1°C affects all continents and

is generally larger over land than over the
oceans in both observations and models.
Across most regions, observed and
simulated patterns are consistent.

(b) Annual mean temperature change (°C) Across warming levels, land areas warm more than ocean areas, and the
Arctic and Antarctica warm more than the tropics.
relative to 1850–1900
Simulated change at 1.5°C global warming Simulated change at 2°C global warming Simulated change at 4°C global warming

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7

Change (°C)
Warmer

31 Monthly average sea ice area of less than 1 million km2, which is about 15% of the average September sea ice area observed in 1979–1988.

16
 Summary for Policymakers

l a.
(c) Annual mean precipitation change (%) Precipitation is projected to increase over high latitudes, the equatorial
Pacific and parts of the monsoon regions, but decrease over parts of the
relative to 1850–1900 subtropics and in limited areas of the tropics.

Simulated change at 1.5°C global warming Simulated change at 2°C global warming Simulated change at 4°C global warming

SPM

Relatively small absolute changes

may appear as large % changes in
–40 –30 –20 –10 0 10 20 30 40
regions with dry baseline conditions.
Change (%)
Drier Wetter

(d) Annual mean total column soil Across warming levels, changes in soil moisture largely follow changes in
precipitation but also show some differences due to the influence of
moisture change (standard deviation) evapotranspiration.

Simulated change at 1.5°C global warming Simulated change at 2°C global warming Simulated change at 4°C global warming

Relatively small absolute changes

may appear large when expressed –1.5 –1.0 –0.5 0 0.5 1.0 1.5
in units of standard deviation in dry
Change (standard deviation
regions with little interannual
Drier of interannual variability) Wetter
variability in baseline conditions.

Figure SPM.5 | Changes in annual mean surface temperature, precipitation, and soil moisture
Panel (a) Comparison of observed and simulated annual mean surface temperature change. The left map shows the observed changes in annual
mean surface temperature in the period 1850–2020 per °C of global warming (°C). The local (i.e., grid point) observed annual mean surface temperature changes
are linearly regressed against the global surface temperature in the period 1850–2020. Observed temperature data are from Berkeley Earth, the dataset with
the largest coverage and highest horizontal resolution. Linear regression is applied to all years for which data at the corresponding grid point is available. The
regression method was used to take into account the complete observational time series and thereby reduce the role of internal variability at the grid point level.
White indicates areas where time coverage was 100 years or less and thereby too short to calculate a reliable linear regression. The right map is based on model
simulations and shows change in annual multi-model mean simulated temperatures at a global warming level of 1°C (20-year mean global surface temperature
change relative to 1850–1900). The triangles at each end of the colour bar indicate out-of-bound values, that is, values above or below the given limits.
Panel (b) Simulated annual mean temperature change (°C), panel (c) precipitation change (%), and panel (d) total column soil moisture change
(standard deviation of interannual variability) at global warming levels of 1.5°C, 2°C and 4°C (20-year mean global surface temperature change relative
to 1850–1900). Simulated changes correspond to Coupled Model Intercomparison Project Phase 6 (CMIP6) multi-model mean change (median change for soil
moisture) at the corresponding global warming level, that is, the same method as for the right map in panel (a).
In panel (c), high positive percentage changes in dry regions may correspond to small absolute changes. In panel (d), the unit is the standard deviation
of interannual variability in soil moisture during 1850–1900. Standard deviation is a widely used metric in characterizing drought severity. A projected
reduction in mean soil moisture by one standard deviation corresponds to soil moisture conditions typical of droughts that occurred about once every six years
during 1850–1900. In panel (d), large changes in dry regions with little interannual variability in the baseline conditions can correspond to small absolute
change. The triangles at each end of the colour bars indicate out-of-bound values, that is, values above or below the given limits. Results from all models
reaching the corresponding warming level in any of the five illustrative scenarios (SSP1-1.9, SSP1-2.6, SSP2-4.5, SSP3-7.0 and SSP5-8.5) are averaged.
Maps of annual mean temperature and precipitation changes at a global warming level of 3°C are available in Figure 4.31 and Figure 4.32 in Section 4.6.
Corresponding maps of panels (b), (c) and (d), including hatching to indicate the level of model agreement at grid-cell level, are found in Figures 4.31, 4.32 and
11.19, respectively; as highlighted in Cross-Chapter Box Atlas.1, grid-cell level hatching is not informative for larger spatial scales (e.g., over AR6 reference regions)
where the aggregated signals are less affected by small-scale variability, leading to an increase in robustness.
{Figure 1.14, 4.6.1, Cross-Chapter Box 11.1, Cross-Chapter Box Atlas.1, TS.1.3.2, Figures TS.3 and TS.5}

17
 Summary for Policymakers

Projected changes in extremes are larger in frequency and intensity with

every additional increment of global warming

Hot temperature extremes over land

10-year event 50-year event
SPM
Frequency and increase in intensity of extreme temperature Frequency and increase in intensity of extreme temperature
event that occurred once in 10 years on average event that occurred once in 50 years on average
in a climate without human influence in a climate without human influence

Future global warming levels Future global warming levels

1850–1900 Present 1°C 1.5°C 2°C 4°C 1850–1900 Present 1°C 1.5°C 2°C 4°C

FREQUENCY per 50 years

FREQUENCY per 10 years

Once now likely will likely will likely will likely Once now likely will likely will likely will likely
occurs occur occur occur occurs occur occur occur
2.8 times 4.1 times 5.6 times 9.4 times 4.8 times 8.6 times 13.9 times 39.2 times
(1.8–3.2) (2.8–4.7) (3.8–6.0) (8.3–9.6) (2.3–6.4) (4.3–10.7) (6.9–16.6) (27.0–41.4)

+6°C +6°C
INTENSITY increase

INTENSITY increase
+5°C +5°C
+4°C +4°C
+3°C +3°C
+2°C +2°C
+1°C +1°C
0°C 0°C
+1.2°C +1.9°C +2.6°C +5.1°C +1.2°C +2.0°C +2.7°C +5.3°C
hotter hotter hotter hotter hotter hotter hotter hotter

Heavy precipitation over land Agricultural & ecological droughts in drying regions
10-year event 10-year event
Frequency and increase in intensity of heavy 1-day Frequency and increase in intensity of an agricultural and ecological
precipitation event that occurred once in 10 years on drought event that occurred once in 10 years on average across
average in a climate without human influence drying regions in a climate without human influence

Future global warming levels Future global warming levels

1850–1900 Present 1°C 1.5°C 2°C 4°C 1850–1900 Present 1°C 1.5°C 2°C 4°C
FREQUENCY per 10 years
FREQUENCY per 10 years

Once now likely will likely will likely will likely Once now likely will likely will likely will likely
occurs occur occur occur occurs occur occur occur
1.3 times 1.5 times 1.7 times 2.7 times 1.7 times 2.0 times 2.4 times 4.1 times
(1.2–1.4) (1.4–1.7) (1.6–2.0) (2.3–3.6) (0.7–4.1) (1.0–5.1) (1.3–5.8) (1.7–7.2)

+40% +2 sd
INTENSITY increase

INTENSITY increase

+30%

+20% +1 sd

+10%

0% 0 sd
+6.7% +10.5% +14.0% +30.2% +0.3 sd +0.5 sd +0.6 sd +1.0 sd
wetter wetter wetter wetter drier drier drier drier

Figure SPM.6 | Projected changes in the intensity and frequency of hot temperature extremes over land, extreme precipitation over land,
and agricultural and ecological droughts in drying regions
Projected changes are shown at global warming levels of 1°C, 1.5°C, 2°C, and 4°C and are relative to 1850–1900,9 representing a climate without human
influence. The figure depicts frequencies and increases in intensity of 10- or 50-year extreme events from the base period (1850–1900) under different global
warming levels.
Hot temperature extremes are defined as the daily maximum temperatures over land that were exceeded on average once in a decade (10-year event) or once
in 50 years (50-year event) during the 1850–1900 reference period. Extreme precipitation events are defined as the daily precipitation amount over land that

18
 Summary for Policymakers

was exceeded on average once in a decade during the 1850–1900 reference period. Agricultural and ecological drought events are defined as the annual
average of total column soil moisture below the 10th percentile of the 1850–1900 base period. These extremes are defined on model grid box scale. For hot
temperature extremes and extreme precipitation, results are shown for the global land. For agricultural and ecological drought, results are shown for drying regions
only, which correspond to the AR6 regions in which there is at least medium confidence in a projected increase in agricultural and ecological droughts at the 2°C
warming level compared to the 1850–1900 base period in the Coupled Model Intercomparison Project Phase 6 (CMIP6). These regions include Western North
America, Central North America, Northern Central America, Southern Central America, Caribbean, Northern South America, North-Eastern South America, South
American Monsoon, South-Western South America, Southern South America, Western and Central Europe, Mediterranean, West Southern Africa, East Southern
Africa, Madagascar, Eastern Australia, and Southern Australia (Caribbean is not included in the calculation of the figure because of the too-small number of full land
SPM
grid cells). The non-drying regions do not show an overall increase or decrease in drought severity. Projections of changes in agricultural and ecological droughts
in the CMIP Phase 5 (CMIP5) multi-model ensemble differ from those in CMIP6 in some regions, including in parts of Africa and Asia. Assessments of projected
changes in meteorological and hydrological droughts are provided in Chapter 11.
In the ‘frequency’ section, each year is represented by a dot. The dark dots indicate years in which the extreme threshold is exceeded, while light dots are years
when the threshold is not exceeded. Values correspond to the medians (in bold) and their respective 5–95% range based on the multi-model ensemble from
simulations of CMIP6 under different Shared Socio-economic Pathway scenarios. For consistency, the number of dark dots is based on the rounded-up median.
In the ‘intensity’ section, medians and their 5–95% range, also based on the multi-model ensemble from simulations of CMIP6, are displayed as dark and
light bars, respectively. Changes in the intensity of hot temperature extremes and extreme precipitation are expressed as degree Celsius and percentage. As for
agricultural and ecological drought, intensity changes are expressed as fractions of standard deviation of annual soil moisture.
{11.1; 11.3; 11.4; 11.6; 11.9; Figures 11.12, 11.15, 11.6, 11.7, and 11.18}

B.3 Continued global warming is projected to further intensify the global water cycle, including its variability,
global monsoon precipitation and the severity of wet and dry events.
{4.3, 4.4, 4.5, 4.6, 8.2, 8.3, 8.4, 8.5, Box 8.2, 11.4, 11.6, 11.9, 12.4, Atlas.3} (Figure SPM.5, Figure SPM.6)

B.3.1 There is strengthened evidence since AR5 that the global water cycle will continue to intensify as global temperatures
rise (high confidence), with precipitation and surface water flows projected to become more variable over most land
regions within seasons (high confidence) and from year to year (medium confidence). The average annual global land
precipitation is projected to increase by 0–5% under the very low GHG emissions scenario (SSP1-1.9), 1.5–8% for the
intermediate GHG emissions scenario (SSP2-4.5) and 1–13% under the very high GHG emissions scenario (SSP5-8.5) by
2081–2100 relative to 1995–2014 (likely ranges). Precipitation is projected to increase over high latitudes, the equatorial
Pacific and parts of the monsoon regions, but decrease over parts of the subtropics and limited areas in the tropics
in SSP2-4.5, SSP3-7.0 and SSP5-8.5 (very likely). The portion of the global land experiencing detectable increases or
decreases in seasonal mean precipitation is projected to increase (medium confidence). There is high confidence in an
earlier onset of spring snowmelt, with higher peak flows at the expense of summer flows in snow-dominated regions
globally.
{4.3, 4.5, 4.6, 8.2, 8.4, Atlas.3, TS.2.6, TS.4.3, Box TS.6} (Figure SPM.5)

B.3.2 A warmer climate will intensify very wet and very dry weather and climate events and seasons, with implications for
flooding or drought (high confidence), but the location and frequency of these events depend on projected changes in
regional atmospheric circulation, including monsoons and mid-latitude storm tracks. It is very likely that rainfall variability
related to the El Niño–Southern Oscillation is projected to be amplified by the second half of the 21st century in the
SSP2-4.5, SSP3-7.0 and SSP5-8.5 scenarios.
{4.3, 4.5, 4.6, 8.2, 8.4, 8.5, 11.4, 11.6, 11.9, 12.4, TS.2.6, TS.4.2, Box TS.6} (Figure SPM.5, Figure SPM.6)

B.3.3 Monsoon precipitation is projected to increase in the mid- to long term at the global scale, particularly over South and
South East Asia, East Asia and West Africa apart from the far west Sahel (high confidence). The monsoon season is
projected to have a delayed onset over North and South America and West Africa (high confidence) and a delayed retreat
over West Africa (medium confidence).
{4.4, 4.5, 8.2, 8.3, 8.4, Box 8.2, Box TS.13}

B.3.4 A projected southward shift and intensification of Southern Hemisphere summer mid-latitude storm tracks and associated
precipitation is likely in the long term under high GHG emissions scenarios (SSP3-7.0, SSP5-8.5), but in the near term
the effect of stratospheric ozone recovery counteracts these changes (high confidence). There is medium confidence in
a continued poleward shift of storms and their precipitation in the North Pacific, while there is low confidence in projected
changes in the North Atlantic storm tracks.
{4.4, 4.5, 8.4, TS.2.3, TS.4.2}

B.4 Under scenarios with increasing CO2 emissions, the ocean and land carbon sinks are projected to be less
effective at slowing the accumulation of CO2 in the atmosphere.
{4.3, 5.2, 5.4, 5.5, 5.6} (Figure SPM.7)

19
 Summary for Policymakers

B.4.1 While natural land and ocean carbon sinks are projected to take up, in absolute terms, a progressively larger amount
of CO2 under higher compared to lower CO2 emissions scenarios, they become less effective, that is, the proportion of
emissions taken up by land and ocean decrease with increasing cumulative CO2 emissions. This is projected to result in
a higher proportion of emitted CO2 remaining in the atmosphere (high confidence).
{5.2, 5.4, Box TS.5} (Figure SPM.7)

SPM B.4.2 Based on model projections, under the intermediate GHG emissions scenario that stabilizes atmospheric CO2 concentrations
this century (SSP2-4.5), the rates of CO2 taken up by the land and ocean are projected to decrease in the second half of
the 21st century (high confidence). Under the very low and low GHG emissions scenarios (SSP1-1.9, SSP1-2.6), where CO2
concentrations peak and decline during the 21st century, the land and ocean begin to take up less carbon in response
to declining atmospheric CO2 concentrations (high confidence) and turn into a weak net source by 2100 under SSP1-1.9
(medium confidence). It is very unlikely that the combined global land and ocean sink will turn into a source by 2100
under scenarios without net negative emissions (SSP2-4.5, SSP3-7.0, SSP5-8.5).32
{4.3, 5.4, 5.5, 5.6, Box TS.5, TS.3.3}

B.4.3 The magnitude of feedbacks between climate change and the carbon cycle becomes larger but also more uncertain
in high CO2 emissions scenarios (very high confidence). However, climate model projections show that the uncertainties in
atmospheric CO2 concentrations by 2100 are dominated by the differences between emissions scenarios (high confidence).
Additional ecosystem responses to warming not yet fully included in climate models, such as CO2 and CH4 fluxes from
wetlands, permafrost thaw and wildfires, would further increase concentrations of these gases in the atmosphere
(high confidence).
{5.4, Box TS.5, TS.3.2}

The proportion of CO₂ emissions taken up by land and ocean carbon sinks
is smaller in scenarios with higher cumulative CO₂ emissions
Total cumulative CO₂ emissions taken up by land and ocean (colours) and remaining in the atmosphere (grey)
under the five illustrative scenarios from 1850 to 2100
GtCO₂
12,000
For scenarios with
higher cumulative
10,000 CO₂ emissions…

8000
ATMOSPHERE

6000
ATMOSPHERE …the amount of CO₂ emissions
taken up by land and ocean
4000 ATMOSPHERE carbon sinks is larger,
OCEAN
but more of the emitted
ATMOSPHERE
ATMOSPHERE
OCEAN CO₂ remains in the
2000 OCEAN
OCEAN atmosphere…
OCEAN

LAND LAND LAND LAND LAND

0 …meaning that the proportion
AT
MOSPHER
AT
MOSPHER MOSPHERE AT
MOSPHER
AT
MOSPHERE of CO₂ emissions taken up by
E E AT E
land and ocean carbon sinks
from the atmosphere
70% 65% 54% 44% 38% is smaller in scenarios
with higher CO₂ emissions.
A A A A A
N

ND ND ND ND ND
L

E E E E E
A

OC OC OC OC OC

SSP1-1.9 SSP1-2.6 SSP2-4.5 SSP3-7.0 SSP5-8.5

Figure SPM.7 | Cumulative anthropogenic CO2 emissions taken up by land and ocean sinks by 2100 under the five illustrative scenarios
The cumulative anthropogenic (human-caused) carbon dioxide (CO2) emissions taken up by the land and ocean sinks under the five illustrative scenarios (SSP1-1.9,
SSP1-2.6, SSP2-4.5, SSP3-7.0 and SSP5-8.5) are simulated from 1850 to 2100 by Coupled Model Intercomparison Project Phase 6 (CMIP6) climate models in the
concentration-driven simulations. Land and ocean carbon sinks respond to past, current and future emissions; therefore, cumulative sinks from 1850 to 2100 are
presented here. During the historical period (1850–2019) the observed land and ocean sink took up 1430 GtCO2 (59% of the emissions).

32 These projected adjustments of carbon sinks to stabilization or decline of atmospheric CO2 are accounted for in calculations of remaining carbon budgets.

20
 Summary for Policymakers

The bar chart illustrates the projected amount of cumulative anthropogenic CO2 emissions (GtCO2) between 1850 and 2100 remaining in the atmosphere (grey
part) and taken up by the land and ocean (coloured part) in the year 2100. The doughnut chart illustrates the proportion of the cumulative anthropogenic
CO2 emissions taken up by the land and ocean sinks and remaining in the atmosphere in the year 2100. Values in % indicate the proportion of the cumulative
anthropogenic CO2 emissions taken up by the combined land and ocean sinks in the year 2100. The overall anthropogenic carbon emissions are calculated by
adding the net global land-use emissions from the CMIP6 scenario database to the other sectoral emissions calculated from climate model runs with prescribed CO2
concentrations.33 Land and ocean CO2 uptake since 1850 is calculated from the net biome productivity on land, corrected for CO2 losses due to land-use change by
adding the land-use change emissions, and net ocean CO2 flux.
{5.2.1; Table 5.1; 5.4.5; Figure 5.25; Box TS.5; Box TS.5, Figure 1} SPM

B.5 Many changes due to past and future greenhouse gas emissions are irreversible for centuries to millennia,
especially changes in the ocean, ice sheets and global sea level.
{2.3, Cross-Chapter Box 2.4, 4.3, 4.5, 4.7, 5.3, 9.2, 9.4, 9.5, 9.6, Box 9.4} (Figure SPM.8)

B.5.1 Past GHG emissions since 1750 have committed the global ocean to future warming (high confidence). Over the rest of
the 21st century, likely ocean warming ranges from 2–4 (SSP1-2.6) to 4–8 times (SSP5-8.5) the 1971–2018 change. Based
on multiple lines of evidence, upper ocean stratification (virtually certain), ocean acidification (virtually certain) and ocean
deoxygenation (high confidence) will continue to increase in the 21st century, at rates dependent on future emissions.
Changes are irreversible on centennial to millennial time scales in global ocean temperature (very high confidence),
deep-ocean acidification (very high confidence) and deoxygenation (medium confidence).
{4.3, 4.5, 4.7, 5.3, 9.2, TS.2.4} (Figure SPM.8)

B.5.2 Mountain and polar glaciers are committed to continue melting for decades or centuries (very high confidence). Loss of
permafrost carbon following permafrost thaw is irreversible at centennial time scales (high confidence). Continued ice
loss over the 21st century is virtually certain for the Greenland Ice Sheet and likely for the Antarctic Ice Sheet. There is
high confidence that total ice loss from the Greenland Ice Sheet will increase with cumulative emissions. There is limited
evidence for low-likelihood, high-impact outcomes (resulting from ice-sheet instability processes characterized by deep
uncertainty and in some cases involving tipping points) that would strongly increase ice loss from the Antarctic Ice Sheet
for centuries under high GHG emissions scenarios.34
{4.3, 4.7, 5.4, 9.4, 9.5, Box 9.4, Box TS.1, TS.2.5}

B.5.3 It is virtually certain that global mean sea level will continue to rise over the 21st century. Relative to 1995–2014, the likely
global mean sea level rise by 2100 is 0.28–0.55 m under the very low GHG emissions scenario (SSP1-1.9); 0.32–0.62 m
under the low GHG emissions scenario (SSP1-2.6); 0.44–0.76 m under the intermediate GHG emissions scenario (SSP2-4.5);
and 0.63–1.01 m under the very high GHG emissions scenario (SSP5-8.5); and by 2150 is 0.37–0.86 m under the very
low scenario (SSP1-1.9); 0.46–0.99 m under the low scenario (SSP1-2.6); 0.66–1.33 m under the intermediate scenario
(SSP2-4.5); and 0.98–1.88 m under the very high scenario (SSP5-8.5) (medium confidence).35 Global mean sea level rise
above the likely range – approaching 2 m by 2100 and 5 m by 2150 under a very high GHG emissions scenario (SSP5-8.5)
(low confidence) – cannot be ruled out due to deep uncertainty in ice-sheet processes.
{4.3, 9.6, Box 9.4, Box TS.4} (Figure SPM.8)

B.5.4 In the longer term, sea level is committed to rise for centuries to millennia due to continuing deep-ocean warming and
ice-sheet melt and will remain elevated for thousands of years (high confidence). Over the next 2000 years, global mean
sea level will rise by about 2 to 3 m if warming is limited to 1.5°C, 2 to 6 m if limited to 2°C and 19 to 22 m with 5°C of
warming, and it will continue to rise over subsequent millennia (low confidence). Projections of multi-millennial global
mean sea level rise are consistent with reconstructed levels during past warm climate periods: likely 5–10 m higher than
today around 125,000 years ago, when global temperatures were very likely 0.5°C–1.5°C higher than 1850–1900; and very
likely 5–25 m higher roughly 3 million years ago, when global temperatures were 2.5°C–4°C higher (medium confidence).
{2.3, Cross-Chapter Box 2.4, 9.6, Box TS.2, Box TS.4, Box TS.9}

33 The other sectoral emissions are calculated as the residual of the net land and ocean CO2 uptake and the prescribed atmospheric CO2 concentration changes in the CMIP6
simulations. These calculated emissions are net emissions and do not separate gross anthropogenic emissions from removals, which are included implicitly.
34 Low-likelihood, high-impact outcomes are those whose probability of occurrence is low or not well known (as in the context of deep uncertainty) but whose potential impacts on
society and ecosystems could be high. A tipping point is a critical threshold beyond which a system reorganizes, often abruptly and/or irreversibly. (Glossary) {1.4, Cross-Chapter Box
1.3, 4.7}
35 To compare to the 1986–2005 baseline period used in AR5 and SROCC, add 0.03 m to the global mean sea level rise estimates. To compare to the 1900 baseline period used in
Figure SPM.8, add 0.16 m.

21
 Summary for Policymakers

Human activities affect all the major climate system components, with
some responding over decades and others over centuries
(a) Global surface temperature change relative to 1850–1900 (e) Global mean sea
°C
level change in 2300
SPM 5
relative to 1900
SSP5-8.5
4 SSP3-7.0 Sea level rise greater than
15 m cannot be ruled out
3 SSP2-4.5
with high emissions

2 SSP1-2.6
SSP1-1.9
1

0 9m
–1
1950 2000 2015 2050 2100

(b) September Arctic sea ice area

8m
106 km2
10

7m
6

SSP1-1.9
2 SSP1-2.6
6m
Practically ice-free SSP2-4.5
SSP3-7.0
0 SSP5-8.5
1950 2000 2015 2050 2100

(c) Global ocean surface pH (a measure of acidity) 5m

8.2

SSP5-8.5
8.1
SSP1-1.9
8.0 SSP1-2.6
7.9 SSP2-4.5
4m
ocean
7.8 acidification
SSP3-7.0
7.7
SSP5-8.5
7.6
1950 2000 2015 2050 2100 3m

(d) Global mean sea level change relative to 1900

m
SSP1-2.6

2 2m

1.5
Low-likelihood, high-impact storyline,
including ice-sheet instability
processes, under SSP5-8.5
1 SSP5-8.5 1m
SSP3-7.0
SSP2-4.5
SSP1-2.6
0.5 SSP1-1.9

0 0m
1950 2000 2020 2050 2100 2300

Figure SPM.8 | Selected indicators of global climate change under the five illustrative scenarios used in this Report
The projections for each of the five scenarios are shown in colour. Shades represent uncertainty ranges – more detail is provided for each panel below. The black
curves represent the historical simulations (panels a, b, c) or the observations (panel d). Historical values are included in all graphs to provide context for the
projected future changes.

22
 Summary for Policymakers

Panel (a) Global surface temperature changes in °C relative to 1850–1900. These changes were obtained by combining Coupled Model Intercomparison
Project Phase 6 (CMIP6) model simulations with observational constraints based on past simulated warming, as well as an updated assessment of equilibrium climate
sensitivity (see Box SPM.1). Changes relative to 1850–1900 based on 20-year averaging periods are calculated by adding 0.85°C (the observed global surface
temperature increase from 1850–1900 to 1995–2014) to simulated changes relative to 1995–2014. Very likely ranges are shown for SSP1-2.6 and SSP3-7.0.
Panel (b) September Arctic sea ice area in 106 km2 based on CMIP6 model simulations. Very likely ranges are shown for SSP1-2.6 and SSP3-7.0. The Arctic is
projected to be practically ice-free near mid-century under intermediate and high GHG emissions scenarios.
Panel (c) Global ocean surface pH (a measure of acidity) based on CMIP6 model simulations. Very likely ranges are shown for SSP1-2.6 and SSP3-7.0.
Panel (d) Global mean sea level change in metres, relative to 1900. The historical changes are observed (from tide gauges before 1992 and altimeters SPM
afterwards), and the future changes are assessed consistently with observational constraints based on emulation of CMIP, ice-sheet, and glacier models. Likely
ranges are shown for SSP1-2.6 and SSP3-7.0. Only likely ranges are assessed for sea level changes due to difficulties in estimating the distribution of deeply
uncertain processes. The dashed curve indicates the potential impact of these deeply uncertain processes. It shows the 83rd percentile of SSP5-8.5 projections that
include low-likelihood, high-impact ice-sheet processes that cannot be ruled out; because of low confidence in projections of these processes, this curve does not
constitute part of a likely range. Changes relative to 1900 are calculated by adding 0.158 m (observed global mean sea level rise from 1900 to 1995–2014) to
simulated and observed changes relative to 1995–2014.
Panel (e) Global mean sea level change at 2300 in metres relative to 1900. Only SSP1-2.6 and SSP5-8.5 are projected at 2300, as simulations that extend
beyond 2100 for the other scenarios are too few for robust results. The 17th–83rd percentile ranges are shaded. The dashed arrow illustrates the 83rd percentile
of SSP5-8.5 projections that include low-likelihood, high-impact ice-sheet processes that cannot be ruled out.
Panels (b) and (c) are based on single simulations from each model, and so include a component of internal variability. Panels (a), (d) and (e) are based on long-term
averages, and hence the contributions from internal variability are small.
{4.3; Figures 4.2, 4.8, and 4.11; 9.6; Figure 9.27; Figures TS.8 and TS.11; Box TS.4, Figure 1}

C. Climate Information for Risk Assessment

and Regional Adaptation
Physical climate information addresses how the climate system responds to the interplay between human influence, natural drivers
and internal variability. Knowledge of the climate response and the range of possible outcomes, including low-likelihood, high
impact outcomes, informs climate services, the assessment of climate-related risks, and adaptation planning. Physical climate
information at global, regional and local scales is developed from multiple lines of evidence, including observational products,
climate model outputs and tailored diagnostics.

C.1 Natural drivers and internal variability will modulate human-caused changes, especially at regional scales and
in the near term, with little effect on centennial global warming. These modulations are important to consider
in planning for the full range of possible changes.
{1.4, 2.2, 3.3, Cross-Chapter Box 3.1, 4.4, 4.6, Cross-Chapter Box 4.1, Box 7.2, 8.3, 8.5, 9.2, 10.3, 10.4, 10.6,
11.3, 12.5, Atlas.4, Atlas.5, Atlas.8, Atlas.9, Atlas.10, Atlas.11, Cross-Chapter Box Atlas.2}

C.1.1 The historical global surface temperature record highlights that decadal variability has both enhanced and masked
underlying human-caused long-term changes, and this variability will continue into the future (very high confidence). For
example, internal decadal variability and variations in solar and volcanic drivers partially masked human-caused surface
global warming during 1998–2012, with pronounced regional and seasonal signatures (high confidence). Nonetheless,
the heating of the climate system continued during this period, as reflected in both the continued warming of the global
ocean (very high confidence) and in the continued rise of hot extremes over land (medium confidence).
{1.4, 3.3, Cross-Chapter Box 3.1, 4.4, Box 7.2, 9.2, 11.3, Cross-Section Box TS.1} (Figure SPM.1)

C.1.2 Projected human-caused changes in mean climate and climatic impact-drivers (CIDs),36 including extremes, will be either
amplified or attenuated by internal variability (high confidence).37 Near-term cooling at any particular location with
respect to present climate could occur and would be consistent with the global surface temperature increase due to
human influence (high confidence).
{1.4, 4.4, 4.6, 10.4, 11.3, 12.5, Atlas.5, Atlas.10, Atlas.11, TS.4.2}

36 Climatic impact-drivers (CIDs) are physical climate system conditions (e.g., means, events, extremes) that affect an element of society or ecosystems. Depending on system tolerance,
CIDs and their changes can be detrimental, beneficial, neutral, or a mixture of each across interacting system elements and regions (Glossary). CID types include heat and cold, wet
and dry, wind, snow and ice, coastal and open ocean.
37 The main internal variability phenomena include El Niño–Southern Oscillation, Pacific Decadal Variability and Atlantic Multi-decadal Variability through their regional influence.

23
 Summary for Policymakers

C.1.3 Internal variability has largely been responsible for the amplification and attenuation of the observed human-caused
decadal-to-multi-decadal mean precipitation changes in many land regions (high confidence). At global and regional
scales, near-term changes in monsoons will be dominated by the effects of internal variability (medium confidence).
In addition to the influence of internal variability, near-term projected changes in precipitation at global and regional
scales are uncertain because of model uncertainty and uncertainty in forcings from natural and anthropogenic aerosols
(medium confidence).
SPM {1.4, 4.4, 8.3, 8.5, 10.3, 10.4, 10.5, 10.6, Atlas.4, Atlas.8, Atlas.9, Atlas.10, Atlas.11, Cross-Chapter Box Atlas.2, TS.4.2,
Box TS.6, Box TS.13}

C.1.4 Based on paleoclimate and historical evidence, it is likely that at least one large explosive volcanic eruption would occur
during the 21st century.38 Such an eruption would reduce global surface temperature and precipitation, especially over land,
for one to three years, alter the global monsoon circulation, modify extreme precipitation and change many CIDs (medium
confidence). If such an eruption occurs, this would therefore temporarily and partially mask human-caused climate change.
{2.2, 4.4, Cross-Chapter Box 4.1, 8.5, TS.2.1}

C.2 With further global warming, every region is projected to increasingly experience concurrent and multiple
changes in climatic impact-drivers. Changes in several climatic impact-drivers would be more widespread
at 2°C compared to 1.5°C global warming and even more widespread and/or pronounced for higher
warming levels.
{8.2, 9.3, 9.5, 9.6, Box 10.3, 11.3, 11.4, 11.5, 11.6, 11.7, 11.9, Box 11.3, Box 11.4, Cross-Chapter Box 11.1, 12.2,
12.3, 12.4, 12.5, Cross-Chapter Box 12.1, Atlas.4, Atlas.5, Atlas.6, Atlas.7, Atlas.8, Atlas.9, Atlas.10, Atlas.11}
(Table SPM.1, Figure SPM.9)

C.2.1 All regions39 are projected to experience further increases in hot climatic impact-drivers (CIDs) and decreases in cold
CIDs (high confidence). Further decreases are projected in permafrost; snow, glaciers and ice sheets; and lake and Arctic
sea ice (medium to high confidence).40 These changes would be larger at 2°C global warming or above than at 1.5°C
(high confidence). For example, extreme heat thresholds relevant to agriculture and health are projected to be exceeded
more frequently at higher global warming levels (high confidence).
{9.3, 9.5, 11.3, 11.9, Cross-Chapter Box 11.1, 12.3, 12.4, 12.5, Cross-Chapter Box 12.1, Atlas.4, Atlas.5, Atlas.6, Atlas.7,
Atlas.8, Atlas.9, Atlas.10, Atlas.11, TS.4.3} (Table SPM.1, Figure SPM.9)

C.2.2 At 1.5°C global warming, heavy precipitation and associated flooding are projected to intensify and be more frequent
in most regions in Africa and Asia (high confidence), North America (medium to high confidence)40 and Europe (medium
confidence). Also, more frequent and/or severe agricultural and ecological droughts are projected in a few regions in all
inhabited continents except Asia compared to 1850–1900 (medium confidence); increases in meteorological droughts are
also projected in a few regions (medium confidence). A small number of regions are projected to experience increases or
decreases in mean precipitation (medium confidence).
{11.4, 11.5, 11.6, 11.9, Atlas.4, Atlas.5, Atlas.7, Atlas.8, Atlas.9, Atlas.10, Atlas.11, TS.4.3} (Table SPM.1)

C.2.3 At 2°C global warming and above, the level of confidence in and the magnitude of the change in droughts and heavy
and mean precipitation increase compared to those at 1.5°C. Heavy precipitation and associated flooding events
are projected to become more intense and frequent in the Pacific Islands and across many regions of North America
and Europe (medium to high confidence).40 These changes are also seen in some regions in Australasia and Central and
South America (medium confidence). Several regions in Africa, South America and Europe are projected to experience an
increase in frequency and/or severity of agricultural and ecological droughts with medium to high confidence;40 increases
are also projected in Australasia, Central and North America, and the Caribbean with medium confidence. A small number
of regions in Africa, Australasia, Europe and North America are also projected to be affected by increases in hydrological
droughts, and several regions are projected to be affected by increases or decreases in meteorological droughts, with
more regions displaying an increase (medium confidence). Mean precipitation is projected to increase in all polar, northern
European and northern North American regions, most Asian regions and two regions of South America (high confidence).
{11.4, 11.6, 11.9, Cross-Chapter Box 11.1, 12.4, 12.5, Cross-Chapter Box 12.1, Atlas.5, Atlas.7, Atlas.8, Atlas.9, Atlas.11,
TS.4.3} (Table SPM.1, Figure SPM.5, Figure SPM.6, Figure SPM.9)

38 Based on 2500 year reconstructions, eruptions more negative than –1 W m–2 occur on average twice per century.
39 Regions here refer to the AR6 WGI reference regions used in this Report to summarize information in sub-continental and oceanic regions. Changes are compared to averages over
the last 20–40 years unless otherwise specified. {1.4, 12.4, Atlas.1}.
40 The specific level of confidence or likelihood depends on the region considered. Details can be found in the Technical Summary and the underlying Report.

24
 Summary for Policymakers

C.2.4 More CIDs across more regions are projected to change at 2°C and above compared to 1.5°C global warming
(high confidence). Region-specific changes include intensification of tropical cyclones and/or extratropical storms
(medium confidence), increases in river floods (medium to high confidence),40 reductions in mean precipitation and
increases in aridity (medium to high confidence),40 and increases in fire weather (medium to high confidence).40 There
is low confidence in most regions in potential future changes in other CIDs, such as hail, ice storms, severe storms, dust
storms, heavy snowfall and landslides.
{11.7, 11.9, Cross-Chapter Box 11.1, 12.4, 12.5, Cross-Chapter Box 12.1, Atlas.4, Atlas.6, Atlas.7, Atlas.8, Atlas.10, TS.4.3.1, SPM
TS.4.3.2, TS.5} (Table SPM.1, Figure SPM.9)

C.2.5 It is very likely to virtually certain40 that regional mean relative sea level rise will continue throughout the 21st century,
except in a few regions with substantial geologic land uplift rates. Approximately two-thirds of the global coastline has
a projected regional relative sea level rise within ±20% of the global mean increase (medium confidence). Due to relative
sea level rise, extreme sea level events that occurred once per century in the recent past are projected to occur at least
annually at more than half of all tide gauge locations by 2100 (high confidence). Relative sea level rise contributes to
increases in the frequency and severity of coastal flooding in low-lying areas and to coastal erosion along most sandy
coasts (high confidence).
{9.6, 12.4, 12.5, Cross-Chapter Box 12.1, Box TS.4, TS.4.3} (Figure SPM.9)

C.2.6 Cities intensify human-induced warming locally, and further urbanization together with more frequent hot extremes will
increase the severity of heatwaves (very high confidence). Urbanization also increases mean and heavy precipitation
over and/or downwind of cities (medium confidence) and resulting runoff intensity (high confidence). In coastal cities,
the combination of more frequent extreme sea level events (due to sea level rise and storm surge) and extreme rainfall/
riverflow events will make flooding more probable (high confidence).
{8.2, Box 10.3, 11.3, 12.4, Box TS.14}

C.2.7 Many regions are projected to experience an increase in the probability of compound events with higher global warming
(high confidence). In particular, concurrent heatwaves and droughts are likely to become more frequent. Concurrent
extremes at multiple locations, including in crop-producing areas, become more frequent at 2°C and above compared to
1.5°C global warming (high confidence).
{11.8, Box 11.3, Box 11.4, 12.3, 12.4, Cross-Chapter Box 12.1, TS.4.3} (Table SPM.1)

25
 Summary for Policymakers

Multiple climatic impact-drivers are projected to change in all regions

of the world
Climatic impact-drivers (CIDs) are physical climate system conditions (e.g., means, events, extremes) that affect an element
of society or ecosystems. Depending on system tolerance, CIDs and their changes can be detrimental, beneficial, neutral,
or a mixture of each across interacting system elements and regions. The CIDs are grouped into seven types, which are
summarized under the icons in the figure. All regions are projected to experience changes in at least 5 CIDs. Almost all
(96%) are projected to experience changes in at least 10 CIDs and half in at least 15 CIDs. For many CID changes, there is
SPM wide geographical variation, and so each region is projected to experience a specific set of CID changes. Each bar in the
chart represents a specific geographical set of changes that can be explored in the WGI Interactive Atlas. interactive-atlas.ipcc.ch

Number of land & coastal regions (a) and open-ocean regions (b) where each climatic impact-driver (CID) is projected
to increase or decrease with high confidence (dark shade) or medium confidence (light shade)

(a) (b)
Heat and Cold Wet and Dry Wind Snow and Ice Other Coastal Open Ocean
Heavy precipitation and pluvial flood

Agricultural and ecological drought

& COASTAL REGIONS

NUMBER OF OPEN-OCEAN REGIONS

Heavy snowfall and ice storm

Atmospheric CO2 at surface

Snow, glacier and ice sheet
Mean surface temperature

Mean ocean temperature

Lake, river and sea ice

Air pollution weather

Hydrological drought

Sand and dust storm

Radiation at surface
Mean precipitation

Severe wind storm

Mean wind speed

Dissolved oxygen
Marine heatwave

Marine heatwave
Relative sea level
Snow avalanche
Tropical cyclone

Coastal erosion

Ocean salinity
Ocean acidity

Ocean acidity
NUMBER OF LAND

Extreme heat

Coastal flood
Fire weather

Permafrost
River flood
Cold spell

Landslide
Aridity
Frost

Hail

55

45

35

25

15 15

5 5

5 5

15 15

25

35

45

55

BAR CHART LEGEND LIGHTER-SHADED ‘ENVELOPE’ LEGEND ASSESSED FUTURE CHANGES

Regions with high confidence increase The height of the lighter shaded ‘envelope’ behind each bar Changes refer to a 20–30
Regions with medium confidence increase represents the maximum number of regions for which each year period centred around
CID is relevant. The envelope is symmetrical about the x-axis 2050 and/or consistent
Regions with high confidence decrease showing the maximum possible number of relevant regions with 2°C global warming
Regions with medium confidence decrease for CID increase (upper part) or decrease (lower part). compared to a similar
period within 1960–2014
or 1850–1900.

Figure SPM.9 | Synthesis of the number of AR6 WGI reference regions where climatic impact-drivers are projected to change
A total of 35 climatic impact-drivers (CIDs) grouped into seven types are shown: heat and cold; wet and dry; wind; snow and ice; coastal; open ocean; and other.
For each CID, the bar in the graph below displays the number of AR6 WGI reference regions where it is projected to change. The colours represent the direction
of change and the level of confidence in the change: purple indicates an increase while brown indicates a decrease; darker and lighter shades refer to high and
medium confidence, respectively. Lighter background colours represent the maximum number of regions for which each CID is broadly relevant.
Panel (a) shows the 30 CIDs relevant to the land and coastal regions, while panel (b) shows the five CIDs relevant to the open-ocean regions. Marine heatwaves
and ocean acidity are assessed for coastal ocean regions in panel (a) and for open-ocean regions in panel (b). Changes refer to a 20–30-year period centred around 2050
and/or consistent with 2°C global warming compared to a similar period within 1960–2014, except for hydrological drought and agricultural and ecological drought, which
is compared to 1850–1900. Definitions of the regions are provided in Sections 12.4 and Atlas.1 and the Interactive Atlas (see https://interactive-atlas.ipcc.ch/).
{11.9, 12.2, 12.4, Atlas.1, Table TS.5, Figures TS.22 and TS.25} (Table SPM.1)

26
 Summary for Policymakers

C.3 Low-likelihood outcomes, such as ice-sheet collapse, abrupt ocean circulation changes, some compound
extreme events, and warming substantially larger than the assessed very likely range of future warming,
cannot be ruled out and are part of risk assessment.
{1.4, Cross-Chapter Box 1.3, 4.3, 4.4, 4.8, Cross-Chapter Box 4.1, 8.6, 9.2, Box 9.4, 11.8, Box 11.2, Cross-Chapter
Box 12.1} (Table SPM.1)

C.3.1 If global warming exceeds the assessed very likely range for a given GHG emissions scenario, including low GHG emissions SPM
scenarios, global and regional changes in many aspects of the climate system, such as regional precipitation and other
CIDs, would also exceed their assessed very likely ranges (high confidence). Such low-likelihood, high-warming outcomes
are associated with potentially very large impacts, such as through more intense and more frequent heatwaves and heavy
precipitation, and high risks for human and ecological systems, particularly for high GHG emissions scenarios.
{Cross-Chapter Box 1.3, 4.3, 4.4, 4.8, Box 9.4, Box 11.2, Cross-Chapter Box 12.1, TS.1.4, Box TS.3, Box TS.4} (Table SPM.1)

C.3.2 Low-likelihood, high-impact outcomes34 could occur at global and regional scales even for global warming within the
very likely range for a given GHG emissions scenario. The probability of low-likelihood, high-impact outcomes increases
with higher global warming levels (high confidence). Abrupt responses and tipping points of the climate system, such as
strongly increased Antarctic ice-sheet melt and forest dieback, cannot be ruled out (high confidence).
{1.4, 4.3, 4.4, 4.8, 5.4, 8.6, Box 9.4, Cross-Chapter Box 12.1, TS.1.4, TS.2.5, Box TS.3, Box TS.4, Box TS.9} (Table SPM.1)

C.3.3 If global warming increases, some compound extreme events18 with low likelihood in past and current climate will become
more frequent, and there will be a higher likelihood that events with increased intensities, durations and/or spatial extents
unprecedented in the observational record will occur (high confidence).
{11.8, Box 11.2, Cross-Chapter Box 12.1, Box TS.3, Box TS.9}

C.3.4 The Atlantic Meridional Overturning Circulation is very likely to weaken over the 21st century for all emissions scenarios.
While there is high confidence in the 21st century decline, there is only low confidence in the magnitude of the trend.
There is medium confidence that there will not be an abrupt collapse before 2100. If such a collapse were to occur, it
would very likely cause abrupt shifts in regional weather patterns and water cycle, such as a southward shift in the
tropical rain belt, weakening of the African and Asian monsoons and strengthening of Southern Hemisphere monsoons,
and drying in Europe.
{4.3, 8.6, 9.2, TS2.4, Box TS.3}

C.3.5 Unpredictable and rare natural events not related to human influence on climate may lead to low-likelihood, high-impact
outcomes. For example, a sequence of large explosive volcanic eruptions within decades has occurred in the past, causing
substantial global and regional climate perturbations over several decades. Such events cannot be ruled out in the future,
but due to their inherent unpredictability they are not included in the illustrative set of scenarios referred to in this Report
{2.2, Cross-Chapter Box 4.1, Box TS.3} (Box SPM.1)

D. Limiting Future Climate Change

Since AR5, estimates of remaining carbon budgets have been improved by a new methodology first presented in SR1.5, updated
evidence, and the integration of results from multiple lines of evidence. A comprehensive range of possible future air pollution
controls in scenarios is used to consistently assess the effects of various assumptions on projections of climate and air pollution.
A novel development is the ability to ascertain when climate responses to emissions reductions would become discernible above
natural climate variability, including internal variability and responses to natural drivers.

D.1 From a physical science perspective, limiting human-induced global warming to a specific level requires
limiting cumulative CO2 emissions, reaching at least net zero CO2 emissions, along with strong reductions in
other greenhouse gas emissions. Strong, rapid and sustained reductions in CH4 emissions would also limit the
warming effect resulting from declining aerosol pollution and would improve air quality.
{3.3, 4.6, 5.1, 5.2, 5.4, 5.5, 5.6, Box 5.2, Cross-Chapter Box 5.1, 6.7, 7.6, 9.6} (Figure SPM.10, Table SPM.2)

27
 Summary for Policymakers

D.1.1 This Report reaffirms with high confidence the AR5 finding that there is a near-linear relationship between cumulative
anthropogenic CO2 emissions and the global warming they cause. Each 1000 GtCO2 of cumulative CO2 emissions is assessed
to likely cause a 0.27°C to 0.63°C increase in global surface temperature with a best estimate of 0.45°C.41 This is a narrower
range compared to AR5 and SR1.5. This quantity is referred to as the transient climate response to cumulative CO2 emissions
(TCRE). This relationship implies that reaching net zero anthropogenic CO2 emissions42 is a requirement to stabilize
human-induced global temperature increase at any level, but that limiting global temperature increase to a specific level
SPM would imply limiting cumulative CO2 emissions to within a carbon budget.43 {5.4, 5.5, TS.1.3, TS.3.3, Box TS.5} (Figure SPM.10)

Every tonne of CO₂ emissions adds to global warming

Global surface temperature increase since 1850–1900 (OC) as a function of cumulative CO₂ emissions (GtCO₂)
OC
3

SSP5-8.5
The near-linear relationship
2.5 SSP3-7.0
between the cumulative
CO₂ emissions and global SSP2-4.5
warming for five illustrative
scenarios until year 2050 SSP1-2.6
2
SSP1-1.9

1.5

Historical global
warming
0.5

Cumulative CO₂ emissions since 1850

0
1000 2000 3000 4000 4500 GtCO₂

–0.5 Future cumulative

SSP1-1.9 CO₂ emissions differ
SSP1-2.6 across scenarios and
SSP2-4.5 determine how much
SSP3-7.0 warming we will
SSP5-8.5 experience.
time
1850
1900

1950

2000

2019
2020

2030

2040

2050

HISTORICAL PROJECTIONS
Cumulative CO₂ emissions between 1850 and 2019 Cumulative CO₂ emissions between 2020 and 2050

Figure SPM.10 | Near-linear relationship between cumulative CO2 emissions and the increase in global surface temperature
Top panel: Historical data (thin black line) shows observed global surface temperature increase in °C since 1850–1900 as a function of historical cumulative carbon
dioxide (CO2) emissions in GtCO2 from 1850 to 2019. The grey range with its central line shows a corresponding estimate of the historical human-caused surface
warming (see Figure SPM.2). Coloured areas show the assessed very likely range of global surface temperature projections, and thick coloured central lines show the
median estimate as a function of cumulative CO2 emissions from 2020 until year 2050 for the set of illustrative scenarios (SSP1-1.9, SSP1-2.6, SSP2-4.5, SSP3-7.0, and
SSP5-8.5; see Figure SPM.4). Projections use the cumulative CO2 emissions of each respective scenario, and the projected global warming includes the contribution
from all anthropogenic forcers. The relationship is illustrated over the domain of cumulative CO2 emissions for which there is high confidence that the transient climate
response to cumulative CO2 emissions (TCRE) remains constant, and for the time period from 1850 to 2050 over which global CO2 emissions remain net positive under
all illustrative scenarios, as there is limited evidence supporting the quantitative application of TCRE to estimate temperature evolution under net negative CO2 emissions.
Bottom panel: Historical and projected cumulative CO2 emissions in GtCO2 for the respective scenarios.
{Section 5.5, Figure 5.31, Figure TS.18}

41 In the literature, units of °C per 1000 PgC (petagrams of carbon) are used, and the AR6 reports the TCRE likely range as 1.0°C to 2.3°C per 1000 PgC in the underlying report, with
a best estimate of 1.65°C.
42 The condition in which anthropogenic carbon dioxide (CO2) emissions are balanced by anthropogenic CO2 removals over a specified period (Glossary).
43 The term ‘carbon budget’ refers to the maximum amount of cumulative net global anthropogenic CO2 emissions that would result in limiting global warming to a given level with
a given probability, taking into account the effect of other anthropogenic climate forcers. This is referred to as the total carbon budget when expressed starting from the pre-industrial
period, and as the remaining carbon budget when expressed from a recent specified date (Glossary). Historical cumulative CO2 emissions determine to a large degree warming to
date, while future emissions cause future additional warming. The remaining carbon budget indicates how much CO2 could still be emitted while keeping warming below a specific
temperature level.

28
 Summary for Policymakers

D.1.2 Over the period 1850–2019, a total of 2390 ± 240 (likely range) GtCO2 of anthropogenic CO2 was emitted. Remaining
carbon budgets have been estimated for several global temperature limits and various levels of probability, based on the
estimated value of TCRE and its uncertainty, estimates of historical warming, variations in projected warming from non-
CO2 emissions, climate system feedbacks such as emissions from thawing permafrost, and the global surface temperature
change after global anthropogenic CO2 emissions reach net zero.
{5.1, 5.5, Box 5.2, TS.3.3} (Table SPM.2)
SPM

Table SPM.2 | Estimates of historical carbon dioxide (CO2) emissions and remaining carbon budgets. Estimated remaining carbon budgets are
calculated from the beginning of 2020 and extend until global net zero CO2 emissions are reached. They refer to CO2 emissions, while accounting for the global
warming effect of non-CO2 emissions. Global warming in this table refers to human-induced global surface temperature increase, which excludes the impact
of natural variability on global temperatures in individual years.
{Table 3.1, 5.5.1, 5.5.2, Box 5.2, Table 5.1, Table 5.7, Table 5.8, Table TS.3}

Global Warming Between

Historical Cumulative CO2 Emissions from 1850 to 2019 (GtCO2)
1850–1900 and 2010–2019 (°C)
1.07 (0.8–1.3; likely range) 2390 (± 240; likely range)

Estimated remaining carbon budgets

Approximate global Additional global from the beginning of 2020 (GtCO2)
warming relative warming relative to Variations in reductions
to 1850–1900 until 2010–2019 until tem- Likelihood of limiting global warming in non-CO2 emissionsc
temperature limit (°C)a perature limit (°C) to temperature limitb

17% 33% 50% 67% 83%

1.5 0.43 900 650 500 400 300

Higher or lower reductions in
accompanying non-CO2 emissions can
1.7 0.63 1450 1050 850 700 550
increase or decrease the values on
the left by 220 GtCO2 or more
2.0 0.93 2300 1700 1350 1150 900

a
Values at each 0.1°C increment of warming are available in Tables TS.3 and 5.8.
b
This likelihood is based on the uncertainty in transient climate response to cumulative CO2 emissions (TCRE) and additional Earth system feedbacks and provides the
probability that global warming will not exceed the temperature levels provided in the two left columns. Uncertainties related to historical warming (±550 GtCO2)
and non-CO2 forcing and response (±220 GtCO2) are partially addressed by the assessed uncertainty in TCRE, but uncertainties in recent emissions since 2015
(±20 GtCO2) and the climate response after net zero CO2 emissions are reached (±420 GtCO2) are separate.
c
Remaining carbon budget estimates consider the warming from non-CO2 drivers as implied by the scenarios assessed in SR1.5. The Working Group III Contribution
to AR6 will assess mitigation of non-CO2 emissions.

D.1.3 Several factors that determine estimates of the remaining carbon budget have been re-assessed, and updates to these
factors since SR1.5 are small. When adjusted for emissions since previous reports, estimates of remaining carbon budgets
are therefore of similar magnitude compared to SR1.5 but larger compared to AR5 due to methodological improvements.44
{5.5, Box 5.2, TS.3.3} (Table SPM.2)

D.1.4 Anthropogenic CO2 removal (CDR) has the potential to remove CO2 from the atmosphere and durably store it in reservoirs
(high confidence). CDR aims to compensate for residual emissions to reach net zero CO2 or net zero GHG emissions or, if
implemented at a scale where anthropogenic removals exceed anthropogenic emissions, to lower surface temperature.
CDR methods can have potentially wide-ranging effects on biogeochemical cycles and climate, which can either weaken
or strengthen the potential of these methods to remove CO2 and reduce warming, and can also influence water availability
and quality, food production and biodiversity45 (high confidence).
{5.6, Cross-Chapter Box 5.1, TS.3.3}

D.1.5 Anthropogenic CO2 removal (CDR) leading to global net negative emissions would lower the atmospheric CO2 concentration
and reverse surface ocean acidification (high confidence). Anthropogenic CO2 removals and emissions are partially

44 Compared to AR5, and when taking into account emissions since AR5, estimates in AR6 are about 300–350 GtCO2 larger for the remaining carbon budget consistent with limiting
warming to 1.5°C; for 2°C, the difference is about 400–500 GtCO2.
45 Potential negative and positive effects of CDR for biodiversity, water and food production are methods-specific and are often highly dependent on local context, management, prior
land use, and scale. IPCC Working Groups II and III assess the CDR potential and ecological and socio-economic effects of CDR methods in their AR6 contributions.

29
 Summary for Policymakers

compensated by CO2 release and uptake respectively, from or to land and ocean carbon pools (very high confidence).
CDR would lower atmospheric CO2 by an amount approximately equal to the increase from an anthropogenic emission of
the same magnitude (high confidence). The atmospheric CO2 decrease from anthropogenic CO2 removals could be up to
10% less than the atmospheric CO2 increase from an equal amount of CO2 emissions, depending on the total amount of
CDR (medium confidence).
{5.3, 5.6, TS.3.3}
SPM
D.1.6 If global net negative CO2 emissions were to be achieved and be sustained, the global CO2-induced surface temperature
increase would be gradually reversed but other climate changes would continue in their current direction for decades to
millennia (high confidence). For instance, it would take several centuries to millennia for global mean sea level to reverse
course even under large net negative CO2 emissions (high confidence).
{4.6, 9.6, TS.3.3}

D.1.7 In the five illustrative scenarios, simultaneous changes in CH4, aerosol and ozone precursor emissions, which also
contribute to air pollution, lead to a net global surface warming in the near and long term (high confidence). In the
long term, this net warming is lower in scenarios assuming air pollution controls combined with strong and sustained
CH4 emissions reductions (high confidence). In the low and very low GHG emissions scenarios, assumed reductions in
anthropogenic aerosol emissions lead to a net warming, while reductions in CH4 and other ozone precursor emissions
lead to a net cooling. Because of the short lifetime of both CH4 and aerosols, these climate effects partially counterbalance
each other, and reductions in CH4 emissions also contribute to improved air quality by reducing global surface ozone
(high confidence).
{6.7, Box TS.7} (Figure SPM.2, Box SPM.1)

D.1.8 Achieving global net zero CO2 emissions, with anthropogenic CO2 emissions balanced by anthropogenic removals of
CO2, is a requirement for stabilizing CO2-induced global surface temperature increase. This is different from achieving
net zero GHG emissions, where metric-weighted anthropogenic GHG emissions equal metric-weighted anthropogenic
GHG removals. For a given GHG emissions pathway, the pathways of individual GHGs determine the resulting climate
response,46 whereas the choice of emissions metric47 used to calculate aggregated emissions and removals of different
GHGs affects what point in time the aggregated GHGs are calculated to be net zero. Emissions pathways that reach and
sustain net zero GHG emissions defined by the 100-year global warming potential are projected to result in a decline in
surface temperature after an earlier peak (high confidence).
{4.6, 7.6, Box 7.3, TS.3.3}

D.2 Scenarios with very low or low GHG emissions (SSP1-1.9 and SSP1-2.6) lead within years to discernible effects
on greenhouse gas and aerosol concentrations and air quality, relative to high and very high GHG emissions
scenarios (SSP3-7.0 or SSP5-8.5). Under these contrasting scenarios, discernible differences in trends of global
surface temperature would begin to emerge from natural variability within around 20 years, and over longer
time periods for many other climatic impact-drivers (high confidence).
{4.6, 6.6, 6.7, Cross-Chapter Box 6.1, 9.6, 11.2, 11.4, 11.5, 11.6, Cross-Chapter Box 11.1, 12.4, 12.5} (Figure
SPM.8, Figure SPM.10)

D.2.1 Emissions reductions in 2020 associated with measures to reduce the spread of COVID-19 led to temporary but detectable
effects on air pollution (high confidence) and an associated small, temporary increase in total radiative forcing, primarily
due to reductions in cooling caused by aerosols arising from human activities (medium confidence). Global and regional
climate responses to this temporary forcing are, however, undetectable above natural variability (high confidence).
Atmospheric CO2 concentrations continued to rise in 2020, with no detectable decrease in the observed CO2 growth rate
(medium confidence).48
{Cross-Chapter Box 6.1, TS.3.3}

D.2.2 Reductions in GHG emissions also lead to air quality improvements. However, in the near term,49 even in scenarios with
strong reduction of GHGs, as in the low and very low GHG emissions scenarios (SSP1-2.6 and SSP1-1.9), these improvements

46 A general term for how the climate system responds to a radiative forcing (Glossary).
47 The choice of emissions metric depends on the purposes for which gases or forcing agents are being compared. This Report contains updated emissions metric values and assesses
new approaches to aggregating gases.
48 For other GHGs, there was insufficient literature available at the time of the assessment to assess detectable changes in their atmospheric growth rate during 2020.
49 Near term: 2021–2040.

30
 Summary for Policymakers

are not sufficient in many polluted regions to achieve air quality guidelines specified by the World Health Organization
(high confidence). Scenarios with targeted reductions of air pollutant emissions lead to more rapid improvements in air
quality within years compared to reductions in GHG emissions only, but from 2040, further improvements are projected
in scenarios that combine efforts to reduce air pollutants as well as GHG emissions, with the magnitude of the benefit
varying between regions (high confidence).
{6.6, 6.7, Box TS.7}.
SPM
D.2.3 Scenarios with very low or low GHG emissions (SSP1-1.9 and SSP1-2.6) would have rapid and sustained effects to limit
human-caused climate change, compared with scenarios with high or very high GHG emissions (SSP3-7.0 or SSP5-8.5),
but early responses of the climate system can be masked by natural variability. For global surface temperature, differences
in 20-year trends would likely emerge during the near term under a very low GHG emissions scenario (SSP1-1.9), relative
to a high or very high GHG emissions scenario (SSP3-7.0 or SSP5-8.5). The response of many other climate variables would
emerge from natural variability at different times later in the 21st century (high confidence).
{4.6, Cross-Section Box TS.1} (Figure SPM.8, Figure SPM.10)

D.2.4 Scenarios with very low and low GHG emissions (SSP1-1.9 and SSP1-2.6) would lead to substantially smaller changes
in a range of CIDs36 beyond 2040 than under high and very high GHG emissions scenarios (SSP3-7.0 and SSP5-8.5).
By the end of the century, scenarios with very low and low GHG emissions would strongly limit the change of several
CIDs, such as the increases in the frequency of extreme sea level events, heavy precipitation and pluvial flooding, and
exceedance of dangerous heat thresholds, while limiting the number of regions where such exceedances occur, relative
to higher GHG emissions scenarios (high confidence). Changes would also be smaller in very low compared to low GHG
emissions scenarios, as well as for intermediate (SSP2-4.5) compared to high or very high GHG emissions scenarios (high
confidence).
{9.6, 11.2, 11.3, 11.4, 11.5, 11.6, 11.9, Cross-Chapter Box 11.1, 12.4, 12.5, TS.4.3}

31
Foreword
Technical
and Preface
Summary

33
34
TS Technical Summary

Coordinating Authors:
Paola A. Arias (Colombia), Nicolas Bellouin (United Kingdom/France), Erika Coppola (Italy),
Richard G. Jones (United Kingdom), Gerhard Krinner (France/Germany, France), Jochem Marotzke
(Germany), Vaishali Naik (United States of America), Matthew D. Palmer (United Kingdom),
Gian-Kasper Plattner (Switzerland), Joeri Rogelj (United Kingdom/Belgium), Maisa Rojas (Chile),
Jana Sillmann (Norway/Germany), Trude Storelvmo (Norway), Peter W. Thorne (Ireland/United
Kingdom), Blair Trewin (Australia)

Authors:
Krishna Achuta Rao (India), Bhupesh Adhikary (Nepal), Richard P. Allan (United Kingdom),
Kyle Armour (United States of America), Govindasamy Bala (India/United States of America),
Rondrotiana Barimalala (South Africa/Madagascar), Sophie Berger (France/Belgium),
Josep G. Canadell (Australia), Christophe Cassou (France), Annalisa Cherchi (Italy), William Collins
(United Kingdom), William D. Collins (United States of America), Sarah L. Connors (France/United
Kingdom), Susanna Corti (Italy), Faye Cruz (Philippines), Frank J. Dentener (EU/The Netherlands),
Claudine Dereczynski (Brazil), Alejandro Di Luca (Australia, Canada/Argentina), Aida Diongue
Niang (Senegal), Francisco J. Doblas-Reyes (Spain), Alessandro Dosio (Italy), Hervé Douville (France),
François Engelbrecht (South Africa), Veronika Eyring (Germany), Erich Fischer (Switzerland), Piers
Forster (United Kingdom), Baylor Fox-Kemper (United States of America), Jan S. Fuglestvedt
(Norway), John C. Fyfe (Canada), Nathan P. Gillett (Canada), Leah Goldfarb (France/United States
of America), Irina Gorodetskaya (Portugal/Russian Federation, Belgium), Jose Manuel Gutierrez
(Spain), Rafiq Hamdi (Belgium), Ed Hawkins (United Kingdom), Helene T. Hewitt (United Kingdom),
Pandora Hope (Australia), AKM Saiful Islam (Bangladesh), Christopher Jones (United Kingdom),
Darrell S. Kaufman (United States of America), Robert E. Kopp (United States of America),
Yu Kosaka (Japan), James Kossin (United States of America), Svitlana Krakovska (Ukraine),
June-Yi Lee (Republic of Korea), Jian Li (China), Thorsten Mauritsen (Sweden, Denmark), Thomas
K. Maycock (United States of America), Malte Meinshausen (Australia/Germany), Seung-Ki Min
(Republic of Korea), Pedro M. S. Monteiro (South Africa), Thanh Ngo-Duc (Vietnam), Friederike
Otto (United Kingdom/Germany), Izidine Pinto (South Africa/Mozambique), Anna Pirani (Italy),
Krishnan Raghavan (India), Roshanka Ranasinghe (The Netherlands/Sri Lanka, Australia), Alex
C. Ruane (United States of America), Lucas Ruiz (Argentina), Jean-Baptiste Sallée (France), Bjørn
H. Samset (Norway), Shubha Sathyendranath (UK/Canada, United Kingdom, Overseas Citizen of
India), Sonia I. Seneviratne (Switzerland), Anna A. Sörensson (Argentina), Sophie Szopa (France),
Izuru Takayabu (Japan), Anne-Marie Treguier (France), Bart van den Hurk (The Netherlands),

35
 Technical Summary

Robert Vautard (France), Karina von Schuckmann (France/Germany), Sönke Zaehle (Germany),
Xuebin Zhang (Canada), Kirsten Zickfeld (Canada/Germany)

Contributing Authors:
Guðfinna Aðalgeirsdóttir (Iceland), Lincoln M. Alves (Brazil), Terje Berntsen (Norway),
Sara M. Blichner (Norway), Lisa Bock (Germany), Gregory G. Garner (United States of America),
Joelle Gergis (Australia), Sergey K. Gulev (Russian Federation), Mathias Hauser (Switzerland),
Flavio Lehner (United States of America/Switzerland), Chao Li (China), Marianne T. Lund
(Norway), Daniel J. Lunt (United Kingdom), Sebastian Milinski (Germany), Gemma Teresa
Narisma (Philippines), Zebedee R. J. Nicholls (Australia), Dirk Notz (Germany), Sophie Nowicki
(United States of America/France, United States of America), Bette Otto-Bliesner (United States
of America), Brodie Pearson (United States of America/United Kingdom), Adam S. Phillips
(United States of America), James Renwick (New Zealand), Stéphane Sénési (France), Lucas Silva
(Portugal/Switzerland), Aimee B. A. Slangen (The Netherlands), Thomas F. Stocker (Switzerland),
TS Claudia Tebaldi (United States of America), Laurent Terray (France), Sabin Thazhe Purayil (India),
Andrew Turner (United Kingdom), Steven Turnock (United Kingdom), Carolina Vera (Argentina),
Cunde Xiao (China), Panmao Zhai (China)

Review Editors:
Valérie Masson-Delmotte (France), Gregory M. Flato (Canada), Noureddine Yassa (Algeria)

This Technical Summary should be cited as:

Arias, P.A., N. Bellouin, E. Coppola, R.G. Jones, G. Krinner, J. Marotzke, V. Naik, M.D. Palmer, G.-K. Plattner, J. Rogelj,
M. Rojas, J. Sillmann, T. Storelvmo, P.W. Thorne, B. Trewin, K. Achuta Rao, B. Adhikary, R.P. Allan, K. Armour, G. Bala,
R. Barimalala, S. Berger, J.G. Canadell, C. Cassou, A. Cherchi, W. Collins, W.D. Collins, S.L. Connors, S. Corti, F. Cruz,
F.J. Dentener, C. Dereczynski, A. Di Luca, A. Diongue Niang, F.J. Doblas-Reyes, A. Dosio, H. Douville, F. Engelbrecht,
V. Eyring, E. Fischer, P. Forster, B. Fox-Kemper, J.S. Fuglestvedt, J.C. Fyfe, N.P. Gillett, L. Goldfarb, I. Gorodetskaya,
J.M. Gutierrez, R. Hamdi, E. Hawkins, H.T. Hewitt, P. Hope, A.S. Islam, C. Jones, D.S. Kaufman, R.E. Kopp, Y. Kosaka,
J. Kossin, S. Krakovska, J.-Y. Lee, J. Li, T. Mauritsen, T.K. Maycock, M. Meinshausen, S.-K. Min, P.M.S. Monteiro,
T. Ngo-Duc, F. Otto, I. Pinto, A. Pirani, K. Raghavan, R. Ranasinghe, A.C. Ruane, L. Ruiz, J.-B. Sallée, B.H. Samset,
S. Sathyendranath, S.I. Seneviratne, A.A. Sörensson, S. Szopa, I. Takayabu, A.-M. Tréguier, B. van den Hurk, R. Vautard,
K. von Schuckmann, S. Zaehle, X. Zhang, and K. Zickfeld, 2021: Technical Summary. In Climate Change 2021: The
Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental
Panel on Climate Change [Masson-Delmotte, V., P. Zhai, A. Pirani, S.L. Connors, C. Péan, S. Berger, N. Caud, Y. Chen,
L. Goldfarb, M.I. Gomis, M. Huang, K. Leitzell, E. Lonnoy, J.B.R. Matthews, T.K. Maycock, T. Waterfield, O. Yelekçi, R. Yu,
and B. Zhou (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, pp. 33−144.
doi:10.1017/9781009157896.002.

36
 Technical Summary

Table of Contents

Introduction ��������������������������������������������������������������������������������������������������������� 38 TS.3 Understanding the Climate System Response

and Implications for Limiting Global Warming �� 90
Box TS.1 |
Core Concepts Central to This Report ����������������������������������� 39 TS.3.1 Radiative Forcing and Energy Budget ���������������������� 90
TS.3.2 Climate Sensitivity and
TS.1 A Changing Climate ������������������������������������������������������������������ 43 Earth System Feedbacks ���������������������������������������������������� 93
TS1.1 Context of a Changing Climate ������������������������������������ 43 TS.3.3 Temperature Stabilization,
Net Zero Emissions and Mitigation ���������������������������� 97
Box TS.2 |
Paleoclimate ������������������������������������������������������������������������������������������������ 45 Box TS.7 |
Climate and Air Quality Responses TS
TS.1.2 Progress in Climate Science �������������������������������������������� 47 to Short-lived Climate Forcers
TS.1.3 Assessing Future Climate Change ������������������������������ 52 in Shared Socio-economic Pathways ����������������������������������� 103
TS.1.4 From Global to Regional Climate Information Box TS.8 |
for Impact and Risk Assessment ���������������������������������� 57 Earth System Response
to Solar Radiation Modification ���������������������������������������������� 104
Cross-Section Box TS.1|
Box TS.9 |
Global Surface Temperature Change ����������������������������������������������� 59
Irreversibility, Tipping Points
and Abrupt Changes �������������������������������������������������������������������������� 106
TS.2 Large-scale Climate Change: Mean Climate,
Variability and Extremes ������������������������������������������������������� 63
TS.4 Regional Climate Change �������������������������������������������������� 107
TS.2.1 Changes Across the Global Climate System ��������� 63
TS.4.1 Generation and Communication of Regional
TS.2.2 Changes in the Drivers of the Climate System ���� 67 Climate Change Information ��������������������������������������� 107
TS.2.3 Upper Air Temperatures Box TS.10 |
and Atmospheric Circulation ������������������������������������������ 70 Event Attribution ���������������������������������������������������������������������������������� 108
Box TS.3 | Box TS.11 |
Low-likelihood, High-warming Storylines �������������������������� 72 Climate Services ������������������������������������������������������������������������������������ 111
TS.2.4 The Ocean ���������������������������������������������������������������������������������� 74
Box TS.12 |
TS.2.5 The Cryosphere ����������������������������������������������������������������������� 76 Multiple Lines of Evidence for Assessing
Box TS.4 | Regional Climate Change and the Interactive Atlas  111
Sea Level �������������������������������������������������������������������������������������������������������� 77 TS.4.2 Drivers of Regional Climate Variability
Box TS.5 | and Change ���������������������������������������������������������������������������� 113
The Carbon Cycle ������������������������������������������������������������������������������������� 79 Box TS.13 |
TS.2.6 Land Climate, Including Biosphere Monsoons ��������������������������������������������������������������������������������������������������� 118
and Extremes ��������������������������������������������������������������������������� 82 TS.4.3 Regional Climate Change
Box TS.6 | and Implications for Climate Extremes
Water Cycle �������������������������������������������������������������������������������������������������� 85 and Climatic Impact-Drivers ���������������������������������������� 120
Box TS.14 |
Infographic TS.1 | Urban Areas ����������������������������������������������������������������������������������������������� 144
Climate Futures ������������������������������������������������������������������������������������������������� 88

37
 Technical Summary

Introduction TS.3 summarizes knowledge and understanding of climate forcings,

feedbacks and responses. Infographic TS.1 uses a storyline approach
The Working Group I (WGI) contribution to the Intergovernmental to integrate findings on possible climate futures. Finally, Section TS.4
Panel on Climate Change (IPCC) Sixth Assessment Report (AR6) provides a synthesis of climate information at regional scales.3 The
assesses the physical science basis of climate change. As part of list of acronyms used in the WGI Report is in Annex VIII.
that contribution, this Technical Summary (TS) is designed to bridge
between the comprehensive assessment of the WGI Chapters Text at the beginning of a section presented in dark blue
and its Summary for Policymakers (SPM). It is primarily built from with a blue vertical bar at the left, as shown here, provides
the Executive Summaries of the individual chapters and Atlas and a summary of the findings discussed in that section.
provides a synthesis of key findings based on multiple lines of
evidence (e.g., analyses of observations, models, paleoclimate The AR6 WGI Report promotes best practices in traceability
information and understanding of physical, chemical and biological and reproducibility, including through adoption of the Findable,
processes and components of the climate system). All the findings Accessible, Interoperable, and Reusable (FAIR) principles for scientific
and figures here are supported by and traceable to the underlying data. Each chapter has a data table (in its Supplementary Material)
TS chapters, with relevant chapter sections indicated in curly brackets. documenting the input data and code used to generate its figures
and tables. In addition, a collection of data and code from the report
Throughout this Technical Summary, key assessment findings are has been made freely-available online via long-term archives.4
reported using the IPCC calibrated uncertainty language (Chapter 1,
Box 1.1). Two calibrated approaches are used to communicate the These FAIR principles are central to the WGI Interactive Atlas5, an
degree of certainty in key findings, which are based on author teams’ online tool that complements the WGI Report by providing flexible
evaluations of underlying scientific understanding: spatial and temporal analyses of past, observed and projected
climate change information. It comprises a regional information
1) Confidence1 is a qualitative measure of the validity of a finding, component that supports many of the chapters of the Report and
based on the type, amount, quality and consistency of evidence a regional synthesis component that supports the Technical Summary
(e.g., data, mechanistic understanding, theory, models, expert and Summary for Policymakers.
judgment) and the degree of agreement.
2) Likelihood2 provides a quantified measure of confidence in Regarding the representation of robustness and uncertainty in maps,
a finding expressed probabilistically (e.g., based on statistical the method chosen for the AR66 differs from the method used in the
analysis of observations or model results, or both, and expert Fifth Assessment Report (AR5). This choice is based on new research
judgement by the author team or from a formal quantitative on the visualization of uncertainty and on user surveys.
survey of expert views, or both).

Where there is sufficient scientific confidence, findings can also be

formulated as statements of fact without uncertainty qualifiers.
Throughout IPCC reports, the calibrated language is clearly identified
by being typeset in italics.

The context and progress in climate science (Section TS.1) is followed

by a Cross-Section Box TS.1 on global surface temperature change.
Section TS.2 provides information about past and future large-
scale changes in all components of the climate system. Section

1 In this Technical Summary, the following summary terms are used to describe the available evidence: limited, medium, or robust; and for the degree of agreement: low, medium, or high. A level of
confidence is expressed using five qualifiers: very low, low, medium, high, and very high, and typeset in italics, e.g., medium confidence. For a given evidence and agreement statement, different
confidence levels can be assigned, but increasing levels of evidence and degrees of agreement are correlated with increasing confidence (see Chapter 1, Box 1.1 for more details).
2 In this Technical Summary, the following terms are used to indicate the assessed likelihood of an outcome or a result: virtually certain 99–100% probability, very likely 90–100%, likely 66–100%,
about as likely as not 33–66%, unlikely 0–33%, very unlikely 0–10%, exceptionally unlikely 0–1%. Additional terms (extremely likely: 95–100%, more likely than not >50–100%, and extremely
unlikely 0–5%) may also be used when appropriate. Assessed likelihood is typeset in italics, e.g., very likely (see Chapter 1, Box 1.1 for more details). Throughout the WGI report and unless stated
otherwise, uncertainty is quantified using 90% uncertainty intervals. The 90% uncertainty interval, reported in square brackets [x to y], is estimated to have a 90% likelihood of covering the value
that is being estimated. The range encompasses the median value, and there is an estimated 10% combined likelihood of the value being below the lower end of the range (x) and above its upper
end (y). Often, the distribution will be considered symmetric about the corresponding best estimate, but this is not always the case. In this Report, an assessed 90% uncertainty interval is referred
to as a ‘very likely range’. Similarly, an assessed 66% uncertainty interval is referred to as a ‘likely range’.
3 The regional traceback matrices that provide the location of the assessment findings synthesized in Section TS.4 are in the Supplementary Material (SM) of Chapter 10.
4 Data archive is available at https://catalogue.ceda.ac.uk/uuid/3234e9111d4f4354af00c3aaecd879b7.
5 https://interactive-atlas.ipcc.ch/
6 The AR6 figures use one of the following approaches. For observations, the absence of ‘x’ symbols shows areas with statistical significance, while the presence of ‘x’ indicates non-significance.
For model projections, the method offers two approaches with varying complexity. In the simple approach, high agreement (≥80%) is indicated with no overlay, and diagonal lines (///) show low
agreement (<80%); In the advanced approach, areas with no overlay display robust signal (≥66% of models show change greater than the variability threshold and ≥80% of all models agree on
the sign of change), reverse diagonal lines (\\\) show no robust signal, and crossed lines show conflicting signals (i.e., significant change but low agreement). Cross-Chapter Box Atlas.1 provides
more information on the AR6 method for visualizing robustness and uncertainty on maps.

38
 Technical Summary

Box TS.1 | Core Concepts Central to This Report

This box provides short descriptions of key concepts that are relevant to the AR6 WGI assessment, with a focus on their use in the
Technical Summary and the Summary for Policymakers. The Glossary (Annex VII) includes more information on these concepts along
with definitions of many other important terms and concepts used in this Report.

Characteristics of Climate Change Assessment

Global warming: Global warming refers to the change of global surface temperature relative to a baseline depending upon the
application. Specific global warming levels, such as 1.5°C, 2°C, 3°C or 4°C, are defined as changes in global surface temperature
relative to the years 1850–1900 as the baseline (the earliest period of reliable observations with sufficient geographic coverage). They
are used to assess and communicate information about global and regional changes, linking to scenarios and used as a common basis
for Working Group II (WGII) and Working Group III (WGIII) assessments. (Section TS.1.3, Cross-Section Box TS.1) {1.4.1, 1.6.2, 4.6.1,
Cross-Chapter Boxes 1.5, 2.3, 11.1, and 12.1, Atlas Sections 3–11, Glossary} TS

Emergence: Emergence refers to the experience or appearance of novel conditions of a particular climate variable in a given region.
This concept is often expressed as the ratio of the change in a climate variable relative to the amplitude of natural variations of that
variable (often termed a ‘signal-to-noise’ ratio, with emergence occurring at a defined threshold of this ratio). Emergence can be
expressed in terms of a time or a global warming level at which the novel conditions appear and can be estimated using observations
or model simulations. (Sections TS.1.2.3 and TS.4.2) {1.4.2, FAQ 1.2, 7.5.5, 10.3, 10.4, 12.5.2, Cross-Chapter Box Atlas.1, Glossary}

Cumulative carbon dioxide (CO2) emissions: The total net amount of CO2 emitted into the atmosphere as a result of human
activities. Given the nearly linear relationship between cumulative CO2 emissions and increases in global surface temperature,
cumulative CO2 emissions are relevant for understanding how past and future CO2 emissions affect global surface temperature.
A related term – remaining carbon budget – is used to describe the total net amount of CO2 that could be released in the future by
human activities while keeping global warming to a specific global warming level, such as 1.5°C, taking into account the warming
contribution from non-CO2 forcers as well. The remaining carbon budget is expressed from a recent specified date, while the total
carbon budget is expressed starting from the pre-industrial period. (Sections TS.1.3 and TS.3.3) {1.6.3, 5.5, Glossary}

Net zero CO2 emissions: A condition that occurs when the amount of CO2 emitted into the atmosphere by human activities equals
the amount of CO2 removed from the atmosphere by human activities over a specified period of time. Net negative CO2 emissions
occur when anthropogenic removals exceed anthropogenic emissions. (Section TS.3.3) {Box 1.4, Glossary}

Human Influence on the Climate System

Earth’s energy imbalance: In a stable climate, the amount of energy that Earth receives from the Sun is approximately in balance
with the amount of energy that is lost to space in the form of reflected sunlight and thermal radiation. ‘Climate drivers’, such as an
increase in greenhouse gases or aerosols, interfere with this balance, causing the system to either gain or lose energy. The strength
of a climate driver is quantified by its effective radiative forcing (ERF), measured in W m–2. Positive ERF leads to warming, and
negative ERF leads to cooling. That warming or cooling in turn can change the energy imbalance through many positive (amplifying)
or negative (dampening) climate feedbacks. (Sections TS.2.2, TS.3.1 and TS.3.2) {2.2.8, 7.2, 7.3, 7.4, Box 7.1, Box 7.2, Glossary}

Attribution: Attribution is the process of evaluating the relative contributions of multiple causal factors to an observed change in
climate variables (e.g., global surface temperature, global mean sea level), or to the occurrence of extreme weather or climate-related
events. Attributed causal factors include human activities (such as increases in greenhouse gas concentration and aerosols, or land-use
change) or natural external drivers (solar and volcanic influences), and in some cases internal variability. (Sections TS.1.2.4 and TS.2,
Box TS.10) {Cross-Working Group Box: Attribution in Chapter 1; 3.5; 3.8; 10.4; 11.2.4; Glossary}

Committed change, long-term commitment: Changes in the climate system, resulting from past, present and future human
activities, which will continue long into the future (centuries to millennia) even with strong reductions in greenhouse gas emissions.
Some aspects of the climate system, including the terrestrial biosphere, the deep ocean and the cryosphere, respond much more slowly
than surface temperatures to changes in greenhouse gas concentrations. As a result, there are already substantial committed changes
associated with past greenhouse gas emissions. For example, global mean sea level will continue to rise for thousands of years, even
if future CO2 emissions are reduced to net zero and global warming halted, as excess energy due to past emissions continues to
propagate into the deep ocean and as glaciers and ice sheets continue to melt. (Section TS.2.1, Box TS.4, Box TS.9) {1.2.1, 1.3, Box 1.2,
Cross-Chapter Box 5.3}

39
 Technical Summary

Box TS.1 (continued)

Climate Information for Regional Climate Change and Risk Assessment

Distillation: The process of synthesizing information about climate change from multiple lines of evidence obtained from a variety
of sources, taking into account user context and values. It leads to an increase in the usability, usefulness and relevance of climate
information, enhances stakeholder trust, and expands the foundation of evidence used in climate services. It is particularly relevant in
the context of co-producing regional-scale climate information to support decision-making. (Section TS.4.1, Box TS.11) {10.1, 10.5, 12.6}

(Climate change) risk: The concept of risk is a key aspect of how the IPCC assesses and communicates to decision-makers about the
potential for adverse consequences for human or ecological systems, recognizing the diversity of values and objectives associated with
such systems. In the context of climate change, risks can arise from potential impacts of climate change as well as human responses to
climate change. WGI contributes to the common IPCC risk framing through the assessment of relevant climate information, including
TS climatic impact-drivers and low-likelihood, high-impact outcomes. (Sections TS.1.4 and TS.4.1, Box TS.4) {Cross-Chapter Boxes 1.3 and
12.1, Glossary}

Climatic impact-drivers: Physical climate system conditions (e.g., means, events, extremes) that can be directly connected with
having impacts on human or ecological systems are described as ‘climatic impact-drivers’ (CIDs) without anticipating whether their
impacts are detrimental (i.e., as for hazards in the context of climate change risks) or provide potential opportunities. A range of indices
may capture the sector- or application-relevant characteristics of a climatic impact-driver and can reflect exceedances of identified
tolerance thresholds. (Sections TS.1.4 and TS.4.3) {12.1–12.3, FAQ 12.1, Glossary}

Storylines: The term storyline is used both in connection to scenarios (related to a future trajectory of emissions or socio-economic
developments) or to describe plausible trajectories of weather and climate conditions or events, especially those related to high levels
of risk. Physical climate storylines are introduced in AR6 to explore uncertainties in climate change and natural climate variability, to
develop and communicate integrated and context-relevant regional climate information, and to address issues with deep uncertainty7,
including low-likelihood, high-impact outcomes. (Section TS.1.4, Box TS.3, Infographic TS.1) {1.4.4, Box 10.2, Glossary}

Low-likelihood, high impact outcomes: Outcomes/events whose probability of occurrence is low or not well known (as in the context
of deep uncertainty) but whose potential impacts on society and ecosystems could be high. To better inform risk assessment and
decision-making, such low-likelihood outcomes are considered if they are associated with very large consequences and may therefore
constitute material risks, even though those consequences do not necessarily represent the most likely outcome. (Section TS.1.4,
Box TS.3, Figure TS.6) {1.4.4, 4.8, Cross Chapter Box 1.3, Glossary}

As part of the AR6 cycle, the IPCC produced three Special Reports The Report has been peer-reviewed by the scientific community
in 2018 and 2019: the Special Report on Global Warming of and governments (Annex X provides the Expert Reviewer list). The
1.5°C (SR1.5), the Special Report on the Ocean and Cryosphere in substantive introduction provided by Chapter 1 is followed by a first
a Changing Climate (SROCC), and the Special Report on Climate set of chapters dedicated to large-scale climate knowledge (Chapters
Change and Land (SRCCL). 2–4), which encompasses observations and paleoclimate evidence,
causes of observed changes, and projections; these are complemented
The AR6 WGI Report provides a full and comprehensive assessment by Chapter 11 for large-scale changes in extremes. The second set of
of the physical science basis of climate change that builds on the chapters (Chapters 5–9) is orientated around the understanding of
previous assessments and these Special Reports and considers new key climate system components and processes, including the global
information and knowledge from the recent scientific literature8, cycles of carbon, energy and water; short-lived climate forcers and their
including longer observational datasets and new scenarios and link to air quality; and the ocean, cryosphere and sea level change.
model results. The last set of chapters (Chapters 10–12 and the Atlas) is dedicated to
the assessment and distillation of regional climate information from
The structure of the AR6 WGI Report is designed to enhance the multiple lines of evidence at sub-continental to local scales (including
visibility of knowledge developments and to facilitate the integration urban climate), with a focus on recent and projected regional changes
of multiple lines of evidence, thereby improving confidence in findings. in mean climate, extremes, and climatic impact-drivers. The new online

7 Although not a core concept of the WGI Report, deep uncertainty is used in the Technical Summary in the following sense: ‘A situation of deep uncertainty exists when experts or stakeholders do
not know or cannot agree on: (1) appropriate conceptual models that describe relationships among key driving forces in a system; (2) the probability distributions used to represent uncertainty
about key variables and parameters; and/or (3) how to weigh and value desirable alternative outcomes’ (Lempert et al., 2003). Lempert, R. J., Popper, S. W., and Bankes, S. C. (2003). Shaping the
next one hundred years: New methods for quantitative long-term strategy analysis (MR-1626-RPC). Santa Monica, CA: The RAND Pardee Center.
8 The assessment covers scientific literature accepted for publication by 31 January 2021.

40
 Technical Summary

Interactive Atlas allows users to interact in a flexible manner through • Updated assessment of recent warming: The AR5 reported
maps, time series and summary statistics with climate information for a smaller rate of increase in global mean surface temperature over
a set of updated WGI reference regions. The Report also includes 34 the period 1998–2012 than the rate calculated since 1951. Based
Frequently Asked Questions and answers for the general public (https:// on updated observational datasets showing a larger trend over
www.ipcc.ch/report/ar6/wg1/faqs). 1998–2012 than earlier estimates, there is now high confidence
that the observed 1998–2012 global surface temperature trend
Together, this Technical Summary and the underlying chapters aim at is consistent with ensembles of climate model simulations,
providing a comprehensive picture of knowledge progress since the and there is now very high confidence that the slower rate of
WGI contribution to AR5. Multiple lines of scientific evidence confirm global surface temperature increase observed over this period
that the climate is changing due to human influence. Important was a temporary event induced by internal and naturally forced
advances in the ability to understand past, present and possible variability that partly offset the anthropogenic surface warming
future changes should result in better-informed decision-making. trend over this period, while heat uptake continued to increase
in the ocean. Since 2012, strong warming has been observed,
Some of the new results and main updates to key findings in this with the past five years (2016–2020) being the hottest five-
Report compared to AR5, SR1.5, SRCCL, and SROCC are summarized year period in the instrumental record since at least 1850 (high TS
below. Relevant Technical Summary sections with further details are confidence). (Section TS.1.2, Cross-Section Box TS.1)
shown in parentheses after each bullet point. • Magnitude of climate system response: In this Report, it has
been possible to reduce the long-standing uncertainty ranges
for metrics that quantify the response of the climate system to
Selected Updates and/or New Results since AR5 radiative forcing, such as the equilibrium climate sensitivity (ECS)
and the transient climate response (TCR), due to substantial
• Human influence9 on the climate system is now an advances (e.g., a 50% reduction in the uncertainty range of cloud
established fact: The Fourth Assessment Report (AR4) stated feedbacks) and improved integration of multiple lines of evidence,
in 2007 that ‘warming of the climate system is unequivocal’, and including paleoclimate information. Improved quantification
AR5 stated in 2013 that ‘human influence on the climate system of ERF, the climate system radiative response, and the observed
is clear’. Combined evidence from across the climate system energy increase in the Earth system over the past five decades
strengthens this finding. It is unequivocal that the increase of demonstrate improved consistency between independent
CO2, methane (CH4) and nitrous oxide (N2O) in the atmosphere estimates of climate drivers, the combined climate feedbacks, and
over the industrial era is the result of human activities and that the observed energy increase relative to AR5. (Section TS.3.2)
human influence is the main driver10 of many changes observed • Improved constraints on projections of future climate
across the atmosphere, ocean, cryosphere and biosphere. change: For the first time in an IPCC report, the assessed future
(Sections TS.1.2, TS.2.1 and TS.3.1) change in global surface temperature is consistently constructed
• Observed global warming to date: A combination of by combining scenario-based projections (which AR5 focused
improved observational records and a series of very warm years on) with observational constraints based on past simulations
since AR5 have resulted in a substantial increase in the estimated of warming as well as the updated assessment of ECS and TCR.
level of global warming to date. The contribution of changes in In addition, initialized forecasts have been used for the period
observational understanding alone between AR5 and AR6 leads 2019–2028. The inclusion of these lines of evidence reduces
to an increase of about 0.1°C in the estimated warming since the assessed uncertainty for each scenario. (Section TS.1.3,
1850–1900. For the decade 2011–2020, the increase in global Cross-Section Box TS.1)
surface temperature since 1850–1900 is assessed to be 1.09 • Air quality: The AR5 assessed that projections of air quality
[0.95 to 1.20] °C.11 Estimates of crossing times of global warming are driven primarily by precursor emissions, including CH4. New
levels and estimates of remaining carbon budgets are updated scenarios explore a diversity of future options in air pollution
accordingly. (Section TS.1.2, Cross-Section Box TS.1) management. The AR6 reports rapid recent shifts in the
• Paleoclimate evidence: The AR5 assessed that many of the geographical distribution of some of these precursor emissions,
changes observed since the 1950s are unprecedented over confirms the AR5 finding, and shows higher warming effects
decades to millennia. Updated paleoclimate evidence strengthens of short-lived climate forcers in scenarios with the highest air
this assessment; over the past several decades, key indicators of pollution. (Sections TS.1.3 and TS.2.2, Box TS.7)
the climate system are increasingly at levels unseen in centuries • Effects of short-lived climate forcers on global warming:
to millennia and are changing at rates unprecedented in at least The AR5 assessed the radiative forcing for emitted compounds.
the last 2000 years. (Box TS.2, Section TS.2) The AR6 has extended this by assessing the emissions-based ERFs

9 Human influence on the climate system refers to human-driven activities that lead to changes in the climate system due to perturbations of Earth’s energy budget (also called anthropogenic
forcing). Human influence results from emissions of greenhouse gases, aerosols and tropospheric ozone precursors, ozone-depleting substances, and land-use change.
10 Throughout this Technical Summary, ‘main driver’ means responsible for more than 50% of the change.
11 Throughout the WGI report and unless stated otherwise, uncertainty is quantified using 90% uncertainty intervals. The 90% uncertainty interval, reported in square brackets [x to y], is estimated
to have a 90% likelihood of covering the value that is being estimated. The range encompasses the median value and there is an estimated 10% combined likelihood of the value being below the
lower end of the range (x) and above its upper end (y). Often the distribution will be considered symmetric about the corresponding best estimate, but this is not always the case. In this Report,
an assessed 90% uncertainty interval is referred to as a ‘very likely range’. Similarly, an assessed 66% uncertainty interval is referred to as a ‘likely range’.

41
 Technical Summary

also accounting for aerosol–cloud interactions. The best estimates • Effect of short-lived climate forcers on global warming
of ERF attributed to sulphur dioxide (SO2) and CH4 emissions are in coming decades: The SR1.5 stated that reductions in
substantially greater than in AR5, while that of black carbon is emissions of cooling aerosols partially offset greenhouse gas
substantially reduced. The magnitude of uncertainty in the ERF mitigation effects for two to three decades in pathways limiting
due to black carbon emissions has also been reduced relative to global warming to 1.5°C. The AR6 assessment updates the AR5
AR5. (Section TS.3.1) assessment of the net cooling effect of aerosols and confirms
• Global water cycle: The AR5 assessed that anthropogenic that changes in short-lived climate forcers will very likely cause
influences have likely affected the global water cycle since further warming in the next two decades across all scenarios.
1960. The dedicated chapter in AR6 (Chapter 8) concludes with (Section TS.1.3, Box TS.7)
high confidence that human-caused climate change has driven • COVID-19: Temporary emissions reductions in 2020 associated
detectable changes in the global water cycle since the mid-20th with COVID-19 containment led to small and positive net radiative
century, with a better understanding of the response to aerosol effect (warming influence). However, global and regional climate
and greenhouse gas changes. The AR6 further projects with high responses to this forcing are undetectable above internal climate
confidence an increase in the variability of the water cycle in most variability due to the temporary nature of emissions reductions.
TS regions of the world and under all emissions scenarios. (Box TS.6) (Section TS.3.3)
• Extreme events: The AR5 assessed that human influence
had been detected in changes in some climate extremes.
A dedicated chapter in AR6 (Chapter 11) concludes that it is Selected Updates and/or New Results Since AR5, SRCCL
now an established fact that human-induced greenhouse gas and SROCC
emissions have led to an increased frequency and/or intensity
of some weather and climate extremes since 1850, in particular • Atmospheric concentration of methane: The SRCCL reported
for temperature extremes. Evidence of observed changes and a resumption of atmospheric CH4 concentration growth since
attribution to human influence has strengthened for several types 2007. The AR6 reports a faster growth over 2014–2019 and
of extremes since AR5, in particular for extreme precipitation, assesses growth since 2007 to be largely driven by emissions
droughts, tropical cyclones and compound extremes (including from the fossil fuels and agriculture (dominated by livestock)
fire weather). (Sections TS.1.2 and TS.2.1, Box TS.10) sectors. (Section TS.2.2)
• Land and ocean carbon sinks: The SRCCL assessed that the
persistence of the land carbon sink is uncertain due to climate
Selected Updates and/or New Results Since AR5 and SR1.5 change. The AR6 finds that land and ocean carbon sinks are
projected to continue to grow until 2100 with increasing
• Timing of crossing 1.5°C global warming: Slightly different atmospheric concentrations of CO2, but the fraction of emissions
approaches are used in SR1.5 and in this Report. SR1.5 assessed taken up by land and ocean is expected to decline as the CO2
a likely range of 2030 to 2052 for reaching a global warming concentration increases, with a much larger uncertainty range for
level of 1.5°C (for a 30-year period), assuming a continued, the land sink. The AR5, SR1.5 and SRCCL assessed carbon dioxide
constant rate of warming. In AR6, combining the larger estimate removal options and scenarios. The AR6 finds that the carbon
of global warming to date and the assessed climate response to cycle response is asymmetric for pulse emissions or removals,
all considered scenarios, the central estimate of crossing 1.5°C of which means that CO2 emissions would be more effective at
global warming (for a 20-year period) occurs in the early 2030s, in raising atmospheric CO2 than CO2 removals are at lowering
the early part of the likely range assessed in SR1.5, assuming no atmospheric CO2. (Section TS.3.3, Box TS.5)
major volcanic eruption. (Section TS.1.3, Cross-Section Box TS.1) • Ocean stratification increase12 : Refined analyses of available
• Remaining carbon budgets: The AR5 had assessed the observations in the AR6 lead to a reassessment of the rate of
transient climate response to cumulative emissions of CO2 to be increase of the global stratification in the upper 200 m to be double
likely in the range of 0.8°C to 2.5°C per 1000 GtC (1 Gigatonne that estimated in SROCC from 1970 to 2018. (Section TS.2.4)
of carbon, GtC, = 1 Petagram of carbon, PgC, = 3.664 Gigatonnes • Projected ocean oxygen loss: Future subsurface oxygen
of carbon dioxide, GtCO2), and this was also used in SR1.5. The decline in new projections assessed in WGI AR6 is substantially
assessment in AR6, based on multiple lines of evidence, leads to greater in 2080–2099 than assessed in SROCC. (Section TS.2.4)
a narrower likely range of 1.0°C–2.3°C per 1000 GtC. This has • Ice loss from glaciers and ice sheets: Since SROCC, globally
been incorporated in updated estimates of remaining carbon resolved glacier changes have improved estimates of glacier mass
budgets (see Section TS.3.3.1), together with methodological loss over the past 20 years, and estimates of the Greenland and
improvements and recent observations. (Sections TS.1.3 Antarctic Ice Sheet loss have been extended to 2020. (Section TS.2.5)
and TS.3.3) • Observed global mean sea level change: new observation-
based estimates published since SROCC lead to an assessed sea
level rise estimate from 1901 to 2018 that is now consistent with
the sum of individual components and consistent with closure of
the global energy budget. (Box TS.4)

12 Increased stratification reduces the vertical exchange of heat, salinity, oxygen, carbon and nutrients. Stratification is an important indicator for ocean circulation.

42
 Technical Summary

• Projected global mean sea level change: The AR6 projections TS.1 A Changing Climate
of global mean sea level are based on projections from ocean
thermal expansion and land ice contribution estimates, which are This section introduces the assessment of the physical science basis of
consistent with the assessed ECS and assessed changes in global climate change in the AR6 and presents the climate context in which
surface temperature. They are underpinned by new land ice model this assessment takes place, recent progress in climate science and
intercomparisons and consideration of processes associated with the relevance of global and regional climate information for impact
low confidence to characterize the deep uncertainty in future ice and risk assessments. The future emissions scenarios and global
loss from Antarctica. The AR6 projections based on new models warming levels, used to integrate assessments across this Report,
and methods are broadly consistent with SROCC findings. are introduced and their applications for future climate projections
(Box TS.4) are briefly addressed. Paleoclimate science provides a long-term
context for observed climate change of the past 150 years and the
projected changes in the 21st century and beyond (Box TS.2). The
assessment of past, current and future global surface temperature
changes relative to the standard baselines and reference periods13
used throughout this Report is summarized in Cross-Section Box TS.1. TS

TS1.1 Context of a Changing Climate

This Report assesses new scientific evidence relevant

for a world whose climate system is rapidly changing,
overwhelmingly due to human influence. The five IPCC
assessment cycles since 1990 have comprehensively and
consistently laid out the rapidly accumulating evidence
of a changing climate system, with the Fourth Assessment
Report in 2007 being the first to conclude that warming
of the climate system is unequivocal. Sustained changes
have been documented in all major elements of the climate
system: the atmosphere, land, cryosphere, biosphere and
ocean (Section TS.2). Multiple lines of evidence indicate
the recent large-scale climatic changes are unprecedented
in a multi-millennial context and that they represent
a millennial-scale commitment for the slow-responding
elements of the climate system, resulting in continued
worldwide loss of ice, increase in ocean heat content, sea
level rise and deep ocean acidification (Box TS.2; Section
TS.2). {1.2.1, 1.3, Box 1.2, 2.2, 2.3, Figure 2.34, 5.1, 5.3, 9.2,
9.4–9.6, Appendix 1.A}

Earth’s climate system has evolved over many millions of years, and
evidence from natural archives provides a long-term perspective on
observed changes and projected changes over the coming centuries.
These reconstructions of past climate also show that atmospheric
CO2 concentrations and global surface temperature are strongly
coupled (Figure TS.1), based on evidence from a variety of proxy
records over multiple time scales (Box TS.2, Section TS.2). Levels of
global warming (see Core Concepts Box) that have not been seen
in millions of years could be reached by 2300, depending on the
emissions pathway that is followed (Section TS.1.3). For example,
there is medium confidence that, by 2300, an intermediate scenario14
used in this Report leads to global surface temperatures of [2.3°C
to 4.6°C] higher than 1850–1900, similar to the mid-Pliocene Warm

13 Several baselines or reference periods are used consistently throughout this Report. Baseline refers to a period against which anomalies (i.e., differences from the average value for the baseline
period) are calculated. Examples include the 1750 baseline (used for anthropogenic radiative forcings), the 1850–1900 baseline (an approximation for pre-industrial global surface temperature
from which global warming levels are calculated) and the 1995–2014 baseline (used for many climate model projections). A reference period indicates a time period over which various statistics
are calculated (e.g., the near-term reference period, 2021–2040). Paleo reference periods are listed in Box TS.2. {1.4.1, Cross-Chapter Boxes 1.2 and 2.1}
14 Please refer to Section TS.1.3.1 for an overview of the climate change scenarios used in this Report.

43
 Technical Summary

Atmospheric CO2 concentration (ppm) Atmospheric CO2 concentration and global surface temperature change
during the last 60 million years and projections for the next 300 years SSP5-8.5
2000 SSP5-8.5

1000
SSP2-4.5 2300

400
SSP1-2.6

200
2100
early Eocene mid-Pliocene 2020

TS SSP1-2.6
Global surface temperature change (°C)

15

SSP5-8.5 2300
10
(relative to 1850–1900)

5 SSP2-4.5

0 SSP1-2.6 2100

-5

0 2 4 6 8 10 12 14 16 18
60 50 40 30 20 10 9 7 5 3 1 800 600 400 200 0 1850 2000 2150 2300
Temperature (°C)
Millions of years Thousands of years Year CE relative to 1850–1900

Figure TS.1 | Changes in atmospheric CO2 and global surface temperature (relative to 1850–1900) from the deep past to the next 300 years. The intent of
this figure is to show that CO2 and temperature covary, both in the past and into the future, and that projected CO2 and temperatures are similar to those only from many millions
of years ago. CO2 concentrations from millions of years ago are reconstructed from multiple proxy records (grey dots are data from Section 2.2.3.1, Figure 2.3 shown with cubic-
spline fit). CO2 levels for the last 800,000 years through the mid-20th century are from air trapped in polar ice; recent values are from direct air measurements. Global surface
temperature prior to 1850 is estimated from marine oxygen isotopes, one of multiple sources of evidence used to assess paleo temperatures in this Report. Temperature of the
past 170 years is the AR6 assessed mean. CO2 levels and global surface temperature change for the future are shown for three Shared Socio-economic Pathway (SSP) scenarios
through 2300 CE, using Earth system model emulators calibrated to the assessed global surface temperatures. Their smooth trajectories do not account for inter-annual to
inter-decadal variability, including transient response to potential volcanic eruptions. Global maps for two paleo reference periods are based on Coupled Model Intercomparison
Project Phase 6 (CMIP6) and pre-CMIP6 multi-model means, with site-level proxy data for comparison (squares and circles are marine and terrestrial, respectively). The map
for 2020 is an estimate of the total observed warming since 1850–1900. Global maps at right show two SSP scenarios at 2100 (2081–2100) and at 2300 (2281–2300; map
from CMIP6 models; temperature assessed in 4.7.1). A brief account of the major climate forcings associated with past global temperature changes is in Cross-Chapter Box 2.1.
(Section TS.1.3, Figure TS.9, Cross-Section Box TS.1, Box TS.2) {1.2.1.2; Figures 1.14 and 1.5; 2.2.3; 2.3.1.1; 2.3.1.1.1; Figures 2.4 and 2.5; Cross-Chapter Box 2.1, Figure 1;
4.5.1; 4.7.1; Cross-Chapter Box 4.1; Cross-Chapter Box 7.1; Figure 7.13}

Period [2.5°C to 4°C], about 3.2 million years ago, whereas the high been studied systematically since the early 20th century. Other
CO2 emissions scenario SSP5-8.5 leads to temperatures of [6.6°C major anthropogenic drivers, such as atmospheric aerosols (fine
to 14.1°C] by 2300, which overlaps with the Early Eocene Climate solid particles or liquid droplets), land-use change and non-CO2
Optimum [10°C to 18°C], about 50 million years ago. {Cross-Chapter greenhouse gases, were identified by the 1970s. Since systematic
Boxes 2.1 and 2.4, 2.3.1, 4.3.1.1, 4.7.1.2, 7.4.4.1} scientific assessments began in the 1970s, the influence of human
activities on the warming of the climate system has evolved from
Understanding of the climate system’s fundamental elements is theory to established fact (see also Section TS.2). The evidence for
robust and well established. Scientists in the 19th century identified human influence on recent climate change strengthened from the
the major natural factors influencing the climate system. They IPCC First Assessment Report in 1990 to the IPCC Fifth Assessment
also hypothesized the potential for anthropogenic climate change Report in 2013/14, and is now even stronger in this assessment
due to CO2 emitted by combustion of fossil fuels (petroleum, (Sections TS.1.2.4 and TS.2). Changes across a greater number of
coal, natural gas). The principal natural drivers of climate change, climate system components, including changes in regional climate
including changes in incoming solar radiation, volcanic activity, and extremes can now be attributed to human influence (see
orbital cycles and changes in global biogeochemical cycles, have Sections TS.2 and TS.4). {1.3.1–1.3.5, 3.1, 11.2, 11.9}

44
 Technical Summary

Box TS.2 | Paleoclimate

Paleoclimate evidence is integrated within multiple lines of evidence across the WGI Report to more fully understand
the climate system. Paleo evidence extends instrument-based observations of climate variables and climate drivers
back in time, providing the long-term context needed to gauge the extent to which recent and potential future
changes are unusual (Section TS.2, Figure TS.1). Pre-industrial climate states complement evidence from climate model
projections by providing real-world examples of climate characteristics for past global warming levels, with empirical
evidence for how the slow-responding components of the climate system operate over centuries to millennia – the
time scale for committed climate change (Core Concepts Box, Box TS.4, Box TS.9). Information about the state of
the climate system during well-described paleoclimate reference periods helps narrow the uncertainty range in the
overall assessment of Earth’s sensitivity to climate forcing (Section TS.3.2.1). {Cross-Chapter Box 2.1, FAQ 1.3, FAQ 2.1}

Paleoclimate reference periods. Over the long evolution of Earth’s climate, several periods have received extensive research
attention as examples of distinct climate states and rapid climate transitions (Box TS.2, Figure 1). These paleoclimate reference periods TS
represent the present geological era (Cenozoic; past 65 million years) and are used across chapters to help structure the assessment
of climate changes prior to industrialization. Cross-Chapter Box 2.1 describes the reference periods, along with a brief account of their
climate forcings, and lists where each is discussed in other chapters. Cross-Chapter Box 2.4 summarizes information on one of the
reference periods, the mid-Pliocene Warm Period. The Interactive Atlas includes model output from the World Climate Research
Programme Coupled Model Intercomparison Project Phase 6 (CMIP6) for four of the paleoclimate reference periods.

Three selected global climate indicators covary across multiple paleoclimate reference periods
(a) (b)
Reference period CO2 Temperature Sea level
(*See Interactive Atlas for climate model output) Age (ppm) (°C) (m)
2000
Recent past 1995–2014 CE 360 397 0.66 to 1.00 0.15 to 0.25 Early Eocene
Atmospheric CO2 (ppm)
Approximate pre-industrial 1850–1900 CE 286 296 –0.15 to +0.11 –0.03 to 0.00
1000
Last Millennium 850–1850 CE 278 to 285 –0.14 ~ 0.24 –0.05 to 0.03
Mid-Holocene* 6.5–5.5 ka 260 to 268 0.2 to 1.0 –3.5 to +0.5
Last Deglacial Transition 500
18–11 ka 193 271 not assessed –120 –50 Recent past Mid-Pliocene
Last Glacial Maximum* 23–19 ka 188 to 194 –5 to –7 –134 to –125 1850–1900
Last Interglacial
Last Interglacial* 129–116 ka 266 to 282 0.5 to 1.5 5 to 10
200 Mid-Holocene
Mid-Pliocene Warm Period* 3.3–3.0 Ma 360 to 420 2.5 to 4.0 5 to 25
Last Glacial Maximum
Early Eocene 53–49 Ma 1150 to 2500 10 to 18 70 to 76
Paleocene-Eocene Thermal Maximum 55.9–55.7 Ma 900 2000 10 to 25 not assessed 100
–10 –5 0 5 10 15 20
X to Y: very likely range (caveats in Figure 2.34) lower 1850–1900 higher
X Y: start to end of period, with no stated uncertainty Global surface temperature (°C)
relative to 1850–1900
X ~ Y: lowest and highest values, with not stated uncertainty

Box TS.2, Figure 1 | Paleoclimate and recent reference periods, with selected key indicators. The intent of this figure is to list the paleoclimate reference
periods used in this Report, to summarize three key global climate indicators, and compare CO2 with global temperature over multiple periods. (a) Three large-scale
climate indicators (atmospheric CO2, global surface temperature relative to 1850–1900, and global mean sea level relative to 1900), based on assessments in
Chapter 2, with confidence levels ranging from low to very high. (b) Comparison between global surface temperature (relative to 1850–1900) and atmospheric CO2
concentration (shown on a log scale) for multiple reference periods (mid-points with 5–95% ranges). {2.2.3, 2.3.1.1, 2.3.3.3, Figure 2.34}

Paleoclimate models and reconstructions. Climate models that target paleoclimate reference periods have been featured by
the IPCC since the First Assessment Report. Under the framework of CMIP6-PMIP4 (Paleoclimate Modelling Intercomparison Project),
new protocols for model intercomparisons have been developed for multiple paleoclimate reference periods. These modelling efforts
have led to improved understanding of the climate response to different external forcings, including changes in Earth’s orbital and
plate movements, solar irradiance, volcanism, ice-sheet size and atmospheric greenhouse gases. Likewise, quantitative reconstructions
of climate variables from proxy records that are compared with paleoclimate simulations have improved as the number of study sites
and variety of proxy types have expanded, and as records have been compiled into new regional and global datasets. {1.3.2, 1.5.1,
Cross-Chapter Boxes 2.1 and 2.4}

Global surface temperature. Since AR5, updated climate forcings, improved models, new understanding of the strengths
and weaknesses of a growing array of proxy records, better chronologies and more robust proxy data products have led to better
agreement between models and reconstructions. For global surface temperature, the mid-point of the AR6-assessed range and the
median of the model-simulated temperatures differ by an average of 0.5°C across five reference periods; they overlap within their

45
 Technical Summary

Box TS.2 (continued)

90% ranges in four of five cases, which together span from about 6 [5 to 7]°C colder during the Last Glacial Maximum to about 14 [10
to 18] °C warmer during the Early Eocene, relative to 1850–1900 (Box TS.2, Figure 2a). Changes in temperature by latitude in response
to multiple forcings show that polar amplification (stronger warming at high latitudes than the global average) is a prominent feature
of the climate system across multiple climate states, and the ability of models to simulate this polar amplification in past warm
climates has improved since AR5 (high confidence). Over the past millennium, and especially since about 1300 CE, simulated global
surface temperature anomalies are well within the uncertainty of reconstructions (medium confidence), except for some short periods
immediately following large volcanic eruptions, for which different forcing datasets disagree (Box TS.2, Figure 2b). {2.3.1.1, 3.3.3.1,
3.8.2.1, 7.4.4.1.2}

Proxy-based and model-simulated estimates of global surface temperature agree across multiple reference periods

TS (a) (b)
20
Global surface temperature relative to 1.0 Observed Reconstructed Simulated
1850–1900 0.8
15
Early Eocene
Simulated temperature (°C)

0.6
Global surface temperature
relative to 1850–1900 (°C)
10
0.4

5 0.2
mid-Pliocene
-
Recent past 0
0 1850–1900 Last Interglacial
mid-Holocene –0.2
–0.4
–5 Last Glacial Maximum
–0.6
–10
–10 –5 0 5 10 15 20 1000 1200 1400 1600 1800 2000
Reconstructed temperature (°C) Year (CE)

Box TS.2, Figure 2 | Global surface temperature as estimated from proxy records (reconstructed) and climate models (simulated). The intent of this
figure is to show the agreement between observations and models of global temperatures during paleo reference periods. (a) For individual paleoclimate reference
periods. (b) For the last millennium, with instrumental temperature (AR6 assessed mean, 10-year smoothed). Model uncertainties in (a) and (b) are 5–95% ranges of
multi-model ensemble means; reconstructed uncertainties are 5–95% ranges (medium confidence) of (a) midpoints and (b) multi-method ensemble median. {2.3.1.1,
Figure 2.34, Figure 3.2c, Figure 3.44}

Equilibrium climate sensitivity. Paleoclimate data provide evidence to estimate equilibrium climate sensitivity (ECS15) (Section
TS.3.2.1). In AR6, refinements in paleo data for paleoclimate reference periods indicate that ECS is very likely greater than 1.5°C and
likely less than 4.5°C, which is largely consistent with other lines of evidence and helps narrow the uncertainty range of the overall
assessment of ECS. Some of the CMIP6 climate models that have either high (>5°C) or low (<2°C) ECS also simulate past global
surface temperature changes outside the range of proxy-based reconstructions for the coldest and warmest reference periods. Since
AR5, independent lines of evidence, including proxy records from past warm periods and glacial–interglacial cycles, indicate that
sensitivity to forcing increases as temperature increases (Section TS.3.2.2). {7.4.3.2, 7.5.3, 7.5.6, Table 7.11}

Water cycle. New hydroclimate reconstructions and model-data comparisons have improved the understanding of the causes
and effects of long-term changes in atmospheric and ocean circulation, including monsoon variability and modes of variability
(Box TS.13, Section TS.4.2). Climate models are able to reproduce decadal drought variability on large regional scales, including the
severity, persistence and spatial extent of past megadroughts known from proxy records (medium confidence). Some long-standing
discrepancies remain, however, such as the magnitude of African monsoon precipitation during the early Holocene (the past 11,700
years), suggesting continuing knowledge gaps. Paleoclimate evidence shows that, in relatively high CO2 climates such as the Pliocene,
Walker circulation over the equatorial Pacific Ocean weakens, supporting the high confidence model projections of weakened Walker
cells by the end of the 21st century. {3.3.2, 8.3.1.6, 8.4.1.6, 8.5.2.1, 9.2}

15 In this Report, equilibrium climate sensitivity is defined as the equilibrium (steady state) change in the surface temperature following a doubling of the atmospheric carbon dioxide (CO2)
concentration from pre-industrial conditions.

46
 Technical Summary

Box TS.2 (continued)

Sea level and ice sheets. Although past and future global warming differ in their forcings, evidence from paleoclimate
records and modelling show that ice-sheet mass and global mean sea level (GMSL) responded dynamically over multiple millennia
(high confidence). This evidence helps to constrain estimates of the committed GMSL response to global warming (Box TS.4). For
example, under a past global warming levels of around [2.5°C to 4°C] relative to 1850–1900, like during the mid-Pliocene Warm
Period, sea level was [5 to 25 m] higher than 1900 (medium confidence); under past global warming levels of [10°C to 18°C], like
during the Early Eocene, the planet was essentially ice free (high confidence). Constraints from these past warm periods, combined
with physical understanding, glaciology and modelling, indicate a committed long-term GMSL rise over 10,000 years, reaching about
8 to 13 m for sustained peak global warming of 2°C and up to 28 to 37 m for 5°C, which exceeds the AR5 estimate. {2.3.3.3, 9.4.1.4,
9.4.2.6, 9.6.2, 9.6.3.5}

Ocean. Since AR5, better integration of paleo-oceanographic data with modelling along with higher-resolution analyses of
transient changes have improved understanding of long-term ocean processes. Low-latitude sea surface temperatures at the Last TS
Glacial Maximum cooled more than previously inferred, resolving some inconsistencies noted in AR5. This paleo context supports
the assessment that ongoing increase in ocean heat content (OHC) represents a long-term commitment (see Core Concepts Box),
essentially irreversible on human time scales (high confidence). Estimates of past global OHC variations generally track those of
sea surface temperatures around Antarctica, underscoring the importance of Southern Ocean processes in regulating deep-ocean
temperatures. Paleoclimate data, along with other evidence of glacial–interglacial changes, show that Antarctic Circumpolar flow
strengthened and that ventilation of Antarctic Bottom Water accelerated during warming intervals, facilitating release of CO2 stored in
the deep ocean to the atmosphere. Paleo evidence suggests significant reduction of deep-ocean ventilation associated with meltwater
input during times of peak warmth. {2.3.1.1, 2.3.3.1, 9.2.2, 9.2.3.2}

Carbon cycle. Past climate states were associated with substantial differences in the inventories of the various carbon reservoirs,
including the atmosphere (Section TS.2.2). Since AR5, the quantification of carbon stocks has improved due to the development of
novel sedimentary proxies and stable-isotope analyses of air trapped in polar ice. Terrestrial carbon storage decreased markedly
during the Last Glacial Maximum by 300–600 PgC, possibly by 850 PgC when accounting for interactions with the lithosphere and
ocean sediments, a larger reduction than previously estimated, owing to a colder and drier climate. At the same time, the storage of
remineralized carbon in the ocean interior increased by as much as 750–950 PgC, sufficient to balance the removal of carbon from the
atmosphere (200 PgC) and terrestrial biosphere reservoirs combined (high confidence). {5.1.2.2}

TS.1.2 Progress in Climate Science satellite measurement techniques are now long enough to be
relevant for climate assessments. For example, globally distributed,
TS.1.2.1 Observation-based Products and their Assessments high-vertical-resolution profiles of temperature and humidity in the
upper troposphere and stratosphere can be obtained from the early
Observational capabilities have continued to improve and 2000s using global navigation satellite systems, leading to updated
expand overall since AR5, enabling improved consistency estimates of recent atmospheric warming. Improved measurements
between independent estimates of climate drivers, the of ocean heat content, warming of the land surface, ice-sheet mass
combined climate feedbacks, and the observed energy loss and sea level changes allow a better closure of the global
and sea level increase. Satellite climate records and energy and sea level budgets relative to AR5. For surface and
improved reanalyses are used as an additional line of balloon-based networks, apparent regional data reductions result
evidence for assessing changes at the global and regional from a combination of data policy issues, data curation/provision
scales. However, there have also been reductions in some challenges, and real cessation of observations, and are to an extent
observational data coverage or continuity and limited counter-balanced by improvements elsewhere. Limited observational
access to data resulting from data policy issues. Natural records of extreme events and spatial data gaps currently limit the
archives of past climate, such as tropical glaciers, have also assessment of some observed regional climate change. {1.5.1, 2.3.2,
been subject to losses (in part due to anthropogenic climate 7.2.2, Box 7.2, Cross-Chapter Box 9.1, 9.6.1, 10.2.2, 10.6, 11.2, 12.4}
change). {1.5.1, 1.5.2, 10.2.2}
New paleoclimate reconstructions from natural archives have enabled
Earth system observations are an essential driver of progress in our more robust reconstructions of the spatial and temporal patterns of
understanding of climate change. Overall, capabilities to observe past climate changes over multiple time scales (Box TS.2). However,
the physical climate system have continued to improve and expand. paleoclimate archives, such as tropical glaciers and modern natural
Improvements are particularly evident in ocean observing networks archives used for calibration (e.g., corals and trees), are rapidly
and remote-sensing systems. Records from several recently instigated disappearing owing to a host of pressures, including increasing

47
 Technical Summary

temperatures (high confidence). Substantial quantities of past Section Box TS.1). Increasing horizontal resolution in global
instrumental observations of weather and other climate variables, climate models improves the representation of small-scale
over both land and ocean, which could fill gaps in existing datasets, features and the statistics of daily precipitation (high
remain un-digitized or inaccessible. These include measurements of confidence). Earth system models, which include additional
temperature (air and sea surface), rainfall, surface pressure, wind biogeochemical feedbacks, often perform as well as their
strength and direction, sunshine amount and many other variables lower-complexity global climate model counterparts, which
dating back into the 19th century. {1.5.1} do not account for these additional feedbacks (medium
confidence). {1.3.6, 1.5.3, 3.1, 3.5.1, 3.8.2, 4.3.1, 4.3.4, 7.5,
Reanalyses combine observations and models (e.g., a numerical 8.5.1, 9.6.3.1}
weather prediction model) using data assimilation techniques to
provide a spatially complete, dynamically consistent estimate of Climate model simulations coordinated and collected as part of the
multiple variables describing the evolving climate state. Since AR5, World Climate Research Programme’s Coupled Model Intercomparison
new reanalyses have been developed for the atmosphere and the Project Phase 6 (CMIP6), complemented by a range of results from
ocean with various combinations of increased resolution, extended the previous phase (CMIP5), constitute a key line of evidence
TS records, more consistent data assimilation and larger availability supporting this Report. The latest generation of CMIP6 models have
of uncertainty estimates. Limitations remain, for example, in how an improved representation of physical processes relative to previous
reanalyses represent global-scale changes to the water cycle. Regional generations, and a wider range of Earth system models now represent
reanalyses use high-resolution, limited-area models constrained by biogeochemical cycles. Higher-resolution models that better capture
regional observations and with boundary conditions from global smaller-scale processes are also increasingly becoming available
reanalyses. There is high confidence that regional reanalyses better for climate change research (Figure TS.2, Panels a and b). Results
represent the frequencies of extremes and variability in precipitation, from coordinated regional climate modelling initiatives, such as the
surface air temperature and surface wind than global reanalyses Coordinated Regional Climate Downscaling Experiment (CORDEX)
and provide estimates that are more consistent with independent complement and add value to the CMIP global models, particularly in
observations than dynamical downscaling approaches. {1.5.2, complex topography zones, coastal areas and small islands, as well
10.2.1.2, Annex I} as for extremes. {1.5.3, 1.5.4, 2.8.2, FAQ 3.3, 6.2.2, 6.4, 6.4.5, 8.5.1,
10.3.3, Atlas.1.4}
TS.1.2.2 Climate Model Performance
Projections of the increase in global surface temperature and the
This report assesses results from climate models pattern of warming from previous IPCC Assessment Reports and other
participating in the Coupled Model Intercomparison Project studies are broadly consistent with subsequent observations (limited
Phase 6 (CMIP6) of the World Climate Research Programme. evidence, high agreement), especially when accounting for the
These models include new and better representation of difference in radiative forcing scenarios used for making projections
physical, chemical and biological processes, as well as and the radiative forcings that actually occurred (Figure TS.3). The
higher resolution, compared to climate models considered AR5 and SROCC projections of GMSL for the 2007–2018 period have
in previous IPCC Assessment Reports. This has improved been shown to be consistent with observed trends in GMSL and
the simulation of the recent mean state of most large-scale regional weighted mean tide gauges. {1.3.6, 9.6.3.1}
indicators of climate change and many other aspects across
the climate system. Some differences from observations For most large-scale indicators of climate change, the simulated
remain, for example in regional precipitation patterns. recent mean climate from CMIP6 models underpinning this
Projections of the increase in global surface temperature, assessment have improved compared to the CMIP5 models used in
the pattern of warming, and global mean sea level rise AR5 (high confidence). This is evident from the performance of 18
from previous IPCC Assessment Reports and other studies simulated atmospheric and land large-scale indicators of climate
are broadly consistent with subsequent observations, change between the three generations of models (CMIP3, CMIP5,
especially when accounting for the difference in radiative and CMIP6) when benchmarked against reanalysis and observational
forcing scenarios used for making projections and the data (Figure TS.2, Panel c). Earth system models, characterized by
radiative forcings that actually occurred. additional biogeochemical feedbacks, often perform at least as well
as related, more constrained, lower-complexity models lacking these
The CMIP6 historical simulations assessed in this report feedbacks (medium confidence). {3.8.2, 10.3.3.3}
have an ensemble mean global surface temperature
change within 0.2°C of the observations over most of the The CMIP6 multi-model mean global surface temperature change
historical period, and observed warming is within the very from 1850–1900 to 2010–2019 is close to the best estimate of
likely range of the CMIP6 ensemble. However, some CMIP6 the observed warming. However, some CMIP6 models simulate
models simulate a warming that is either above or below a warming that is below or above the assessed very likely range. The
the assessed very likely range of observed warming. The CMIP6 models also reproduce surface temperature variations over
information about how well models simulate past warming, the past millennium, including the cooling that follows periods of
as well as other insights from observations and theory, are intense volcanism (medium confidence). For upper air temperature,
used to assess projections of global warming (see Cross- an overestimation of the upper tropical troposphere warming by

48
 Technical Summary

about 0.1°C per decade between 1979 and 2014 persists in most evolution of the satellite-observed Arctic sea ice loss (high confidence).
CMIP5 and CMIP6 models (medium confidence), whereas the The ability to model ice-sheet processes has improved substantially
differences between simulated and improved satellite-derived since AR5. As a consequence, there is medium confidence in the
estimates of change in global mean temperature through the depth representation of key processes related to surface-mass balance and
of the stratosphere have decreased. {3.3.1} retreat of the grounding-line (the junction between a grounded ice
sheet and an ice shelf, where the ice starts to float) in the absence
Some CMIP6 models demonstrate an improvement in how clouds of instabilities. However, there remains low confidence in simulations
are represented. CMIP5 models commonly displayed a negative of ice-sheet instabilities, ice-shelf disintegration and basal melting
shortwave cloud radiative effect that was too weak in the present owing to their high sensitivity to both uncertain oceanic forcing and
climate. These errors have been reduced, especially over the uncertain boundary conditions and parameters. {1.5.3, 2.3.2, 3.4.1,
Southern Ocean, due to a more realistic simulation of supercooled 3.4.2, 3.8.2, 9.3.1, 9.3.2, 9.4.1, 9.4.2}
liquid droplets with sufficient numbers and an associated increase
in the cloud optical depth. Because a negative cloud optical depth CMIP6 models are able to reproduce most aspects of the spatial
feedback in response to surface warming results from ‘brightening’ structure and variance of the El Niño–Southern Oscillation (ENSO)
of clouds via active phase change from ice to liquid cloud particles and Indian Ocean Basin and Dipole modes of variability (medium TS
(increasing their shortwave cloud radiative effect), the extratropical confidence). However, despite a slight improvement in CMIP6, some
cloud shortwave feedback in CMIP6 models tends to be less negative, underlying processes are still poorly represented. Models reproduce
leading to a better agreement with observational estimates (medium observed spatial features and variance of the Southern Annular
confidence). CMIP6 models generally represent more processes Mode (SAM) and Northern Annular Mode (NAM) very well (high
that drive aerosol–cloud interactions than the previous generation confidence). The summertime SAM trend is well captured, with
of climate models, but there is only medium confidence that those CMIP6 models outperforming CMIP5 models (medium confidence).
enhancements improve their fitness-for-purpose of simulating By contrast, the cause of the NAM trend towards its positive phase
radiative forcing of aerosol–cloud interactions. {6.4, 7.4.2, FAQ 7.2} is not well understood. In the Tropical Atlantic basin, which contains
the Atlantic Zonal and Meridional modes, major biases in modelled
CMIP6 models still have deficiencies in simulating precipitation mean state and variability remain. Model performance is limited in
patterns, particularly in the tropical ocean. Increasing horizontal reproducing sea surface temperature anomalies for decadal modes
resolution in global climate models improves the representation of of variability, despite improvements from CMIP5 to CMIP6 (medium
small-scale features and the statistics of daily precipitation (high confidence) (see also Section TS.1.4.2.2, Table TS.4). {3.7.3–3.7.7}
confidence). There is high confidence that high-resolution global,
regional and hydrological models provide a better representation of Earth system models (ESMs) simulate globally averaged land
land surfaces, including topography, vegetation and land-use change, carbon sinks within the range of observation-based estimates
which can improve the accuracy of simulations of regional changes in (high confidence), but global-scale agreement masks large regional
the terrestrial water cycle. {3.3.2, 8.5.1, 10.3.3, 11.2.3} disagreements. There is also high confidence that the ESMs simulate
the weakening of the global net flux of CO2 into the ocean during
There is high confidence that climate models can reproduce the recent the 1990s, as well as the strengthening of the flux from 2000. {3.6}
observed mean state and overall warming of temperature extremes
globally and in most regions, although the magnitude of the trends Two important quantities used to estimate how the climate system
may differ. There is high confidence in the ability of models to capture responds to changes in greenhouse gas (GHG) concentrations are the
the large-scale spatial distribution of precipitation extremes over land. equilibrium climate sensitivity (ECS) and transient climate response
The overall performance of CMIP6 models in simulating the intensity (TCR16). The CMIP6 ensemble has broader ranges of ECS and TCR
and frequency of extreme precipitation is similar to that of CMIP5 values than CMIP5 (see Section TS.3.2 for the assessed range). These
models (high confidence). {Cross-Chapter Box 3.2, 11.3.3, 11.4.3} higher sensitivity values can, in some models, be traced to changes
in extratropical cloud feedbacks (medium confidence). To combine
The structure and magnitude of multi-model mean ocean temperature evidence from CMIP6 models and independent assessments of ECS
biases have not changed substantially between CMIP5 and CMIP6 and TCR, various emulators are used throughout the report. Emulators
(medium confidence). Since AR5, there is improved consistency are a broad class of simple climate models or statistical methods
between recent observed estimates and model simulations of changes that reproduce the behaviour of complex ESMs to represent key
in upper (<700 m) ocean heat content. The mean zonal and overturning characteristics of the climate system, such as global surface temperature
circulations of the Southern Ocean and the mean overturning and sea level projections. The main application of emulators in AR6
circulation of the North Atlantic (AMOC) are broadly reproduced by is to extrapolate insights from ESMs and observational constraints to
CMIP5 and CMIP6 models. {3.5.1, 3.5.4, 9.2.3, 9.3.2, 9.4.2} produce projections from a larger set of emissions scenarios, which
is achieved due to their computational efficiency. These emulated
CMIP6 models better simulate the sensitivity of Arctic sea ice area projections are also used for scenario classification in WGIII. {Box 4.1,
to anthropogenic CO2 emissions, and thus better capture the time 4.3.4, 7.4.2, 7.5.6, Cross-Chapter Box 7.1, FAQ 7.2}

16 In this Report, transient climate response is defined as the surface temperature response for the hypothetical scenario in which atmospheric carbon dioxide (CO2) increases at 1% yr –1 from
pre-industrial to the time of a doubling of atmospheric CO2 concentration.

49
 Technical Summary

(a) Model resolution (b) Model complexity

0 110 60
HighRes 100 CMIP6
50 CMIP6 50 CMIP5
90
CMIP5 CMIP3
100 80
CMIP3 40

Number of models
150 70
60 30

Level
200
km

50
250 40 20

300 30
20 10
350
10
0
400 0
Atm Res Oce Res Atm lev Ocn lev # Mod Aerosols Atm Chem Land carbon N cycle vegetation Ocean BGC
cycle prognostic
(c) Pattern correlation with observational reference
TS 1.0

0.8
Correlation

0.4

CMIP6
0.2 CMIP5
CMIP3
Additional observations
0
Near-Surface Precipitation TOA TOA TOA SW TOA LW Sea Level Temperature Temperature Eastward Eastward Northward Northward Geopotential Specific Soil Leaf Area Gross Primary
Air Temperature Outgoing Outgoing Cloud Rad Cloud Rad Pressure 850 hPa 200 hPa Wind Wind Wind Wind Height Humidity Moisture Index Productivity
Shortwave Longwave Effect Effect 850 hPa 200 hPa 850 hPa 200 hPa 500 hPa 400 hPa
Radiation Radiation

Figure TS.2 | Progress in climate models. The intent of this figure is to show present improvements in climate models in resolution, complexity and representation of key
variables. (a) Evolution of model horizontal resolution and vertical levels (based on Figure 1.19). (b) Evolution of inclusion of processes and resolution from Coupled Model
Intercomparison Project Phase 3 (CMIP3), Phase 5 (CMIP5) and Phase 6 (CMIP6; Annex II). (c) Centred pattern correlations between models and observations for the annual
mean climatology over the period 1980–1999. Results are shown for individual CMIP3 (cyan), CMIP5 (blue) and CMIP6 (red) models (one ensemble member is used) as short
lines, along with the corresponding ensemble averages (long lines). The correlations are shown between the models and the primary reference observational data set (from left
to right: ERA5, GPCP-SG, CERES-EBAF, CERES-EBAF, CERES-EBAF, CERES-EBAF, JR-55, ERA5, ERA5, ERA5, ERA5, ERA5, ERA5, AIRS, ERA5, ESACCI-Soilmoisture, LAI3g, MTE). In
addition, the correlation between the primary reference and additional observational data sets (from left to right: NCEP, GHCN, -, -, -, -, ERA5, HadISST, NCEP, NCEP, NCEP, NCEP,
NCEP, NCEP, ERA5, NCEP, -, -, FLUXCOM) are shown (solid grey circles) if available. To ensure a fair comparison across a range of model resolutions, the pattern correlations
are computed after regridding all datasets to a resolution of 4º in longitude and 5º in latitude. (Expanded from Figure 3.43; produced with ESMValTool version 2). {Figure 3.43}

TS.1.2.3 Understanding Climate Variability and Observational datasets have been extended and improved since
Emerging Changes AR5, providing stronger evidence that the climate is changing and
allowing better estimates of natural climate variability on decadal
Observed changes in climate are unequivocal at the global time scales. There is very high confidence that the slower rate of global
scale and are increasingly apparent on regional and local surface temperature change observed over 1998–2012 compared to
spatial scales. Both the rate of long-term change and 1951–2012 was temporary, and was, with high confidence, induced
the amplitude of year-to-year variations differ between by internal variability (particularly Pacific Decadal Variability) and
regions and across climate variables, thus influencing when variations in solar irradiance and volcanic forcing that partly offset the
changes emerge or become apparent compared to natural anthropogenic warming over this period. Global ocean heat content
variations (see Emergence in Core Concepts Box). The signal continued to increase throughout this period, indicating continuous
of temperature change has emerged more clearly in tropical warming of the entire climate system (very high confidence). Hot
regions, where year-to-year variations tend to be small over extremes also continued to increase during this period over land
land, than in regions with greater warming but larger year- (high confidence). Even in a continually warming climate, periods
to-year variations (high confidence) (Figure TS.3). Long- of reduced and increased trends in global surface temperature at
term changes in other variables have emerged in many decadal time scales will continue to occur in the 21st century (very
regions, such as for some weather and climate extremes high confidence). {Cross-Chapter Box 3.1, 3.3.1, 3.5.1, 4.6.2, 11.3.2}
and Arctic sea ice area. {1.4.2, Cross-Chapter Box 3.1, 9.3.1,
11.3.2, 12.5.2} Since AR5, the increased use of ‘large ensembles’, or multiple
simulations with the same climate model but using different initial
conditions, supports improved understanding of the relative roles

50
 Technical Summary

of internal variability and forced change in the climate system. occurred, and there is medium confidence that many of these changes
Simulations and understanding of modes of climate variability, are attributable to human activities. Several impact-relevant changes
including teleconnections, have improved since AR5 (medium have not yet emerged from natural variability but will emerge sooner
confidence), and larger ensembles allow a better quantification of or later in this century depending on the emissions scenario (high
uncertainty in projections due to internal climate variability. {1.4.2, confidence). Ocean acidification and deoxygenation have already
1.5.3, 1.5.4, 4.2, 4.4.1, Box 4.1, 8.5.2, 10.3.4, 10.4} emerged over most of the global open ocean, as has a reduction in
Arctic sea ice (high confidence). {9.3.1, 9.6.4, 11.2, 11.3, 12.4, 12.5,
Changes in regional climate can be detected even though natural Atlas.3–Atlas.11}
climate variations can temporarily increase or obscure anthropogenic
climate change on decadal time scales. While anthropogenic forcing TS.1.2.4 Understanding of Human Influence
has contributed to multi-decadal mean precipitation changes in
several regions, internal variability can delay emergence of the The evidence for human influence on recent climate change
anthropogenic signal in long-term precipitation changes in many has strengthened progressively from the IPCC Second
land regions (high confidence). {10.4} Assessment Report to AR5 and is even stronger in this
assessment, including for regional scales and for extremes. TS
Mean temperatures and heat extremes have emerged above natural Human influence in the IPCC context refers to the human
variability in almost all land regions with high confidence. Changes activities that lead to or contribute to a climate response,
in temperature-related variables, such as regional temperatures, such as the human-induced emissions of greenhouse
growing season length, extreme heat and frost, have already gases that subsequently alter the atmosphere’s radiative

Emergence of changes in surface temperature

Annual mean temperature change and the change relative to year-to-year variations
(a) Change in temperature at a global warming level of 1°C
Zonal mean
90°N
Observation (Berkeley Earth)
60°N Other observation datasets
CMIP6 multi-model mean (45)
30°N 5-95% model range

30°S

60°S

Missing
90°S
data 0.5 1 2 3 4 5

–2.4 –2.1 –1.8 –1.5 –2.1 –0.9 –0.6 –0.3 0 0.3 0.6 0.9 1.2 1.5 1.8 2.1 2.4
°C
(b) Change in temperature at a global warming level of 1°C relative to the size of year-to-year variations
Zonal mean
90°N

60°N

30°N

30°S

60°S

Missing 90°S
data 0.5 1 2 3 4 5

–4 –3.5 –3 –2.5 –2 –1.5 –1 –0.5 0 0.5 1 1.5 2 2.5 3 3.5 4

Signal to noise ratio

Figure TS.3 | Emergence of changes in temperature over the historical period. The intent of this figure is to show how observed changes in temperature have
emerged and that the emergence pattern agrees with model simulations. The observed change in temperature at a global warming level of 1°C (a), and the signal-to-noise ratio
(the change in temperature at a global warming level of 1°C, divided by the size of year-to-year variations, (b)) using data from Berkeley Earth. The right panels show the zonal
means of the maps and include data from different observational datasets (red) and the Coupled Model Intercomparison Project Phase 6 (CMIP6) simulations (black, including
the 5–95% range) processed in the same way as the observations. {1.4.2, 10.4.3}

51
 Technical Summary

properties, resulting in warming of the atmosphere, ocean of scientific literature combining different lines of evidence, and
and land components of the climate system. Other human improved accessibility to different types of climate models (high
activities influencing climate include the emission of confidence) (see Sections TS.2 and TS.4). {Cross-Working Group Box:
aerosols and other short-lived climate forcers, and land-use Attribution in Chapter 1, 1.5, 3.2, 3.5, 5.2, 6.4.3, 8.3, 9.6, 10.1, 10.2,
change such as urbanization. Progress in our understanding 10.3.3, 10.4.1, 10.4.2, 10.4.3, 10.5, 10.6, Cross-Chapter Box 10.3,
of human influence is gained from longer observational Box 10.3, 11.1.6, 11.2–11.9, 12.4}
datasets, improved paleoclimate information, a stronger
warming signal since AR5, and improvements in climate
models, physical understanding and attribution techniques TS.1.3 Assessing Future Climate Change
(see Core Concepts Box). Since AR5, the attribution to
human influence has become possible across a wider range Various frameworks can be used to assess future climatic changes and
of climate variables and climatic impact-drivers (CIDs, to synthesize knowledge across climate change assessment in WGI,
see Core Concepts Box). New techniques and analyses WGII and WGIII. These frameworks include: (i) scenarios, (ii) global
drawing on several lines of evidence have provided greater warming levels and (iii) cumulative CO2 emissions (see Core Concepts
TS confidence in attributing changes in regional weather and Box). The latter two offer scenario- and path-independent approaches
climate extremes to human influence (high confidence). to assess future projections. Additional choices, for instance with regard
{1.3, 1.5.1, Appendix 1.A, 3.1–3.8, 5.2, 6.4.2, 7.3.5, 7.4.4, to common reference periods and time windows for which changes
8.3.1, 10.4, Cross-Chapter Box 10.3, 11.2–11.9, 12.4} are assessed, can further help to facilitate integration across the WGI
report and across the whole AR6 (see Section TS.1.1). {1.4.1, 1.6, Cross-
Combining the evidence from across the climate system increases Chapter Box 1.4, 4.2.2, 4.2.4, Cross-Chapter Box 11.1}
the level of confidence in the attribution of observed climate change
to human influence and reduces the uncertainties associated with TS.1.3.1 Climate Change Scenarios
assessments based on single variables. {Cross-Chapter Box 10.3}
A core set of five illustrative scenarios based on the Shared
Since AR5, the accumulation of energy in the Earth system has Socio-economic Pathways (SSPs) are used consistently
become established as a robust measure of the rate of global across this Report: SSP1-1.9, SSP1-2.6, SSP2-4.5, SSP3-7.0,
climate change on interannual-to-decadal time scales. The rate of and SSP5-8.5. These scenarios cover a broader range of
accumulation of energy is equivalent to Earth’s energy imbalance greenhouse gas and air pollutant futures than assessed in
and can be quantified by changes in the global energy inventory for earlier WGI reports, and they include high-CO2 emissions
all components of the climate system, including global ocean heat pathways without climate change mitigation as well as new
uptake, warming of the atmosphere, warming of the land and melting low-CO2 emissions pathways (Figure TS.4). In these scenarios,
of ice. Compared to changes in global surface temperature, Earth’s differences in air pollution control and variations in climate
energy imbalance (see Core Concepts Box) exhibits less variability, change mitigation stringency strongly affect anthropogenic
enabling more accurate identification and estimation of trends. {Box emissions trajectories of SLCFs. Modelling studies relying
7.2 and Section 7.2} on the Representative Concentration Pathways (RCPs) used
in AR5 complement the assessment based on SSP scenarios,
Identifying the human-induced components contributing to the for example at the regional scale.
energy budget provides an implicit estimate of the human influence
on global climate change (Sections TS.2 and TS.3.1). {Cross-Working A comparison of simulations from CMIP5 using the RCPs
Group Box: Attribution in Chapter 1, 3.8, 7.2.2, Box 7.2, with SSP-based simulations from CMIP6 shows that about
Cross-Chapter Box 9.1} half of the increase in simulated warming in CMIP6 versus
CMIP5 arises because higher climate sensitivity is more
Regional climate changes can be moderated or amplified by regional prevalent in CMIP6 model versions; the other half arises
forcing from land-use and land-cover changes or from aerosol from higher radiative forcing in nominally corresponding
concentrations and other short-lived climate forcers (SLCFs). For scenarios (e.g., RCP8.5 and SSP5-8.5; medium confidence).
example, the difference in observed warming trends between cities The feasibility or likelihood of individual scenarios is not
and their surroundings can partly be attributed to urbanization part of this assessment, which focuses on the climate
(very high confidence). While established attribution techniques response to a large range of emissions scenarios. {1.5.4, 1.6,
provide confidence in our assessment of human influence on Cross-Chapter Box 1.4, 4.2, 4.3, 4.6, 6.6, 6.7, Cross-Chapter
large-scale climate changes (as described in Section TS.2), new Box 7.1, Atlas.2.1}
techniques developed since AR5, including attribution of individual
events, have provided greater confidence in attributing changes in Climate change projections with climate models require information
climate extremes to climate change (Box TS.10). Multiple attribution about future emissions or concentrations of greenhouse gases,
approaches support the contribution of human influence to several aerosols, ozone-depleting substances, and land use over time
regional multi-decadal mean precipitation changes (high confidence). (Figure TS.4). This information can be provided by scenarios, which
Understanding about past and future changes in weather and climate are internally consistent projections of these quantities based on
extremes has increased due to better observation-based datasets, assumptions of how socio-economic systems could evolve over the
physical understanding of processes, an increasing proportion 21st century. Emissions from natural sources, such as the ocean and

52
 Technical Summary

Human
activities
Other
CO2 (e.g., SLCF,
land use
albedo)

Non-CO2
Greenhouse
gases

Emissions N2O
HFCs etc.

200
NH3
Land use albedo etc.

900
175 NOx TS
125
CO2 800 CH4
100 700 150

MtNOx / yr
600 125

MtCH4 / yr
GtCO2 / yr

75 500 100
50 400
75
25 300
200 50
0 100 25
-25 0 0
1950 2000 2050 2100 1950 2000 2050 2100 1950 2000 2050 2100

Concentrations 4000
N2O
HFCs etc.
140
120
1000
CO2 3500 CH4 100

MtSO2 / yr
3000 80
800
2500
ppm

60
ppb

600 2000 40
400 1500 20 SO2
1000 0
200 1950 2000 2050 2100
1950 2000 2050 2100 1950 2000 2050 2100

12
Radiative
Effective radiative forcing

10
Forcing 8
6
Total anthropogenic
(W/m-2)

4
Carbon- 2
Cycle and 0
non-CO2 Natural
-2
biogeo- 1950 2000 2050 2100
chemical
feedbacks
7
Global
(°C rel. to 1850–1900)
Change in global surface

6
Warming 5 Projections
temperature

4
3
Legend:
2
Historical Observations
1
SSP5-8.5
SSP3-7.0 0
1950 2000 2050 2100
SSP2-4.5 2100
SSP1-2.6 RCP
SSP1-1.9 range Regional Climate Change
Temperature Precipitation
GWL 2°C GWL 2°C

Climatic
Impact-
Robust significant change
No or no robust -5 -4 -3 -2.5 -2 -1.5 0 1.5 2 2.5 3 4 5
Drivers -40 -20 0 +20 +40
Legend
see left
significant change Change in annual mean
Change in annual mean
Conflicting signal surface temperature (°C) precipitation (%)

Figure TS.4 | The climate change cause–effect chain: The intent of this figure is to illustrate the process chain starting from anthropogenic emissions, to changes in
atmospheric concentration, to changes in Earth’s energy balance (‘forcing’), to changes in global climate and ultimately regional climate and climatic impact-drivers. Shown
is the core set of five Shared Socio-economic Pathway (SSP) scenarios as well as emissions and concentration ranges for the previous Representative Concentration Pathway
(RCP) scenarios in year 2100; carbon dioxide (CO2) emissions (GtCO2 yr–1), panel top left; methane (CH4) emissions (middle) and sulphur dioxide (SO2), nitrogen oxide (NOx)
emissions (all in Mt yr–1), top right; concentrations of atmospheric CO2 (ppm) and CH4 (ppb), second row left and right; effective radiative forcing for both anthropogenic and
natural forcings (W m–2), third row; changes in global surface air temperature (°C) relative to 1850–1900, fourth row; maps of projected temperature change (°C) (left) and
changes in annual-mean precipitation (%) (right) at a global warming level (GWL) of 2°C relative to 1850–1900 (see also Figure TS.5), bottom row. Carbon cycle and non-CO2
biogeochemical feedbacks will also influence the ultimate response to anthropogenic emissions (arrows on the left). {1.6.1, Cross-Chapter Box 1.4, 4.2.2, 4.3.1, 4.6.1, 4.6.2}

53
 Technical Summary

the land biosphere, are usually assumed to be constant, or to evolve at the regional scale (Section TS.4). Scenario extensions are based
in response to changes in anthropogenic forcings or to projected on assumptions about the post-2100 evolution of emissions or of
climate change. Natural forcings, such as past changes in solar radiative forcing that are independent from the modelling of socio-
irradiance and historical volcanic eruptions, are represented in model economic dynamics, which does not extend beyond 2100. To explore
simulations covering the historical era. Future simulations assessed specific dimensions, such as air pollution or temporary overshoot of
in this Report account for projected changes in solar irradiance and a given warming level, scenario variants are used in addition to the
for the long-term mean background forcing from volcanoes, but not core set. {1.6.1, Cross-Chapter Box 1.4, 4.2.2, 4.2.6, 4.7.1, Cross-
for individual volcanic eruptions. Scenarios have a long history in Chapter Box 7.1}
IPCC as a method for systematically examining possible futures and
following the cause–effect chain: from anthropogenic emissions, to SSP1-1.9 represents the low end of future emissions pathways,
changes in atmospheric concentrations, to changes in Earth’s energy leading to warming below 1.5°C in 2100 and limited temperature
balance (‘forcing’), to changes in global climate and ultimately overshoot of 1.5°C over the course of the 21st century (see
regional climate and climatic impact-drivers (Figure TS.4, Section Figure TS.6). At the opposite end of the range, SSP5-8.5 represents
TS.2, Infographic TS.1). {1.5.4, 1.6.1, 4.2.2, 4.4.4, Cross-Chapter the very high warming end of future emissions pathways from the
TS Box 4.1, 11.1} literature. SSP3-7.0 has overall lower GHG emissions than SSP5-8.5
but, for example, CO2 emissions still almost double by 2100 compared
The uncertainty in climate change projections that results from to today’s levels. SSP2-4.5 and SSP1-2.6 represent scenarios with
assessing alternative socio-economic futures, the so-called scenario stronger climate change mitigation and thus lower GHG emissions.
uncertainty, is explored through the use of scenario sets. Designed SSP1-2.6 was designed to limit warming to below 2°C. Infographic
to span a wide range of possible future conditions, these scenarios TS.1 presents a narrative depiction of SSP-related climate futures.
do not intend to match how events actually unfold in the future, No likelihood is attached to the scenarios assessed in this Report,
and they do not account for impacts of climate change on the socio- and the feasibility of specific scenarios in relation to current trends
economic pathways. Besides scenario uncertainty, climate change is best informed by the WGIII contribution to AR6. In the scenario
projections are also subject to climate response uncertainty (i.e., the literature, the plausibility of some scenarios with high CO2 emissions,
uncertainty related to our understanding of the key physical processes such as RCP8.5 or SSP5-8.5, has been debated in light of recent
and structural uncertainties in climate models) and irreducible and developments in the energy sector. However, climate projections from
intrinsic uncertainties related to internal variability. Depending on these scenarios can still be valuable because the concentration levels
the spatial and temporal scales of the projection, and on the variable reached in RCP8.5 or SSP5-8.5 and corresponding simulated climate
of interest, the relative importance of these different uncertainties futures cannot be ruled out. That is because of uncertainty in carbon-
may vary substantially. {1.4.3, 1.6, 4.2.5, Box 4.1, 8.5.1} cycle feedbacks which, in nominally lower emissions trajectories, can
result in projected concentrations that are higher than the central
Scenarios in AR6 cover a broader range of emissions futures than concentration levels typically used to drive model projections. {1.6.1;
considered in AR5, including high CO2 emissions scenarios without Cross-Chapter Box 1.4; 4.2.2, 5.4; SROCC; Chapter 3 in WGIII}
climate change mitigation as well as a low CO2 emissions scenario
reaching net zero CO2 emissions (see Core Concepts Box) around mid- The socio-economic narratives underlying SSP-based scenarios
century. In this Report, a core set of five illustrative scenarios is used differ in their assumed level of air pollution control. Together with
to explore climate change over the 21st century and beyond (Section variations in climate change mitigation stringency, this difference
TS.2). They are labelled SSP1-1.9, SSP1-2.6, SSP2-4.5, SSP3-7.0, and strongly affects anthropogenic emissions trajectories of SLCFs,
SSP5-8.517 and span a wide range of radiative forcing levels in 2100. some of which are also air pollutants. SSP1 and SSP5 assume strong
They start in 2015 and include scenarios with high and very high pollution control, projecting a decline of global emissions of ozone
GHG emissions and CO2 emissions that roughly double from current precursors (except methane; CH4) and of aerosols and most of
levels by 2100 and 2050, respectively (SSP3-7.0 and SSP5-8.5); their precursors in the mid- to long term. The reductions due to air
scenarios with intermediate GHG emissions and CO2 emissions pollution controls are further strengthened in scenarios that assume
remaining around current levels until the middle of the century a marked decarbonization, such as SSP1-1.9 or SSP1-2.6. SSP2-4.5
(SSP2-4.5); and scenarios with very low and low GHG emissions and is a medium pollution-control scenario with air pollutant emissions
CO2 emissions declining to net zero around or after 2050, followed by following current trends, and SSP3-7.0 is a weak pollution-control
varying levels of net negative CO2 emissions (SSP1-1.9 and SSP1-2.6). scenario with strong increases in emissions of air pollutants over the
These SSP scenarios offer unprecedented detail of input data for ESM 21st century. Methane emissions in SSP-based scenarios vary with
simulations and allow for a more comprehensive assessment of the overall climate change mitigation stringency, declining rapidly
climate drivers and responses, in particular because some aspects, in SSP1-1.9 and SSP1-2.6 but declining only after 2070 in SSP5-8.5.
such as the temporal evolution of pollutants, emissions or changes in SSP trajectories span a wider range of air pollutant emissions than
land use and land cover, span a broader range in the SSP scenarios considered in the RCP scenarios (see Figure TS.4), reflecting the
than in the RCPs used in AR5. Modelling studies utilizing the RCPs potential for large regional differences in their assumed pollution
complement the assessment based on SSP scenarios, for example,

17 Throughout this Report, scenarios are referred to as SSPx-y, where “SSPx” refers to the Shared Socio-economic Pathway or “SSP” describing the socio-economic trends underlying the scenario,
and “y” refers to the approximate target level of radiative forcing (in W m–-2) resulting from the scenario in the year 2100.

54
 Technical Summary

policies. Their effects on climate and air pollution are assessed in Box TS.1.3.2 Global Warming Levels and Cumulative CO2 Emissions
TS.7. {4.4.4, 6.6.1, Figure 6.4, 6.7.1, Figure 6.19}
Quantifying geographical response patterns of climate
Since the RCPs are also labelled by the level of radiative forcing change at various global warming levels (GWLs), such as
they reach in 2100, they can in principle be related to the core set 1.5°C or 2°C above the 1850–1900 period, is useful for
of AR6 scenarios (Figure TS.4). However, the RCPs and SSP-based characterizing changes in mean climate, extremes and
scenarios are not directly comparable. First, the gas-to-gas climatic impact-drivers. Global warming levels are used
compositions differ; for example, the SSP5-8.5 scenario has higher in this Report as a dimension of integration independent
CO2 but lower CH4 concentrations compared to RCP8.5. Second, the of the timing when the warming level is reached and of
projected 21st-century trajectories may differ, even if they result the emissions scenario that led to the warming. For many
in the same radiative forcing by 2100. Third, the overall effective climate variables the response pattern for a given GWL is
radiative forcing (see Core Concepts Box) may differ, and tends consistent across different scenarios. However, this is not
to be higher for the SSPs compared to RCPs that share the same the case for slowly responding processes, such as ice-sheet
nominal stratospheric-temperature-adjusted radiative forcing label. and glacier mass loss, deep ocean warming, and the related
Comparing the differences between CMIP5 and CMIP6 projections sea level rise. The response of these variables depends on TS
(Cross-Section Box TS.1) that were driven by RCPs and SSP-based the time it takes to reach the GWL, differs if the warming is
scenarios, respectively, indicates that about half of the difference in reached in a transient warming state or after a temporary
simulated warming arises because of higher climate sensitivity being overshoot of the warming level, and will continue to evolve,
more prevalent in CMIP6 model versions; the remainder arises from over centuries to millennia, even after global warming
higher ERF in nominally corresponding scenarios (e.g., RCP8.5 and has stabilized. Different GWLs correspond closely to
SSP5-8.5; medium confidence) (see Section TS.1.2.2). In SSP1-2.6 and specific cumulative CO2 emissions due to their near-linear
SSP2-4.5, changes in ERF also explain about half of the changes in the relationship with global surface temperature. This Report
range of warming (medium confidence). For SSP5-8.5, higher climate uses 1.0°C, 1.5°C, 2.0°C, 3.0°C and 4.0°C above 1850–1900
sensitivity is the primary reason behind the upper end of the CMIP6- conditions as a primary set of GWLs. {1.6.2, 4.2.4, 4.6.1, 5.5,
projected warming being higher than for RCP8.5 in CMIP5 (medium Cross-Chapter Box 11.1, Cross-chapter Box 12.1}
confidence). Note that AR6 uses multiple lines of evidence beyond
CMIP6 results to assess global surface temperature under various For many indicators of climate change, such as seasonal and annual
scenarios (see Cross-Section Box TS.1 for the detailed assessment). mean and extreme surface air temperatures and precipitation, the
{1.6, 4.2.2, 4.6.2.2, Cross-Chapter Box 7.1} geographical patterns of changes are well estimated by the level of
global surface warming, independently of the details of the emissions
Earth system models can be driven by anthropogenic CO2 pathways that caused the warming, or the time at which the level of
emissions (‘emissions-driven’ runs), in which case atmospheric warming is attained. GWLs, defined as a global surface temperature
CO2 concentration is a projected variable; or by prescribed time- increase of, for example, 1.5°C or 2°C relative to the mean of 1850–
varying atmospheric concentrations (‘concentration-driven’ runs). In 1900, are therefore a useful way to integrate climate information
emissions-driven runs, changes in climate feed back on the carbon independently of specific scenarios or time periods. {1.6.2, 4.2.4,
cycle and interactively modify the projected CO2 concentration 4.6.1, 11.2.4, Cross-Chapter Box 11.1}
in each ESM, thus adding the uncertainty in the carbon cycle
response to climate change to the projections. Concentration- The use of GWLs allows disentangling the contribution of changes
driven simulations are based on a central estimate of carbon cycle in global warming from regional aspects of the climate response,
feedbacks, while emissions-driven simulations help quantify the role as scenario differences in response patterns at a given GWL are
of feedback uncertainty. The differences in the few ESMs for which often smaller than model uncertainty and internal variability. The
both emissions and concentration-driven runs were available for the relationship between the GWL and response patterns is often linear,
same scenario are small and do not affect the assessment of global but integration of information can also be done for non-linear
surface temperature projections discussed in Cross-Section Box TS.1 changes, like the frequency of heat extremes. The requirement is that
and Section TS.2 (high confidence). By the end of the 21st century, the relationship to the GWL is broadly independent of the scenario
emissions-driven simulations are on average around 0.1°C cooler and relative contribution of radiative forcing agents. {1.6, 11.2.4,
than concentration-driven runs, reflecting the generally lower CO2 Cross-Chapter Box 11.1}
concentrations simulated by the emissions-driven ESMs, and have a
spread about 0.1°C greater, reflecting the range of simulated CO2 The GWL approach to integration of climate information also has
concentrations. However, these carbon cycle–climate feedbacks do some limitations. Variables that are quick to respond to warming,
affect the transient climate response to cumulative CO2 emissions like temperature and precipitation, including extremes, sea ice area,
(TCRE18), and their quantification is crucial for the assessment of permafrost and snow cover, show little scenario dependence for a
remaining carbon budgets consistent with global warming levels given GWL, whereas slow-responding variables such as glacier and
simulated by ESMs (see Section TS.3). {1.6.1, Cross-Chapter Box 1.4, ice-sheet mass, warming of the deep ocean and their contributions
4.2, 4.3.1, 5.4.5, Cross-Chapter Box 7.1} to sea level rise, have substantial dependency on the trajectory of

18 The transient surface temperature change per unit of cumulative CO2 emissions, usually 1000 GtC.

55
 Technical Summary

warming taken to reach the GWL. A given GWL can also be reached SR1.5 concluded that ‘climate models project robust differences in
for different balances between anthropogenic forcing agents, such regional climate characteristics between present-day and global
as long-lived greenhouse gas and SLCF emissions, and the response warming of 1.5°C, and between 1.5°C and 2°C’. This Report adopts
patterns may depend on this balance. Finally, there is a difference a set of common GWLs across which climate projections, impacts,
in the response even for temperature-related variables if a GWL is adaptation challenges and climate change mitigation challenges can
reached in a rapidly warming transient state or in an equilibrium be integrated, within and across the three Working Groups, relative
state when the land–sea warming contrast is less pronounced. In this to 1850–1900. The core set of GWLs in this Report are 1.0°C (close
Report, the climate responses at different GWLs are calculated based to present day conditions), 1.5°C, 2.0°C, 3.0°C and 4.0°C. {1.4, 1.6.2,
on climate model projections for the 21st century (see Figure TS.5), Cross-Chapter Box 1.2, Table 1.5, Cross-Chapter Box 11.1}
which are mostly not in equilibrium. The SSP1-1.9 scenario allows
assessing the response to a GWL of about 1.5°C after a (relatively) Connecting Scenarios and Global Warming Levels
short-term stabilization by the end of the 21st century. {4.6.2, 9.3.1.1,
9.5.2.3, 9.5.3.3, 11.2.4, Cross-Chapter Box 11.1, Cross-Chapter In this Report, scenario-based climate projections are translated
Box 12.1} into GWLs by aggregating the ESM model response at specific
TS GWLs across scenarios (see Figure TS.5 and Figure TS.6). The climate
Global warming levels are highly relevant as a dimension of response pattern for the 20-year period around when individual
integration across scientific disciplines and socio-economic actors simulations reach a given GWL are averaged across all models and
and are motivated by the long-term goal in the Paris Agreement scenarios that reach that GWL. The best estimate and likely range
of ‘holding the increase in the global average temperature to well of the timing of when a certain GWL is reached under a particular
below 2°C above pre-industrial levels and to pursue efforts to limit scenario (or ‘GWL-crossing time’), however, is based not only on
the temperature increase to 1.5°C above pre-industrial levels’. CMIP6 output, but on a combined assessment taking into account
The evolution of aggregated impacts with temperature levels has the observed warming to date, CMIP6 output and additional lines
also been widely used and embedded in the WGII assessment. of evidence (see Cross-Section Box TS.1). {4.3.4, Cross-Chapter
This includes the ‘Reasons for Concern’ (RFC) and other ‘burning Box 11.1, Atlas.2, Interactive Atlas}
ember’ diagrams in IPCC WGII. The RFC framework has been further
expanded in SR1.5, SROCC and SRCCL by explicitly looking at the Global warming levels are closely related to cumulative CO2 (and
differential impacts between half-degree GWLs and the evolution of in some cases CO2-equivalent) emissions. This Report confirms the
risk for different socio-economic assumptions. {1.4.4, 1.6.2, 11.2.4, assessment of the WGI contribution to AR5 and SR1.5 that a near-
12.5.2, Cross-Chapter Box 11.1, Cross-Chapter Box 12.1} linear relationship exists between cumulative CO2 emissions and the

(a) Global mean temperature in CMIP6 (b) Patterns of change in near-surface air temperature, precipitation and soil moisture
5 SSP3-7.0 (20-yr GSAT means) Temperature change Precipitation change Soil moisture change
SSP1-2.6 (20-yr GSAT means) 48 47 43

+4°C +4°C
4

137 132 119

3
°C

+2°C

+2°C
2
+1.5°C 153 147 132

1 +1.5°C

2000 2020 2040 2060 2080 2100 -5 -4 -3-2.5-2 -1.5 0 1.5 2 2.5 3 4 5
°C
-40 -30 -20 -10 0 10 20 30 40
%
-2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5
σ

Figure TS.5 | Scenarios, global warming levels, and patterns of change. The intent of this figure is to show how scenarios are linked to global warming levels (GWLs)
and to provide examples of the evolution of patterns of change with global warming levels. (a) Illustrative example of GWLs defined as global surface temperature response
to anthropogenic emissions in unconstrained Coupled Model Intercomparison Project Phase 6 (CMIP6) simulations, for two illustrative scenarios (SSP1-2.6 and SSP3-7.0). The
time when a given simulation reaches a GWL, for example, +2°C, relative to 1850–1900 is taken as the time when the central year of a 20-year running mean first reaches
that level of warming. See the dots for +2°C, and how not all simulations reach all levels of warming. The assessment of the timing when a GWL is reached takes into account
additional lines of evidence and is discussed in Cross-Section Box TS.1. (b) Multi-model, multi-simulation average response patterns of change in near-surface air temperature,
precipitation (expressed as percentage change) and soil moisture (expressed in standard deviations of interannual variability) for three GWLs. The number to the top right of
the panels shows the number of model simulations averaged across including all models that reach the corresponding GWL in any of the five Shared Socio-economic Pathways
(SSPs). See Section TS.2 for discussion. {Cross-Chapter Box 11.1}

56
 Technical Summary

resulting increase in global surface temperature (Section TS.3.2). This as sub-continents and oceanic regions, or to typological regions, such
implies that continued CO2 emissions will cause further warming as monsoon regions, coastlines, mountain ranges or cities, as used in
and associated changes in all components of the climate system. For Section TS.4. A new set of standard AR6 WGI reference regions has
declining cumulative CO2 emissions (i.e., if negative net emissions also been included in this Report (Figure TS.6, bottom panels). {1.4.5,
are achieved), the relationship is less strong for some components, 10.1, 11.9, 12.1–12.4, Atlas.1.3.3–1.3.4}
such as the hydrological cycle. The WGI report uses cumulative CO2
emissions to compare climate response across scenarios and provides Global and regional climate models are important sources of climate
a link to the emissions pathways assessment in WGIII. The advantage information at the regional scale. Since AR5, a more comprehensive
of using cumulative CO2 emissions is that it is an inherent emissions assessment of past and future evolution of a range of climate variables
scenario characteristic rather than an outcome of the scenario-based on a regional scale has been enabled by the increased availability
projections, where uncertainties in the cause–effect chain from of coordinated ensemble regional climate model projections and
emissions to temperature change are important (Figure TS.4), for improvements in the level of sophistication and resolution of
example, the uncertainty in ERF and TCR. Cumulative CO2 emissions global and regional climate models. This has been complemented
can also provide a link to the assessments of mitigation options. by observational, attribution and sectoral-vulnerability studies
Cumulative CO2 emissions do not carry information about non-CO2 informing, for instance, about impact-relevant tolerance thresholds. TS
emissions, although these can be included with specific emissions {10.3.3, 11.9, 12.1, 12.3, 12.6, Atlas.3–Atlas.11}
metrics to estimate CO2-equivalent emissions. (Section TS.3.3) {1.3.2,
1.6, 4.6.2, 5.5, 7.6} Multiple lines of evidence derived from observations, model simulations
and other approaches can be used to construct climate information
on a regional scale as described in detail in Sections TS.4.1.1 and
TS.1.4 From Global to Regional Climate Information for TS.4.1.2. Depending on the phenomena and specific context, these
Impact and Risk Assessment sources and methodologies include theoretical understanding of the
relevant processes, drivers and feedbacks of climate at regional scale;
The AR6 WGI Report has an expanded focus on regional trends in observed data from multiple datasets; and the attribution
information supported by the increased availability of of these trends to specific drivers. Furthermore, simulations from
coordinated regional climate model ensemble projections different model types (including global and regional climate models,
and improvements in the sophistication and resolution emulators, statistical downscaling methods, etc.) and experiments
of global and regional climate models (high confidence). (e.g., CMIP, CORDEX, and large ensembles of single-model simulations
Multiple lines of evidence can be used to construct climate with different initial conditions), attribution methodologies and other
information on a global to regional scale and can be further relevant local knowledge (e.g., indigenous knowledge) are utilized
distilled in a co-production process to meet user needs (high (see Box TS.11). {1.5.3, 1.5.4, Cross-Chapter Box 7.1, 10.2–10.6, 11.2,
confidence). To better support risk assessment, a common Atlas.1.4, Cross-Chapter Box 10.3}
risk framework across all three Working Groups has been
implemented in AR6, and low-likelihood but high-impact From the multiple lines of evidence, climate information can be
outcomes are explicitly addressed in WGI by using physical distilled in a co-production process that involves users, related
climate storylines (see Core Concepts Box). stakeholders and producers of climate information, considering the
specific context of the question at stake, the underlying values and
Climatic impact-drivers are physical climate system the challenge of communicating across different communities. The
conditions (e.g., means, events, extremes) that affect co-production process is an essential part of climate services, which
an element of society or ecosystems. They are the WGI are discussed in Section TS.4.1.2. {10.5, 12.6, Cross-Chapter Box 12.2}
contribution to the risk framing without anticipating
whether their impact provides potential opportunities With the aim of informing decision-making at local or regional
or is detrimental (i.e., as for hazards). Many global and scales, a common risk framework has been implemented in AR6.
regional climatic impact-drivers have a direct relation to Methodologies have been developed to construct more impact- and
global warming levels (high confidence). {1.4.4, 1.5.2–1.5.4, risk-relevant climate change information tailored to regions and
Cross-Chapter Box 1.3, 4.8, 10.1, 10.5.1, Box 10.2, Cross- stakeholders. Physical storyline approaches are used in order to
Chapter Box 10.3, 11.2.4, 11.9, Box 11.2, Cross-Chapter build climate information based on multiple lines of evidence, and
Box 11.1, 12.1–12.3, 12.6, Cross-Chapter Boxes 12.1 and which can explicitly address physically plausible, but low-likelihood,
12.2, Atlas.1.3.3–1.3.4, Atlas.1.4, Atlas.1.4.4} high-impact outcomes and uncertainties related to climate variability
for consideration in risk assessments (Figure TS.6). {Cross-Chapter
Climate change is a global phenomenon, but manifests differently Box 1.3, 4.8, Box 9.4, 10.5, Box 10.2, Box 11.2, 12.1–12.3, 12.6,
in different regions. The impacts of climate change are generally Glossary}
experienced at local, national and regional scales, and these are
also the scales at which decisions are typically made. Robust climate The climatic impact-driver framework developed in AR6 supports
change information is increasingly available at regional scales for an assessment of changing climate conditions that are relevant
impact and risk assessments. Depending on the climate information for sectoral impacts and risk assessment. Climatic impact-drivers
context, geographical regions in AR6 may refer to larger areas, such (CIDs) are physical climate system conditions (e.g., means, extremes,

57
 Technical Summary

Schematic ECS likelihood Assessed changes in global surface temperature

Observations Projections
SSP5-8.5
6 Low-likelihood SSP3-7.0

High ECS
high warming SSP2-4.5
SSP1-2.6
Equilibrium climate sensitivity (ECS, °c)

5
SSP1-1.9

1980 2000 2020 2040 2060 2080

4 SSP5-8.5

Mid-range ECS
SSP3-7.0
Likely range SSP2-4.5
3
SSP1-2.6
TS SSP1-1.9
2 1980 2000 2020 2040 2060 2080
SSP5-8.5
Low-likelihood SSP3-7.0
1
Low ECS

low warming SSP2-4.5

SSP1-2.6
0 SSP1-1.9
Relative likelihood
Global Warming Level (GWL, °C)
0.5 1 1.5 2 3 4

GWL 1.5°C GWL 2.0°C GWL 4.0°C

Heat
warning
index

1 day/year 3 days/year 2 weeks/year 1 month/year 3 months/year 6 months/year

GWL 1.5°C GWL 2.0°C GWL 4.0°C

Changes in
extreme
rainfall

0 5 10 15 20 25 30
% more rain on wettest day of the year

Figure TS.6 | A graphical abstract for key aspects of the Technical Summary. The intent of this figure is to summarize many different aspects of the Technical Summary
related to observed and projected changes in global temperature and associated regional changes in climatic impact-drivers relevant for impact and risk assessment. Top left:
a schematic representation of the likelihood for equilibrium climate sensitivity (ECS), consistent with the AR6 assessment (see Chapter 7 and Section TS.3). ECS values above
5°C and below 2°C are termed low-likelihood, high warming (LLHW) and low-likelihood, low warming, respectively (Box TS.3). Top right: Observed (see Cross-Section Box
TS.1) and projected global surface temperature changes, shown as global warming levels (GWLs) relative to 1850–1900, using the assessed 95% (top), 50% (middle) and 5%
(bottom) likelihood time series (see Chapter 4 and Section TS.2). Bottom panels show maps of Coupled Model Intercomparison Project Phase 6 (CMIP6) median projections of
two climatic impact-drivers (CIDs, see Section TS.1.4) at three different GWLs (columns for 1.5, 2 and 4°C) for the AR6 land regions (see Chapters 1, 10, and Atlas and Section
TS.4). The heat warning index is the number of days per year averaged across each region at which a heat warning for human health at level ‘danger’ would be issued according
to the U.S. National Oceanic and Atmospheric Administration (NOAA) (NOAA HI41, see Chapter 12 and Annex VI). The maps of extreme rainfall changes show the percentage
change in the amount of rain falling on the wettest day of a year (Rx1day, relative to 1995–2014, see Chapter 11) averaged across each region when the respective GWL is
reached. Additional CIDs are discussed in Section TS.4. {1.4.4, Box 4.1, 7.5, 11.4.3, 12.4}

58
 Technical Summary

events) that affect an element of society or ecosystems and are thus Many global- and regional-scale CIDs, including extremes, have
a potential priority for providing climate information. For instance, a direct relation to global warming levels (GWLs) and can thus inform
the heat index used by the U.S. National Oceanic and Atmospheric the hazard component of ‘Representative Key Risks’ and ‘Reasons for
Administration (NOAA HI) for issuing heat warnings is a CID index Concern’ assessed by AR6 WGII. These include heat, cold, wet and dry
that can be associated with adverse human health impacts due to hazards, both mean and extremes; cryospheric hazards (snow cover,
heat stress (see Figure TS.6). Depending on system tolerance, CIDs and ice extent, permafrost) and oceanic hazards (marine heatwaves)
their changes can be detrimental (i.e., hazards in the risk framing), (high confidence) (Figure TS.6). Establishing links between specific
beneficial, neutral, or a mixture of each across interacting system GWLs with tipping points and irreversible behaviour is challenging
elements, regions and sectors (aligning with WGII Sectoral Chapters due to model uncertainties and lack of observations, but their
2–8). Each sector is affected by multiple CIDs, and each CID affects occurrence cannot be excluded, and their likelihood of occurrence
multiple sectors. Climate change has already altered CID profiles and generally increases at greater warming levels (Box TS.1, Section TS.9).
resulted in shifting magnitude, frequency, duration, seasonality and {11.2.4, Box 11.2, Cross-Chapter Boxes 11.1 and 12.1}
spatial extent of associated indices (high confidence) (see regional
details in Section TS.4.3). {12.1–12.4, Table 12.1, Table 12.2, Annex VI}
TS

Cross-Section Box TS.1: Global Surface Temperature Change

This box synthesizes the outcomes of the assessment of past, current and future global surface temperature. Global mean surface
temperature (GMST) and global surface air temperature (GSAT) are the two primary metrics of global surface temperature used to
estimate global warming in IPCC reports. GMST merges sea surface temperature (SST) over the ocean and 2 m air temperature over
land and sea ice areas and is used in most paleo, historical and present-day observational estimates. The GSAT metric is 2 m air
temperature over all surfaces and is the diagnostic generally used from climate models. Changes in GMST and GSAT over time differ
by at most 10% in either direction (high confidence), but conflicting lines of evidence from models and direct observations, combined
with limitations in theoretical understanding, lead to low confidence in the sign of any difference in long-term trend. Therefore,
long-term changes in GMST/GSAT are presently assessed to be identical, with expanded uncertainty in GSAT estimates. Hence the term
global surface temperature is used in reference to both quantities in the text of the TS and SPM. {Cross-Chapter Box 2.3}

Global surface temperature has increased by 0.99 [0.84 to 1.10] °C from 1850–1900 to the first two decades of the
21st century (2001–2020) and by 1.09 [0.95 to 1.20] °C from 1850–1900 to 2011–2020. Temperatures as high as
during the most recent decade (2011–2020) exceed the warmest centennial-scale range reconstructed for the present
interglacial, around 6500 years ago [0.2°C to 1°C] (medium confidence). The next most recent warm period was
about 125,000 years ago during the last interglacial when the multi-centennial temperature range [0.5°C to 1.5°C]
encompasses the 2011–2020 values (medium confidence). The likely range of human-induced change in global surface
temperature in 2010–2019 relative to 1850–1900 is 0.8°C to 1.3°C, with a central estimate of 1.07°C, encompassing
the best estimate of observed warming for that period, which is 1.06°C with a very likely range of [0.88°C to 1.21°C],
while the likely range of the change attributable to natural forcing is only –0.1°C to +0.1°C.

Compared to 1850–1900, average global surface temperature over the period 2081–2100 is very likely to be higher by
[1.0°C to 1.8°C] in the low CO2 emissions scenario SSP1-1.9 and by [3.3°C to 5.7°C] in the high CO2 emissions scenario
SSP5-8.5. In all scenarios assessed here except SSP5-8.5, the central estimate of 20-year averaged global surface
warming crossing the 1.5°C level lies in the early 2030s, which is in the early part of the likely range (2030–2052)
assessed in SR1.5. It is more likely than not that under SSP1-1.9, global surface temperature relative to 1850–1900 will
remain below 1.6°C throughout the 21st century, implying a potential temporary overshoot of 1.5°C global warming
of no more than 0.1°C. Global surface temperature in any individual year could exceed 1.5°C relative to 1850–1900
by 2030 with a likelihood between 40% and 60% across the scenarios considered here (medium confidence). A 2°C
increase in global surface temperature relative to 1850–1900 will be crossed under SSP5-8.5 but is extremely unlikely
to be crossed under SSP1-1.9. Periods of reduced and increased global surface temperature trends at decadal time
scales will continue to occur in the 21st century (very high confidence). The effect of strong mitigation on 20-year
global surface temperature trends would be likely to emerge during the near term (2021–2040), assuming no major
volcanic eruptions occur. (Figure TS.8, Cross-Section Box TS.1, Figure 1) {2.3, 3.3, 4.3, 4.4, 4.5, 4.6, 7.3}

Surface Temperature History

Dataset innovations, particularly more comprehensive representation of polar regions, and the availability of new datasets have led
to an assessment of increased global surface temperature change relative to the directly equivalent estimates reported in AR5. The
contribution of changes in observational understanding alone between AR5 and AR6 in assessing temperature changes from 1850–1900

59
 Technical Summary

Cross-Section Box TS.1 (continued)

to 1986–2005 is estimated at 0.08 [–0.01 to 0.12] °C. Global surface temperature increased from 1850–1900 to 1995–2014 by 0.85
[0.69 to 0.95] °C, between 1850–1900 and the first two decades of the 21st century (2001–2020) by 0.99 [0.84 to 1.20] °C, and to
the most recent decade (2011–2020) by 1.09 [0.95 to 1.20] °C. Each of the last four decades has in turn been warmer than any decade
that preceded it since 1850. Temperatures have increased faster over land than over the ocean since 1850–1900, with warming to
2011–2020 of 1.59 [1.34 to 1.83] °C over land and 0.88 [0.68 to 1.01] °C over the ocean. {2.3.1, Cross-Chapter Box 2.3}

Global surface temperature during the period 1850–1900 is used as an approximation for pre-industrial conditions for consistency
with AR5 and AR6 Special Reports, whilst recognizing that radiative forcings have a baseline of 1750 for the start of anthropogenic
influences. It is likely that there was a net anthropogenic forcing of 0.0–0.3 Wm–2 in 1850–1900 relative to 1750 (medium confidence),
and from the period around 1750 to 1850–1900, there was a change in global surface temperature of around 0.1°C (likely range
–0.1 to +0.3°C, medium confidence), with an anthropogenic component of 0.0°C to 0.2°C (likely range, medium confidence). {Cross-
TS Chapter Box 1.2, 7.3.5}

Global surface temperature has evolved over geological time (Figure TS.1, Box TS.2). Beginning approximately 6500 years ago, global
surface temperature generally decreased, culminating in the coldest multi-century interval of the post-glacial period (since roughly
7000 years ago), which occurred between around 1450 and 1850 (high confidence). Over the last 50 years, global surface temperature
has increased at an observed rate unprecedented in at least the last two thousand years (high confidence). Temperatures as high as
during the most recent decade (2011–2020) exceed the warmest centennial-scale range reconstructed for the present interglacial,
around 6500 years ago [0.2°C to 1°C] (medium confidence). The next most recent warm period was about 125,000 years ago during
the Last Interglacial when the multi-centennial temperature range [0.5°C to 1.5°C] encompasses the 2011–2020 values (medium
confidence) (Cross-Section Box TS.1, Figure 1). During the mid-Pliocene Warm Period, around 3.3–3.0 million years ago, global surface
temperature was 2.5°C to 4°C warmer (medium confidence). {2.3.1, Cross-Chapter Box 2.1 and 2.4}

Current Warming
There is very high confidence that the CMIP6 model ensemble reproduces observed global surface temperature trends and variability
since 1850 with errors small enough to allow for detection and attribution of human-induced warming. The CMIP6 multi-model mean
global surface warming between 1850–1900 and 2010–2019 is close to the best estimate of observed warming, though some CMIP6
models simulate a warming that is outside the assessed very likely observed range. {3.3.1}

The likely range of human-induced change in global surface temperature in 2010–2019 relative to 1850–1900 is 0.8°C to 1.3°C,
with a central estimate of 1.07°C (Figure Cross-Section Box TS.1, Figure 1), encompassing the best estimate of observed warming
for that period, which is 1.06°C with a very likely range of [0.88°C to 1.21°C], while the likely range of the change attributable
to natural forcing is only –0.1°C to +0.1°C. This assessment is consistent with an estimate of the human-induced global surface
temperature rise based on assessed ranges of perturbations to the top of the atmosphere (effective radiative forcing) and with metrics
of feedbacks of the climate response (equilibrium climate sensitivity and the transient climate response). Over the same period, well-
mixed greenhouse gas forcing likely warmed global surface temperature by 1.0°C to 2.0°C, while aerosols and other anthropogenic
forcings likely cooled global surface temperature by 0.0°C to 0.8°C. {2.3.1, 3.3.1, 7.3.5, Cross-Chapter Box 7.1}

The observed slower increase in global surface temperature (relative to preceding and following periods) in the 1998–2012 period,
sometimes referred to as ‘the hiatus’, was temporary (very high confidence). The increase in global surface temperature during the
1998–2012 period is also greater in the data sets used in the AR6 assessment than in those available at the time of AR5. Using these
updated observational data sets and a like-for-like consistent comparison of simulated and observed global surface temperature, all
observed estimates of the 1998–2012 trend lie within the very likely range of CMIP6 trends. Furthermore, the heating of the climate
system continued during this period, as reflected in the continued warming of the global ocean (very high confidence) and in the
continued rise of hot extremes over land (medium confidence). Since 2012, global surface temperature has risen strongly, with the
past five years (2016–2020) being the hottest five-year period between 1850 and 2020 (high confidence). {2.3.1, 3.3.1, 3.5.1, Cross-
Chapter Box 3.1}

Future Changes in Global Surface Temperature

The AR6 assessment of future change in global surface temperature is, for the first time in an IPCC report, explicitly constructed by
combining new projections for the SSP scenarios with observational constraints based on past simulated warming as well as the AR6-
updated assessment of equilibrium climate sensitivity and transient climate response. In addition, climate forecasts initialized from
the observed climate state have been used for the period 2019–2028. The inclusion of additional lines of evidence has reduced the
assessed uncertainty ranges for each scenario (Cross-Section Box TS.1, Figure 1). {4.3.1, 4.3.4, Box 4.1, 7.5}

60
 Technical Summary

Changes in surface temperature

(a) Global surface temperatures are more likely than not unprecedented in the past 125,000 years
Kaufman et al. The latest decade was warmer than any multi-century period
1.5 Pages 2k after the Last Interglacial, around 125,000 years ago
AR6 mean
Over the last 50 years, global temperature
Global surface temperature
relative to 1850–1900 (°C)

1.0 has increased at a rate unprecedented in

at least the last 2000 years

0.5

Mid-Holocene
0.0

Last Interglacial
Latest decade
−0.5 TS
−10,000 −6000 −2000 1000 1400 1800 1900 2000
Year (BCE) Year (CE)
Age: Agricultural Historical Industrial
Resolution: Centuries Decades Annual
(b) Observed and projected warming are stronger over (c) Global surface temperature has risen more than 1°C
land than oceans, and strongest in the Arctic from 1850–1900
1.0 HadCRUT .5.0 NOAAGlobalTemp
Kadow et al. Berkeley Earth
1981–2020 0.5
Global surface temperature relative to 1850–1900 (°C)

0.0

−0.5
1850 1900 1950 2000
(d) Internal variability will influence near-term warming rates
2.5 CMIP6 historical
HadCRUT5
assessed SSP1-2.6
2.0 assessed SSP2-4.5
assessed SSP3-7.0
1.5 models 1-4

−0.6 −0.4 −0.2 −0.1 0.0 0.1 0.2 0.4 0.6 1.0

Colour Significant Trend (°C per decade)

0.5
Non significant
0.0
1980 1990 2000 2010 2020 2030 2040 2050
SSP3-7.0 (2081–2100) (e) Warming to 2100 depends on the scenario
5
GSAT 2081–2100 relative to 1995–2014 (°C)

SSP1-1.9
SSP1-2.6
SSP2-4.5 5
4 SSP3-7.0
Relative to 1850–1900 (°C)

SSP5-8.5
4
3

3
2

1 2
−6 −4 −3 −2 −1 −0.5 0 0.5 1 2 3 4 6 1.5 °C
Total change (°C)
1
0
2000–2019 2020–2039 2040–2059 2060–2079 2081–2100

Cross-Section Box TS.1, Figure 1 | Earth’s surface temperature history and future with key findings annotated within each panel.

61
 Technical Summary

Cross-Section Box TS.1 (continued)

Cross-Section Box TS.1, Figure 1 (continued): The intent of this figure is to show global surface temperature observed changes from the Holocene to now,
and projected changes. (a) Global surface temperature over the Holocene divided into three time scales: (i) 12,000 to 1000 years ago (10,000 BCE to 1000 CE)
in 100-year time steps, (ii) 1000 to 1900 CE, 10-year smooth, and (iii) 1900 to 2020 CE (mean of four datasets in panel c). Bold lines show the median of the
multi-method reconstruction, with 5% and 95% percentiles of the ensemble members (thin lines). Vertical bars are 5–95th percentile ranges of estimated global
surface temperature for the Last Interglacial and mid-Holocene (medium confidence) (Section 2.3.1.1). All temperatures are relative to 1850–1900. (b) Spatially
resolved trends (°C per decade) for (upper map) HadCRUTv5 over 1981–2020, and (lower map, total change) multi-model mean projected changes from 1995–2014
to 2081–2100 in the SST3-7.0 scenario. Observed trends have been calculated where data are present in both the first and last decade and for at least 70% of all
years within the period using ordinary least squares. Significance is assessed with autoregressive AR(1) model correction and denoted by stippling. Hatched areas
in the lower map show areas of conflicting model evidence on significance of changes. (c) Temperature from instrumental data for 1850–2020, including annually
resolved averages for the four global surface temperature datasets assessed in Section 2.3.1.1.3 (see text for references). The grey shading shows the uncertainty
associated with the HadCRUTv5 estimate. All temperatures are relative to the 1850–1900 reference period. (d) Recent past and 2015–2050 evolution of annual mean
global surface temperature change relative to 1850–1900, from HadCRUTv5 (black), Coupled Model Intercomparison Project Phase 6 (CMIP6) historical simulations
(up to 2014, in grey, ensemble mean solid, 5% and 95% percentiles dashed, individual models thin), and CMIP6 projections under scenario SSP2-4.5, from four
models that have an equilibrium climate sensitivity near the assessed central value (thick yellow). Solid thin coloured lines show the assessed central estimate of
20-year change in global surface temperature for 2015–2050 under three scenarios, and dashed thin coloured lines the corresponding 5% and 95% quantiles. (e)
TS Assessed projected change in 20-year running mean global surface temperature for five scenarios (central estimate solid, very likely range shaded for SSP1-2.6 and
SSP3-7.0), relative to 1995–2014 (left y-axis) and 1850–1900 (right y-axis). The y-axis on the right-hand side is shifted upward by 0.85°C, the central estimate of the
observed warming for 1995–2014, relative to 1850–1900. The right y-axis in (e) is the same as the y-axis in (d). {2.3, 4.3, 4.4}

During the near term (2021–2040), a 1.5°C increase in global surface temperature, relative to 1850–1900, is very likely to occur
in scenario SSP5-8.5, likely to occur in scenarios SSP2-4.5 and SSP3-7.0, and more likely than not to occur in scenarios SSP1-1.9 and
SSP1-2.6. The time of crossing a warming level is defined here as the midpoint of the first 20-year period during which the average
global surface temperature exceeds the level. In all scenarios assessed here except SSP5-8.5, the central estimate of crossing the 1.5°C
level lies in the early 2030s. This is in the early part of the likely range (2030–2052) assessed in SR1.5, which assumed continuation of
the then-current warming rate; this rate has been confirmed in the AR6. Roughly half of this difference arises from a larger historical
warming diagnosed in AR6. The other half arises because for central estimates of climate sensitivity, most scenarios show stronger
warming over the near term than was estimated as ‘current’ in SR1.5 (medium confidence). When considering scenarios similar to SSP1-
1.9 instead of linear extrapolation, the SR1.5 estimate of when 1.5°C global warming is crossed is close to the central estimate reported
here. (Cross-Section Box TS.1, Table 1) {2.3.1, Cross-Chapter Box 2.3, 3.3.1, 4.3.4, Box 4.1}

It is more likely than not that under SSP1-1.9, global surface temperature relative to 1850–1900 will remain below 1.6°C throughout
the 21st century, implying a potential temporary overshoot of 1.5°C global warming of no more than 0.1°C. If climate sensitivity lies
near the lower end of the assessed very likely range, crossing the 1.5°C warming level is avoided in scenarios SSP1-1.9 and SSP1-2.6
(medium confidence). Global surface temperature in any individual year, in contrast to the 20-year average, could by 2030 exceed
1.5°C relative to 1850–1900 with a likelihood between 40% and 60%, across the scenarios considered here (medium confidence).
(Cross-Section Box TS.1, Table 1) {4.3.4, 4.4.1, Box 4.1, 7.5}

During the 21st century, a 2°C increase in global surface temperature relative to 1850–1900 will be crossed under SSP5-8.5 and
SSP3-7.0, is extremely likely to be crossed under SSP2-4.5, but is unlikely to be crossed under SSP1-2.6 and extremely unlikely to be
crossed under SSP1-1.9. For the mid-term period 2041–2060, this 2°C global warming level is very likely to be crossed under SSP5-8.5,
likely to be crossed under SSP3-7.0, and more likely than not to be crossed under SSP2-4.5. (Cross-Section Box TS.1, Table 1) {4.3.4}

Events of reduced and increased global surface temperature trends at decadal time scales will continue to occur in the 21st century
but will not affect the centennial-scale warming (very high confidence). If strong mitigation is applied from 2020 onward as reflected
in SSP1-1.9, its effect on 20-year trends in global surface temperature would likely emerge during the near term (2021–2040),
measured against an assumed non-mitigation scenario such as SSP3-7.0 or SSP5-8.5. All statements about crossing the 1.5°C level
assume that no major volcanic eruption occurs during the near term (Cross-Section Box TS.1, Table 1). {2.3.1, Cross-Chapter Box 2.3,
4.3.4, 4.4.1, 4.6.3, Box 4.1}

Compared to 1850–1900, average global surface temperature over the period 2081–2100 is very likely to be higher by [1.0°C to
1.8°C] in the low CO2 emissions scenario SSP1-1.9 and by [3.3°C to 5.7°C] in the high CO2 emissions scenario SSP5-8.5. For the
scenarios SSP1-2.6, SSP2-4.5, and SSP3-7.0, the corresponding very likely ranges are [1.3°C to 2.4°C], [2.1°C to 3.5°C], and [2.8°C
to 4.6°C], respectively. The uncertainty ranges for the period 2081–2100 continue to be dominated by the uncertainty in equilibrium
climate sensitivity and transient climate response (very high confidence) (Cross-Section Box TS.1, Table 1). {4.3.1, 4.3.4, 4.4.1, 7.5}

The CMIP6 models project a wider range of global surface temperature change than the assessed range (high confidence); furthermore,
the CMIP6 global surface temperature increase tends to be larger than that in CMIP5 (very high confidence). {4.3.1, 4.3.4, 4.6.2, 7.5.6}

62
 Technical Summary

Cross-Section Box TS.1 (continued)

Cross-Section Box TS.1, Table 1 | Assessment results for 20-year averaged change in global surface temperature based on multiple lines of
evidence. The change is displayed in °C relative to the 1850–1900 reference period for selected time periods (first three rows), and as the first 20-year period during
which the average global surface temperature change exceeds the specified level relative to the period 1850–1900 (last four rows). The entries give both the central
estimate and, in parentheses, the very likely (5–95%) range. An entry n.c. means that the global warming level is not crossed during the period 2021–2100.

SSP1-1.9 SSP1-2.6 SSP2-4.5 SSP3-7.0 SSP5-8.5

Near term,
1.5 [1.2 to 1.7] 1.5 [1.2 to 1.8] 1.5 [1.2 to 1.8] 1.5 [1.2 to 1.8] 1.6 [1.3 to 1.9]
2021–2040

Mid-term,
1.6 [1.2 to 2.0] 1.7 [1.3 to 2.2] 2.0 [1.6 to 2.5] 2.1 [1.7 to 2.6] 2.4 [1.9 to 3.0]
2041–2060

Long term,
1.4 [1.0 to 1.8] 1.8 [1.3 to 2.4] 2.7 [2.1 to 3.5] 3.6 [2.8 to 4.6] 4.4 [3.3 to 5.7]
2081–2100

1.5°C
2025–2044 2023–2042 2021–2040 2021–2040 2018–2037 TS
[2013–2032 to n.c.] [2012–2031 to n.c.] [2012–2031 to 2037–2056] [2013–2032 to 2033–2052] [2011–2030 to 2029–2048]

n.c. n.c. 2043–2062 2037–2056 2032–2051

2°C
[n.c. to n.c.] [2031–2050 to n.c.] [2028–2047 to 2075–2094] [2026–2045 to 2053–2072] [2023–2042 to 2044–2063]

n.c. n.c. n.c. 2066–2085 2055–2074

3°C
[n.c. to n.c.] [n.c. to n.c.] [2061–2080 to n.c.] [2050–2069 to n.c.] [2042–2061 to 2074–2093]

n.c. n.c. n.c. n.c. 2075–2094

4°C
[n.c. to n.c.] [n.c. to n.c.] [n.c. to n.c.] [2070–2089 to n.c.] [2058–2077 to n.c.]

TS.2 Large-scale Climate Change: Mean substantial reductions in global GHG emissions. Continued
Climate, Variability and Extremes GHG emissions greatly increase the likelihood of potentially
irreversible changes in the global climate system (Box TS.9),
This section summarizes knowledge about observed and projected in particular with respect to the contribution of ice sheets
large-scale climate change (including variability and extremes), to global sea level change (high confidence). {2.3, 3.8, 4.3,
drivers and attribution of observed changes to human activities. It 4.6, 4.7, 7.2–7.4, Cross-Chapter Box 7.1, 9.2–9.6}
describes observed and projected large-scale changes associated
with major components of the climate system: atmosphere, ocean Earth system model simulations of the historical period since 1850
(including sea level change), land, biosphere and cryosphere, and the are only able to reproduce the observed changes in key climate
carbon, energy and water cycles. In each subsection, reconstructed indicators when anthropogenic forcings are included (Figure TS.7).
past changes, observed and attributed recent changes, and projected Taken together with numerous formal attribution studies across an
near- and long-term changes to mean climate, variability and even broader range of indicators and theoretical understanding,
extremes are presented, where possible, in an integrated way. See this underpins the unequivocal attribution of observed warming of
Section TS.1.3.1 for information on the scenarios used for projections. the atmosphere, ocean, and land to human influence (Table TS.1).
{2.3, 3.8}

TS.2.1 Changes Across the Global Climate System

In addition to global surface temperature (Cross-Section Box

TS.1), a wide range of indicators across all components of
the climate system are changing rapidly (Figure TS.7), with
many at levels unseen in millennia. The observed changes
provide a coherent picture of a warming world, many
aspects of which have now been formally attributed to
human influence, and human influence on the atmosphere,
ocean, and land components of the climate system, taken
together, is assessed as unequivocal for the first time in an
IPCC assessment report (Table TS.1, Figure TS.7).

It is virtually certain that global surface temperature rise

and associated changes can be limited through rapid and

63
 Technical Summary

Global
Near-surface air temperature
over land Near-surface air temperature Ocean heat content

Near-surface air temperature Sea ice

North America Asia Arctic
TS

Central and South America Australasia Antarctic

Europe and North Africa Antarctic Precipitation

Global

Africa Arctic
60°N–90°N

anthropogenic + natural natural observations

Figure TS.7 | Simulated and observed changes compared to the 1850–1900 average in key large-scale indicators of climate change across the climate
system, for continents, ocean basins and globally up to 2014. The intent of this figure is to compare the observed and simulated changes over the historical period
for a range of variables and regions, with and without anthropogenic forcings, for attribution. Black lines show observations, orange lines and shading show the multi-model
mean and 5–95th percentile ranges for Coupled Model Intercomparison Project Phase 6 (CMIP6) historical simulations including anthropogenic and natural forcing, and green
lines and shading show corresponding ensemble means and 5–95th percentile ranges for CMIP6 natural-only simulations. Observations after 2014 (including, for example, a
strong subsequent decrease of Antarctic sea ice area that leads to no significant overall trend since 1979) are not shown because the CMIP6 historical simulations end in 2014.
A 3-year running mean smoothing has been applied to all observational time series. {3.8, Figure 3.41}

64
 Technical Summary

Table TS.1 | Assessment of observed changes in large-scale indicators of mean climate across climate system components and their attribution to human
influence. The colour coding indicates the assessed confidence in/likelihood of the human contribution as a driver or main driver19 (main driver is specified in that case) where
available (see colour key). Otherwise, explanatory text is provided in cells with white background. The relevant chapter section with more detailed information is listed in each
table cell.

Change in Indicator Observed Change Assessment Human Contribution Assessment

Atmosphere and Water Cycle

Warming of global mean surface air temperature since Likely range of human contribution (0.8°C–1.3°C) encom-
{2.3.1, Cross-Chapter Box 2.3}
1850–1900 passes observed warming (0.9°C–1.2°C) {3.3.1}

Warming of the troposphere since 1979 {2.3.1} Main driver {3.3.1}

Cooling of the lower stratosphere Since mid-20th century {2.3.1} Main driver 1979–mid-1990s {3.3.1}
Large-scale precipitation and upper troposphere humidity
{2.3.1} {3.3.2, 3.3.3}
changes since 1979
Expansion of the zonal mean Hadley Circulation since
{2.3.1} Southern Hemisphere {3.3.3}
the 1980s
Ocean TS
Ocean heat content increase since the 1970s {2.3.3, 2.3.4, 9.2.1, Cross-Chapter Box 9.1} Main driver {3.5.1}
Salinity changes since the mid-20th century {2.3.3, 2.3.4, 9.2.2} {3.5.2}
Global mean sea level rise since 1971 {2.3.3, 9.6.1} Main driver {3.5.3}
Cryosphere
Arctic sea ice loss since 1979 {2.3.2, 9.3.1} Main driver {3.4.1}
Reduction in Northern Hemisphere spring snow cover since
{2.3.2, 9.5.3} {3.4.2}
1950
Greenland Ice Sheet mass loss since 1990s {2.3.2, 9.4.1} {3.4.3}
Antarctic Ice Sheet mass loss since 1990s {2.3.2, 9.4.2} Limited evidence and medium agreement {3.4.3}
Retreat of glaciers {2.3.2, 9.5.1} Main driver {3.4.3}
Carbon Cycle
Increased amplitude of the seasonal cycle of atmospheric
{2.3.4} Main driver {3.6.1}
CO2 since the early 1960s
Acidification of the global surface ocean {SROCC, 5.3.2, Cross-Chapter Box 5.3} Main driver {3.6.2}
Land Climate (Extremes, see Table TS.12)
Mean 2 m land warming since 1850–1900 (about 40%
{2.3.1} Main driver {3.3.1}
larger than global mean warming)
Synthesis
Warming of the global climate system since
{2.3.5} {3.8.1}
pre-industrial times

See text description medium confidence likely/high confidence very likely extremely likely virtually certain fact

Future climate change across a range of atmospheric, cryospheric, centuries to millennia. Furthermore, it is likely that at least one large
oceanic and biospheric indicators depends upon future emissions volcanic eruption will occur during the 21st century. Such an eruption
pathways. Outcomes for a broad range of indicators increasingly would reduce global surface temperature for several years, decrease
diverge through the 21st century across the different SSPs (Section land precipitation, alter monsoon circulation and modify extreme
TS.1.3.1, Figure TS.8). Due to the slow response of the deep ocean precipitation, at both global and regional scales. {4.3, 4.7, 9.4, 9.6,
and ice sheets, this divergence continues long after 2100, and 21st Cross-Chapter Box 4.1}
century emissions choices will have implications for GMSL rise for

19 Throughout this Technical Summary, ‘main driver’ means responsible for more than 50% of the change.

65
 Technical Summary

Recent and future change of four key indicators of the climate system
Atmospheric temperature, ocean heat content, Arctic summer sea ice, and land precipitation
(a) Global surface air temperature (b) Global ocean heat content and thermosteric sea level
4
CMIP6 Emulator
Future (assessed) 3
3 Future (assessed) 0.3

2 2 0.2

m
°C

YJ
1 1 Past (observed) 0.1
Past (observed)
0 Past (simulated) 0 Past (simulated) 0

–1 –1 -0.1
1950 2000 [2000/ [2040/ [2080/ Change 1950 2000 2050 2100 Change
2019] 2059] 2100] in 2100 in 2100

(c) Arctic September sea ice area (d) Global land precipitation
10 20
TS 8 Past (simulated) Future (CMIP6)
Past (observed)
10
6
million km2

Past (observed)
Future (CMIP6)

%
4 0
2
Practically sea ice free Past (simulated)
0 –10
1950 2000 2050 2100 Change 1950 2000 2050 2100 Change
in 2100
in 2100

Past (simulated); 5–95% range Future (SSP1-1.9) mean

Past (observed) Future (SSP1-2.6) mean; 5–95% range
Future (SSP2-4.5) mean
Future (SSP3-7.0) mean; 5–95% range
Future (SSP5-8.5) mean

Figure TS.8 | Observed, simulated and projected changes compared to the 1995–2014 average in four key indicators of the climate system through to
2100 differentiated by Shared Socio-economic Pathway (SSP) scenario. The intent of this figure is to show how future emissions choices impact key, iconic large-scale
indicators and to highlight that our collective choices matter. Past simulations are based on the Coupled Model Intercomparison Project Phase 6 (CMIP6) multi-model ensemble.
Future projections are based on the assessed ranges based upon multiple lines of evidence for (a) global surface temperature (Cross-Section Box TS.1) and (b) global ocean heat
content and the associated thermosteric sea level contribution to global mean sea level change (right-hand axis) using a climate model emulator (Cross-Chapter Box 7.1), and
CMIP6 simulations for (c) Arctic September sea ice and (d) global land precipitation. Projections for SSP1-1.9 and SSP1-2.6 show that reduced greenhouse gas emissions lead
to a stabilization of global surface temperature, Arctic sea ice area and global land precipitation over the 21st century. Projections for SSP1-2.6 show that emissions reductions
have the potential to substantially reduce the increase in ocean heat content and thermosteric sea level rise over the 21st century but that some increase is unavoidable. The
brackets in the x axis in panel (a) indicate assessed 20-year-mean periods. {4.3, Figure 4.2, 9.3, 9.6, Figure 9.6}

Observational records show changes in a wide range of climate human influence is the main contributor to the observed increase
extremes that have been linked to human influence on the climate (decrease) in the likelihood and severity of hot (cold) extremes (Table
system (Table TS.2). In many cases, the frequency and intensity of TS.2). The frequency of extreme temperature and precipitation events
future changes in extremes can be directly linked to the magnitude in the current climate will change with warming, with warm extremes
of future projected warming. Changes in extremes have been becoming more frequent (virtually certain), cold extremes becoming
widespread over land since the 1950s, including a virtually certain less frequent (extremely likely) and precipitation extremes becoming
global increase in extreme air temperatures and a likely intensification more frequent in most locations (very likely). {9.6.4, 11.2, 11.3, 11.4,
in global-scale extreme precipitation. It is extremely likely that 11.6, 11.7, 11.8, 11.9, Box 9.2}

66
 Technical Summary

Table TS.2 | Summary table on observed changes in extremes, their attribution since 1950 (except where stated otherwise), and projected changes at
+1.5°C, +2°C and +4°C of global warming, on global and continental scales. An increase in warm/hot extremes refers to warmer and/or more frequent hot days and
nights and warm spells/heatwaves, over most land areas. A decrease in cold extremes refers to warmer and/or fewer cold days and nights and cold spells/cold waves, over most
land areas. Drought events are relative to a predominant fraction of land area. For tropical cyclones, observed changes and attribution refer to Categories 3–5, while projected
changes refer to Categories 4–5. Tables 11.1 and 11.2 are more detailed versions of this table, containing, in particular, information on regional scales. In general, higher
warming levels also imply stronger projected changes for indicators where the confidence level does not depend on the warming level and the table does not explicitly quantify
the global sensitivity. See also Box TS.10. {9.6, Box 9.2, 11.3, 11.7}

Observed Attributed Projected at GWL (°C)

Change in Indicator
(since 1950) (since 1950) +1.5 +2 +4
Warm/hot extremes: Frequency
or intensity ↑ Main driver
↑ ↑ ↑
Cold extremes: Frequency or intensity ↓ Main driver
↓ ↓ ↓

↑ Main driver of the ↑ ↑

TS
Heavy precipitation events: Frequency, Over majority of land
observed intensification
↑
intensity and/or amount regions with good in most land regions in most land regions
of heavy precipitation in
observational coverage
land regions

↑ ↑ ↑
Agricultural and ecological droughts: ↑ in more regions in more regions in more regions
Intensity and/or frequency in some regions in some regions compared to observed compared to 1.5°C compared to 2°C of
changes of global warming global warming

Precipitation associated with

↑ ↑ ↑ ↑
tropical cyclones Rate +11% Rate +14% Rate +28%

Tropical cyclones: Proportion

↑ ↑ ↑ ↑
of intense cyclones +10% +13% +20%

Compound events: Co-occurrent ↑ ↑

heatwaves and droughts (Frequency) (Frequency) (Frequency and intensity increases with warming)

Marine heatwaves: Intensity & frequency ↑ ↑

(since 1900) (since 2006) Strongest in tropical and Arctic Ocean

Extreme sea levels: Frequency ↑ ↑

(since 1960) (Scenario-based assessment for 21st century)

medium confidence likely/high confidence very likely extremely likely virtually certain

TS.2.2 Changes in the Drivers of the Climate System The total anthropogenic effective radiative forcing (ERF)
in 2019, relative to 1750, was 2.72 [1.96 to 3.48] W m–2
Since 1750, changes in the drivers of the climate system (medium confidence) and has likely been growing at an
are dominated by the warming influence of increases in increasing rate since the 1970s. {2.2, 6.4, 7.2, 7.3}
atmospheric GHG concentrations and a cooling influence
from aerosols, both resulting from human activities. In Solar activity since 1900 was high but not exceptional compared to
comparison there has been negligible long-term influence the past 9000 years (high confidence). The average magnitude and
from solar activity and volcanoes. Concentrations of CO2, variability of volcanic aerosols since 1900 has not been unusual
methane (CH4), and nitrous oxide (N2O) have increased to compared to at least the past 2500 years (medium confidence).
levels unprecedented in at least 800,000 years, and there However, sporadic strong volcanic eruptions can lead to temporary
is high confidence that current CO2 concentrations have not drops in global surface temperature lasting 2–5 years. {2.2.1, 2.2.2,
been experienced for at least 2 million years. Global mean 2.2.8, Cross-Chapter Box 4.1}
concentrations of anthropogenic aerosols peaked in the
late 20th century and have slowly declined since in northern Atmospheric CO2 concentrations have changed substantially
mid-latitudes, although they continue to increase in South over millions of years (Figure TS.1). Current levels of atmospheric
Asia and East Africa (high confidence). CO2 have not been experienced for at least 2 million years (high

67
 Technical Summary

(a) Last time CO2 levels were as high as present was at least 2 million years ago
δ11B-foraminifera
δ13C-alkenone
450
Ant. ice core
409.9
CO2 (ppm) 350

250

150
3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0
Age (millions of years ago)

TS (c) Since 1960–1980 several high-accuracy global networks measure

(b) Information from multiple ice cores depicts a strong surface concentrations of CO2, CH4, and N2O. Current concentrations
increase of CO2, CH4, and N2O since the 19th century are higher than measured in ice cores during the last 800,000 years
1960–2019 1960–2019
420
409.9 420
400 WAIS Divide
400
409.9
380 Law Dome
CO2 (ppm)

380 NOAA

CO2 (ppm)
360 EDML
360 CSIRO
340 SIO
340
320
300 1866.3 320
1800 300 1866.3
280 1800
1600 280
1600
CH4 (ppb)

1400

CH4 (ppb)
GISP2 1400
1200
WAIS Divide 1200
1000
Law Dome NOAA 1000
800 CSIRO
AGAGE 800
600 UCI
340 600
332.1 340
320 332.1
320
N2O (ppb)

GISP2
N2O (ppb)

300 NEEM NOAA

H15 300 CSIRO
Law Dome Styx EURO AGAGE
280
280
260
260
0 400 800 1200 1600 2000 1960 1970 1980 1990 2000 2010 2020
Year (CE)
(d) The increase in effective radiative forcing (ERF) since the late 19th century is driven predominantly by
warming GHGs and cooling aerosol. ERF is changing at a faster rate since the 1970s
Carbon dioxide (CO2) Halogenated gases Volcanic
2 Methane (CH4) Tropospheric Aerosol Solar
Nitrous oxide (N2O) Other anthropogenic Total
Ozone (O3)

0
W m−2

−2
(W m−2 per decade−1)

0.5
0.4
Rate of change anthropogenic ERF 0.3
−4 0.2
0.1
0.0

1750 1800 1850 1900 1950 2000

Figure TS.9 | Changes in well-mixed greenhouse gas (WMGHG) concentrations and effective radiative forcing (EFR). The intent of this figure is to show that the
changes of the main drivers of climate system over the industrial period are exceptional in a long-term context. (a) Changes in carbon dioxide (CO2) from proxy records over the
past 3.5 million years. (b) Changes in all three WMGHGs from ice core records over the Common Era. (c) Directly observed WMGHG changes since the mid-20th century. (d)
Evolution of ERF and components since 1750. Further details on data sources and processing are available in the associated FAIR data table. {2.2, Figures 2.3, 2.4 and 2.10}

68
 Technical Summary

confidence, Figure TS.9a). Over 1750–2019, CO2 increased by 131.6 the late 20th century, with low confidence in the magnitude of
± 2.9 ppm (47.3%). The centennial rate of change of CO2 since 1850 post-2014 changes due to conflicting evidence (Section TS.3.1).
has no precedent in at least the past 800,000 years (Figure TS.9), {2.2.6, 6.2.1, 6.3.5, 6.4.1, 7.3.3}
and the fastest rates of change over the last 56 million years were
at least a factor of four lower (low confidence) than over 1900– There is high confidence that tropospheric ozone has been increasing
2019. Several networks of high-accuracy surface observations from 1750 in response to anthropogenic changes in ozone precursor
show that concentrations of CO2 have exceeded 400 ppm, reaching emissions (nitrogen oxides, carbon monoxide, non-methane volatile
409.9 (± 0.3) ppm in 2019 (Figure TS.9c). The ERF from CO2 in 2019 organic compounds, and methane), but with medium confidence in
(relative to 1750) was 2.16 Wm–2. {2.2.3, 5.1.2, 5.2.1, 7.3} the magnitude of this change, due to limited observational evidence
and knowledge gaps. Since the mid-20th century, tropospheric
By 2019, concentrations of CH4 reached 1866.3 (± 3.3) ppb ozone surface concentrations have increased by 30–70% across the
(Figure TS.9c). The increase since 1750 of 1137 ± 10 ppb (157.8%) far Northern Hemisphere (medium confidence); since the mid-1990s,
exceeds the range over multiple glacial–interglacial transitions of the free tropospheric ozone has increased by 2–7% per decade in most
past 800,000 years (high confidence). In the 1990s, CH4 concentrations northern mid-latitude regions and 2–12% per decade in sampled
plateaued, but started to increase again around 2007 at an average tropical regions. Future changes in surface ozone concentrations will TS
rate of 7.6 ± 2.7 ppb yr –1 (2010–2019; high confidence). There is high be primarily driven by changes in precursor emissions rather than
confidence that this recent growth is largely driven by emissions climate change (high confidence). Stratospheric ozone has declined
from fossil fuel exploitation, livestock, and waste, with ENSO driving between 60°S–60°N by 2.2% from 1964–1980 to 2014–2017
multi-annual variability of wetland and biomass burning emissions. (high confidence), with the largest declines during 1980–1995.
In 2019, ERF from CH4 was 0.54 Wm–2. {2.2.3, 5.2.2, 7.3} The strongest loss of stratospheric ozone continues to occur in
austral spring over Antarctica (ozone hole), with emergent signs of
Since 1750, N2O increased by 62.0 ± 6.0 ppb, reaching a level of recovery after 2000. The 1750–2019 ERF for total (stratospheric and
332.1 (± 0.4) ppb in 2019. The increase since 1750 is of comparable tropospheric) ozone is 0.47 [0.24 to 0.71] W m−2, which is dominated
magnitude to glacial–interglacial fluctuations of the past by tropospheric ozone changes. {2.2.5, 6.3.2, 7.3.2, 7.3.5}
800,000 years (Figure TS.9c). N2O concentration trends since 1980
are largely driven by a 30% increase in emissions from the expansion The global mean abundance of hydroxyl (OH) radical, or ‘oxidizing
and intensification of global agriculture (high confidence). By 2019 capacity’, chemically regulates the lifetimes of many SLCFs, and
its ERF was 0.21 W m–2. {2.2.3, 5.2.3} therefore the radiative forcing of CH4, ozone, secondary aerosols
and many halogenated species. Model estimates suggest no
Halogenated gases consist of chlorofluorocarbons (CFCs), significant change in oxidizing capacity from 1850 to 1980 (low
hydrochlorofluorocarbons (HCFCs), hydrofluorocarbons (HFCs) and confidence). Increases of about 9% over 1980–2014 computed by
other gases, many of which can deplete stratospheric ozone and ESMs and carbon cycle models are not confirmed by observationally
warm the atmosphere. In response to controls on production and constrained inverse models, rendering an overall medium confidence
consumption mandated by the Montreal Protocol on Substances in stable OH or positive trends since the 1980s, and implying that OH
that Deplete the Ozone Layer and its amendments, the atmospheric is not the primary driver of recent observed growth in CH4. {6.3.6,
abundances of most CFCs have continued to decline since AR5. Cross-Chapter Box 5.2}
Abundances of HFCs, which are replacements for CFCs and HCFCs,
are increasing (high confidence), though increases of the major HCFCs Land use and land-cover change exert biophysical and biogeochemical
have slowed in recent years. The ERF from halogenated components effects. There is medium confidence that the biophysical effects
in 2019 was 0.41 Wm–2. {2.2.4, 6.3.4, 7.3.2} of land-use change since 1750, most notably the increase in
global albedo, have had an overall cooling on climate, whereas
Tropospheric aerosols mainly act to cool the climate system, biogeochemical effects (i.e., changes in GHG and volatile organic
directly by reflecting solar radiation, and indirectly by enhancing compound emissions or sinks) led to net warming. Overall land-use
cloud reflectance. Ice cores show increases in aerosols across the and land-cover ERF is estimated at –0.2 [–0.3 to –0.1] W m−2. {2.2.7,
Northern Hemisphere mid-latitudes since 1700 and reductions 7.3.4, SRCCL Section 2.5}
since the late 20th century (high confidence). Aerosol optical depth
(AOD), derived from satellite- and ground-based radiometers, The total anthropogenic ERF in 2019 relative to 1750 was 2.72 [1.96
has decreased since 2000 over the mid-latitude continents of to 3.48] W m−2 (Figure TS.9), dominated by GHGs (positive ERF) and
both hemispheres, but increased over South Asia and East Africa partially offset by aerosols (negative ERF). The rate of change of ERF
(high confidence). Trends in AOD are more pronounced from sub- likely has increased since the 1970s, mainly due to growing CO2
micrometre aerosols for which the anthropogenic contribution is concentrations and less negative aerosol ERF (Section TS.3.1). {2.2.8,
particularly large. Global carbonaceous aerosol budgets and trends 7.3}
remain poorly characterized due to limited observations, but black
carbon (BC), a warming aerosol component, is declining in several
regions of the Northern Hemisphere (low confidence). Total aerosol
ERF in 2019, relative to 1750, is −1.1 [−1.7 to −0.4] W m−2 (medium
confidence) and more likely than not became less negative since

69
 Technical Summary

TS.2.3 Upper Air Temperatures and Atmospheric track in the Southern Hemisphere by 2100 under the high
Circulation CO2 emissions scenarios. It is likely that the proportion of
intense tropical cyclones has increased over the last four
The effects of human-induced climate change have decades and that this cannot be explained entirely by
been clearly identified in observations of atmospheric natural variability. There is low confidence in observed
temperature and some aspects of atmospheric circulation, recent changes in the total number of extratropical cyclones
and these effects are likely to intensify in the future. over both hemispheres. The proportion of tropical cyclones
Tropospheric warming and stratospheric cooling are that are intense is expected to increase (high confidence),
virtually certain to continue with continued net emissions but the total global number of tropical cyclones is expected
of greenhouse gases. Several aspects of the atmospheric to decrease or remain unchanged (medium confidence).
circulation have likely changed since the mid-20th century, {2.3, 3.3, 4.3, 4.4, 4.5, 8.3, 8.4, 11.7}
and human influence has likely contributed to the observed
poleward expansion of the Southern Hemisphere Hadley Cell The troposphere has warmed since at least the 1950s, and it is virtually
and very likely contributed to the observed poleward shift certain that the stratosphere has cooled. It is very likely that human-
TS of the Southern Hemisphere extratropical jet in summer. It induced increases in GHGs were the main driver of tropospheric
is likely that the mid-latitude jet will shift poleward and warming since 1979. It is extremely likely that anthropogenic
strengthen, accompanied by a strengthening of the storm forcing, both from increases in GHG concentrations and depletion of

Observed trends Projected long-term change (2081–2100)

GNSS-RO 2002–2019 Multi-model projections (SSP1-2.6) Multi-model projections (SSP3-7.0)
36 32

Pressure (hPa)
Pressure (hPa)

Pressure (hPa)

Colour High model agreement

–0.5–0.4–0.3–0.2–0.1 0 0.1 0.2 0.3 0.4 0.5 –5 –4 –3 –2 –1 0 1 2 3 4 5 Low model agreement
°C per decade °C

ERA5 1979–2019 (DJF) Multi-model projections (DJF, SSP1-2.6) Multi-model projections (DJF, SSP3-7.0)
34 31
Pressure (hPa)

Pressure (hPa)
Pressure (hPa)

Non-signficant trend Contour from –50 to 40 by 10

Contour from –50 to 40 by 10

–1.5 –1 -0.5 0 0.5 1 1.5 –6 –5 –4 –3 –2 –1 0 1 2 3 4 5 6 Colour High model agreement

m/s per decade m/s Low model agreement

Figure TS.10 | Observed and projected upper air temperature and circulation changes. The intent of this figure is to visualize upper air temperature and circulation
changes and the similarity between observed and projected changes. Upper panels: (Left) Zonal cross-section of temperature trends for 2002–2019 in the upper troposphere
region for the ROM SAF radio-occultation dataset. (Middle) Change in the annual and zonal mean atmospheric temperature (°C) in 2081–2100 in SSP1-2.6 relative to
1995–2014 for 36 Coupled Model Intercomparison Project Phase 6 (CMIP6) models. (Right) the same in SSP3-7.0 for 32 models. Lower panels: (Left) Long-term mean (thin
black colour) and linear trend (colour) of zonal mean December–January–February (DJF) zonal winds for ERA5. (Middle) multi-model mean change in annual and zonal mean
wind (m s–1) in 2081–2100 in SSP1-2.6 relative to 1995–2014 based on 34 CMIP6 models. The 1995–2014 climatology is shown in contours with spacing of 10 m s–1. (Right)
the same for SSP3-7.0 for 31 models. {2.3.1; Figures 2.12 and 2.18; 4.5.1; Figure 4.2.6}

70
 Technical Summary

stratospheric ozone due to ozone-depleting substances, was the main the Southern Hemisphere mid-latitude jet is likely to shift poleward
driver of upper stratospheric cooling since 1979. It is very likely that and strengthen under the SSP5-8.5 scenario relative to 1995–2014,
global mean stratospheric cooling will be larger for scenarios with accompanied by an increase in the SAM (Section TS.4.2.2). It is
higher atmospheric CO2 concentrations. In the tropics, since at least likely that wind speeds associated with extratropical cyclones will
2001 (when new techniques permit more robust quantification), the strengthen in the Southern Hemisphere storm track for SSP5-8.5.
upper troposphere has warmed faster than the near-surface (medium There is low confidence in the potential role of Arctic warming and
confidence) (Figure TS.10). There is medium confidence that most sea ice loss on historical or projected mid-latitude atmospheric
CMIP5 and CMIP6 models overestimate the observed warming in variability. {2.3.1, 3.3.3, 3.7.2, 4.3.3, 4.4.3, 4.5.1, 4.5.3, 8.2.2, 8.3.2,
the upper tropical troposphere over the period 1979–2014, in part Cross-Chapter Box 10.1}
because they overestimate tropical SST warming. It is likely that
future tropical upper tropospheric warming will be larger than at the It is likely that the proportion of major (Category 3–5) tropical
tropical surface. {2.3.1, 3.3.1, 4.5.1} cyclones (TCs) and the frequency of rapid TC intensification events
have increased over the past four decades. The average location
The Hadley Circulation has likely widened since at least the 1980s, of peak TC wind-intensity has very likely migrated poleward in
predominantly in the Northern Hemisphere, although there is only the western North Pacific Ocean since the 1940s, and TC forward TS
medium confidence in the extent of the changes. This has been translation speed has likely slowed over the contiguous USA since
accompanied by a strengthening of the Hadley Circulation in the 1900. It is likely that the poleward migration of TCs in the western
Northern Hemisphere (medium confidence). It is likely that human North Pacific and the global increase in TC intensity rates cannot be
influence has contributed to the poleward expansion of the zonal explained entirely by natural variability. There is high confidence
mean Hadley cell in the Southern Hemisphere since the 1980s, which that average peak TC wind speeds and the proportion of Category
is projected to further expand with global warming (high confidence). 4–5 TCs will increase with warming and that peak winds of the
There is medium confidence that the observed poleward expansion most intense TCs will increase. There is medium confidence that the
in the Northern Hemisphere is within the range of internal variability. average location where TCs reach their maximum wind-intensity will
{2.3.1, 3.3.3, 8.4.3} migrate poleward in the western North Pacific Ocean, while the total
global frequency of TC formation will decrease or remain unchanged
Since the 1970s, near-surface average winds have likely weakened with increasing global warming. {11.7.1}
over land. Over the ocean, near-surface average winds likely
strengthened over 1980–2000, but divergent estimates lead to There is low confidence in observed recent changes in the total
low confidence thereafter. Extratropical storm tracks have likely number of extratropical cyclones over both hemispheres. There is also
shifted poleward since the 1980s. There is low confidence in low confidence in past-century trends in the number and intensity of
projected poleward shifts of the Northern Hemisphere mid-latitude the strongest extratropical cyclones over the Northern Hemisphere
jet and storm tracks due to large internal variability and structural due to the large interannual-to-decadal variability and temporal and
uncertainty in model simulations. There is medium confidence in spatial heterogeneities in the volume and type of assimilated data
a projected decrease in the frequency of atmospheric blocking over in atmospheric reanalyses, particularly before the satellite era. Over
Greenland and the North Pacific in boreal winter in 2081–2100 under the Southern Hemisphere, it is likely that the number of extratropical
the SSP3-7.0 and SSP5-8.5 scenarios. There is high confidence that cyclones with low central pressures (<980 hPa) has increased since
Southern Hemisphere storm tracks and associated precipitation have 1979. The frequency of intense extratropical cyclones is projected to
migrated polewards over recent decades, especially in the austral decrease (medium confidence). Projected changes in the intensity
summer and autumn, associated with a trend towards more positive depend on the resolution of climate models (medium confidence).
phases of the Southern Annular Mode (SAM) (Section TS.4.2.2) and There is medium confidence that wind speeds associated with
the strengthening and southward shift of the Southern Hemisphere extratropical cyclones will change following changes in the storm
extratropical jet in austral summer. In the long term (2081–2100), tracks. {2.3.1, 3.3.3, 4.5.1, 4.5.3, 8.3.2, 8.4.2, 11.7.2}

71
 Technical Summary

Box TS.3 | Low-likelihood, High-warming Storylines

Future global warming exceeding the assessed very likely range cannot be ruled out and is potentially associated
with the highest risks for society and ecosystems. Such low-likelihood, high-warming storylines tend to exhibit
substantially greater changes in the intensity of regional drying and wetting than the multi-model mean. Even at
levels of warming within the very likely range, global and regional low-likelihood outcomes might occur, such as large
precipitation changes, additional sea level rise associated with collapsing ice sheets (see Box TS.4), or abrupt ocean
circulation changes. While there is medium confidence that the Atlantic Meridional Overturning Circulation (AMOC)
will not experience an abrupt collapse before 2100, if it were to occur, it would very likely cause abrupt shifts in
regional weather patterns and water cycle. The probability of these low-likelihood outcomes increases with higher
global warming levels. If the real-world climate sensitivity lies at the high end of the assessed range, then global
and regional changes substantially outside the very likely range projections occur for a given emissions scenario.
With increasing global warming, some very rare extremes and some compound events (multivariate or concurrent
TS extremes) with low likelihood in past and current climate will become more frequent, and there is a higher chance
that events unprecedented in the observational record occur (high confidence). Finally, low-likelihood, high-impact
outcomes may also arise from a series of very large volcanic eruptions that could substantially alter the 21st century
climate trajectory compared to SSP-based Earth system model (ESM) projections. {Cross-Chapter Box 4.1, 4.3, 4.4, 4.8,
7.3, 7.4, 7.5, 8.6, 9.2, 9.6, Box 9.4, Box 11.2, Cross-Chapter Box 12.1}

Previous IPCC reports largely focused their assessment on the projected very likely range of future surface warming and associated
climate change. However, a comprehensive risk assessment also requires considering the potentially larger changes in the physical
climate system that are unlikely or very unlikely but possible and potentially associated with the highest risks for society and
ecosystems (Figure TS.6). Since AR5, the development of physical climate storylines of high warming has emerged as a useful approach
for exploring the future risk space that lies outside of the IPCC very likely range projections. {4.8}

Uncertainty in the true values of equilibrium climate sensitivity (ECS) and transient climate response (TCR) dominate uncertainty
in projections of future warming under moderate to strong emissions scenarios (Section TS.3.2). A real-world ECS higher than the
assessed very likely range (2°C–5°C) would require a strong historical aerosol cooling and/or a trend towards stronger warming from
positive feedbacks linked to changes in SST patterns (pattern effects), combined with a strong positive cloud feedback and substantial
biases in paleoclimate reconstructions – each of which is assessed as either unlikely or very unlikely, but not ruled out. Since CMIP6
contains several ESMs that exceed the upper bound of the assessed very likely range in future surface warming, these models can
be used to develop low-likelihood, high warming storylines to explore risks and vulnerabilities, even in the absence of a quantitative
assessment of likelihood. {4.3.4, 4.8, 7.3.2, 7.4.4, 7.5.2, 7.5.5, 7.5.7}

CMIP6 models with surface warming outside, or close to, the upper bound of the very likely range exhibit patterns of large widespread
temperature and precipitation changes that differ substantially from the multi-model mean in all scenarios. For SSP5-8.5, the
high-warming models exhibit widespread warming of more than 6°C over most extratropical land regions and parts of the Amazon.
In the Arctic, annual mean temperatures increase by more than 10°C relative to present-day, corresponding to about 30% more than
the best estimate of warming. Even for SSP1-2.6, high-warming models show on average 2°C–3°C warming relative to present-day
conditions over much of Eurasia and North America (about 40% more than the best estimate of warming) and more than 4°C warming
relative to the present over the Arctic in 2081–2100 (Box TS.3, Figure 1). Such a high global warming storyline would imply that the
remaining carbon budget consistent with a 2°C warming is smaller than the assessed very likely range. Put another way, even if
a carbon budget that likely limits warming to 2°C is met, a low-likelihood, high-warming storyline would result in warming of 2.5°C
or more. {4.8}

CMIP6 models with global warming close to the upper bound of the assessed very likely warming range tend to exhibit greater
changes in the intensity of regional drying and wetting than the multi-model mean. Furthermore, these model projections show
a larger area of drying and tend to show a larger fraction of strong precipitation increases than the multi-model mean. However,
regional precipitation changes arise from both thermodynamic and dynamic processes so that the most pronounced global warming
levels are not necessarily associated with the strongest precipitation response. Abrupt human-caused changes to the water cycle
cannot be ruled out. Positive land surface feedbacks, involving vegetation and dust, can contribute to abrupt changes in aridity,
but there is only low confidence that such changes will occur during the 21st century. Continued Amazon deforestation, combined
with a warming climate, raises the probability that this ecosystem will cross a tipping point into a dry state during the 21st century
(low confidence). (See also Box TS.9). {4.8, 8.6.2}

72
 Technical Summary

Box TS.3 (continued)

While there is medium confidence that the projected decline in the AMOC (Section TS.2.4) will not involve an abrupt collapse
before 2100, such a collapse might be triggered by an unexpected meltwater influx from the Greenland Ice Sheet. If an AMOC collapse
were to occur, it would very likely cause abrupt shifts in the regional weather patterns and water cycle, such as a southward shift in the
tropical rain belt, and could result in weakening of the African and Asian monsoons, strengthening of Southern Hemisphere monsoons,
and drying in Europe. (See also Boxes TS.9 and TS.13). {4.7.2, 8.6.1, 9.2.3}

Very rare extremes and compound or concurrent events, such as the 2018 concurrent heatwaves across the Northern Hemisphere, are
often associated with large impacts. The changing climate state is already altering the likelihood of extreme events, such as decadal
droughts and extreme sea levels, and will continue to do so under future warming. Compound events and concurrent extremes
contribute to increasing probability of low-likelihood, high-impact outcomes and will become more frequent with increasing global
warming (high confidence). Higher warming levels increase the likelihood of events unprecedented in the observational record. {9.6.4,
Box 11.2} TS

Finally, low likelihood storylines need not necessarily relate solely to the human-induced changes in climate. A low-likelihood, high-
impact outcome, consistent with historical precedent in the past 2500 years, would be to see several large volcanic eruptions that
could greatly alter the 21st century climate trajectory compared to SSP-based Earth system model projections. {Cross-Chapter Box 4.1}

SSP1-2.6 (2081–2100)
(a) Best estimate (scaled) (b) High-warming models (c) Very-high-warming models

–4 –3 –2.5 –2 –1.5 –1 0 1 1.5 2 2.5 3 4

°C

SSP5-8.5 (2081–2100)
(d) Best estimate (scaled) (e) High-warming models (f) Very-high-warming models

–8 –6 –5 –4 –3 –2 0 2 3 4 5 6 8
°C

Box TS.3, Figure 1 | High-warming storylines. The intent of this figure is to illustrate high warming storylines compared to the CMIP6 multi-model-mean.
(a) Coupled Model Intercomparison Project Phase 6 (CMIP6) multi-model mean linearly scaled to the assessed best global surface temperature estimate for SSP1-
2.6 in 2081–2100 relative to 1995–2014, (b) mean across five high-warming models with global surface temperature changes nearest to the upper bound of the
assessed very likely range, and (c) mean across five very high-warming models with global surface temperature changes higher than the assessed very likely. (d–f)
Same as (a–c) but for SSP5-8.5. Note the different colour bars in (a–c) and (d–f). {4.7, Figure 4.41}

73
 Technical Summary

TS.2.4 The Ocean Global mean SST has increased since the beginning of the 20th
century by 0.88 [0.68 to 1.01] °C, and it is virtually certain it will
Observations, models and paleo-evidence indicate that continue to increase throughout the 21st century, with increasing
recently observed changes in the ocean are unprecedented hazards to marine ecosystems (medium confidence). Marine
for centuries to millennia (high confidence). Over the past heatwaves have become more frequent over the 20th century (high
four to six decades, it is virtually certain that the global confidence), approximately doubling in frequency (high confidence)
ocean has warmed, with human influence extremely and becoming more intense and longer since the 1980s (medium
likely the main driver since the 1970s, making climate confidence). Most of the marine heatwaves over 2006–2015 have
change irreversible over centuries to millennia (medium been attributed to anthropogenic warming (very likely). Marine
confidence). It is virtually certain that upper ocean salinity heatwaves will continue to increase in frequency, with a likely global
contrasts have increased since the 1950s and extremely increase of 2–9 times in 2081–2100 compared to 1995–2014 under
likely that human influence has contributed. It is virtually SSP1-2.6, and 3–15 times under SSP5-8.5 (Figure TS.11a), with
certain that upper ocean stratification has increased since the largest changes in the tropical and Arctic ocean. {2.3.1, Cross-
1970 and that sea water pH has declined globally over the Chapter Box 2.3, 9.2.1, Box 9.2, 12.4.8}
TS last 40 years, with human influence being the main driver
of the observed surface open ocean acidification (virtually Observed upper-ocean stratification (0–200 m) has increased
certain). A long-term increase in surface open ocean pH globally since at least 1970 (virtually certain). Based on recent refined
occurred over the past 50 million years (high confidence), analyses of the available observations, there is high confidence
and surface ocean pH as low as recent times is uncommon that it increased by 4.9 ± 1.5% from 1970–2018, which is about
in the last 2 million years (medium confidence). There is twice as much as assessed in SROCC, and will continue to increase
high confidence that marine heatwaves have become more throughout the 21st century at a rate depending on the emissions
frequent in the 20th century, and most of those since 2006 scenario (virtually certain). {2.3.3, 9.2.1}
have been attributed to anthropogenic warming (very
likely). There is high confidence that oxygen levels have It is virtually certain that since 1950 near-surface high-salinity
dropped in many regions since the mid 20th century and regions have become more saline, while low-salinity regions have
that the geographic range of many marine organisms has become fresher, with medium confidence that this is linked to an
changed over the last two decades. intensification of the hydrological cycle (Box TS.6). It is extremely
likely that human influence has contributed to this salinity change
The amount of ocean warming observed since 1971 and that the large-scale pattern will grow in amplitude over the 21st
will likely at least double by 2100 under a low warming century (medium confidence). {2.3.3, 3.5.2, 9.2.2, 12.4.8}
scenario (SSP1-2.6) and will increase by 4–8 times under
a high warming scenario (SSP5-8.5). Stratification (virtually The AMOC was relatively stable during the past 8000 years (medium
certain), acidification (virtually certain), deoxygenation confidence). There is low confidence in the quantification of AMOC
(high confidence) and marine heatwave frequency (high changes in the 20th century because of low agreement in quantitative
confidence) will continue to increase in the 21st century. reconstructed and simulated trends, missing key processes in both
While there is low confidence in 20th century AMOC change, models and measurements used for formulating proxies, and new
it is very likely that AMOC will decline over the 21st century model evaluations. Direct observational records since the mid-2000s
(Figure TS.11). {2.3, 3.5, 3.6, 4.3.2, 5.3, 7.2, 9.2, Box 9.2, 12.4} are too short to determine the relative contributions of internal
variability, natural forcing and anthropogenic forcing to AMOC
It is virtually certain that the global ocean has warmed since at least change (high confidence). An AMOC decline over the 21st century
1971, representing about 90% of the increase in the global energy is very likely for all SSP scenarios (Figure TS.11b); a possible abrupt
inventory (Section TS.3.1). The ocean is currently warming faster than decline is assessed further in Box TS.3. {2.3.3, 3.5.4, 4.3.2, 8.6.1,
at any other time since at least the last deglacial transition (medium 9.2.3, Cross-Chapter Box 12.3}
confidence), with warming extending to depths well below 2000 m
(very high confidence). It is extremely likely that human influence There is high confidence that many ocean currents will change in
was the main driver of this recent ocean warming. Ocean warming the 21st century in response to changes in wind stress. There is low
will continue over the 21st century (virtually certain), and will likely confidence in 21st century change of Southern Ocean circulation,
continue until at least to 2300 even for low CO2 emissions scenarios. despite high confidence that it is sensitive to changes in wind
Ocean warming is irreversible over centuries to millennia (medium patterns and increased ice-shelf melt. Western boundary currents
confidence), but the magnitude of warming is scenario-dependent from and subtropical gyres have shifted poleward since 1993 (medium
about the mid-21st century (medium confidence). The warming will confidence). Subtropical gyres, the East Australian Current Extension,
not be globally uniform, with heat primarily stored in Southern Ocean the Agulhas Current, and the Brazil Current are projected to intensify
water-masses and weaker warming in the subpolar North Atlantic (high in the 21st century in response to changes in wind stress, while the
confidence). Limitations in the understanding of feedback mechanisms Gulf Stream and the Indonesian Throughflow are projected to weaken
limit our confidence in future ocean warming close to Antarctica and (medium confidence). All of the four main eastern boundary upwelling
how this will affect sea ice and ice shelves. {2.3.3, 3.5.1, 4.7.2, 7.2.2, systems are projected to weaken at low latitudes and intensify at high
9.2.2, 9.2.3, 9.2.4, 9.3.2, 9.6.1, Cross-Chapter Box 9.1} latitudes in the 21st century (high confidence). {2.3.3, 9.2.3}

74
 Technical Summary

Recent and Future change in the ocean

Marine heatwaves, Atlantic Meridional Overturning Circulation (AMOC), Dissolved oxygen, and pH

(a) Marine heatwaves (c) Dissolved oxygen (100–600 m)

15 5
Future (CMIP6) Past (simulated) Past (observed)
0
multiplication factor

10

% change
-5
5 Past (observed) Future (CMIP6)
-10
Past (simulated)
-15
Change Change
in 2100 in 2100

(b) Atlantic Meridional Overturning Circulation (AMOC) (d) Ocean acidification

5 8.2
Past (observed)
Past (simulated) Past (simulated) Past (observed)
0 8 TS
low acidity

pH
Sv

-5
7.8
high acidity Future (CMIP6)
-10 Future (CMIP6)
7.6
Change Change
1850 1900 1950 2000 2050 2100 in 2100 1850 1900 1950 2000 2050 2100 in 2100

Past (simulated) Future (SSP1-2.6) mean; 5–95% range

Past (observed) Future (SSP2-4.5) mean
Future (SSP3-7.0) mean; 5–95% range
Future (SSP5-8.5) mean; 5–95% range

Recent and future change in ice sheets

Greenland and Antarctic ice sheets
(e) Greenland ice sheet (f) Antarctic ice sheet
4 -0.1 4 Historical Future (RCP/SSP) -0.1
Historical (observation-based)
Box (observation-based) Future (RCP/SSP) 0 0
0 0
Observations ISMIP6 Observations ISMIP6
(2010–2017) (2093–2100) -4 0.1
104 Gt

(1978–2017) (2061–2100)

m
m

104 Gt

-4
1 0.1 -8 1 0.2
Elevation change

Elevation change

0.5 0.5
0.3
(m/yr)

(m/yr)

0 0
-0.5 0.2 -12
-0.5
-8 -1 -1
0.4
ISMIP6 Emulator -16
1980 2000 2020 2040 2060 2080 2100 Change 1980 2000 2020 2040 2060 2080 2100
in 2100
ISMIP6 LARMIP
Observation-based: Emulator
Emulator median (SSP1-2.6); 17-83% & 5–95% ranges
Bamber Change
Emulator median (SSP5-8.5); 17-83% & 5–95% ranges in 2100
Mouginot/Rignot ISMIP6 models (SSP1-2.6/RCP2.6)
IMBIE ISMIP6 models (SSP5-8.5/RCP8.5)

Figure TS.11 | Past and future ocean and ice-sheet changes. The intent of this figure is to show that observed and projected time series of many ocean and cryosphere
indicators are consistent. Observed and simulated historical changes and projected future changes under varying greenhouse gas emissions scenarios. Simulated and projected
ocean changes are shown as Coupled Model Intercomparison Project Phase 6 (CMIP6) ensemble mean, and 5–95% range (shading) is provided for scenarios SSP1-2.6 and
SSP3-7.0 (except in panel a where the range is provided for scenario SSP1-2.6 and SSP5-8.5). Mean and 5–95% range in 2100 are shown as vertical bars on the right-hand side
of each panel. (a) Change in multiplication factor in surface ocean marine heatwave days relative to 1995–2014 (defined as days exceeding the 99th percentile in sea surface
temperature (SST) from 1995–2014 distribution). Assessed observational change span 1982–2019 from AVHRR satellite SST. (b) Atlantic Meridional Overturning Circulation
(AMOC) transport relative to 1995–2014 (defined as maximum transport at 26°N). Assessed observational change spans 2004–2018 from the RAPID array smoothed with
a 12-month running mean (shading around the mean shows the 12-month running standard deviation around the mean). (c) Global mean percent change in ocean oxygen
(100–600 m depth), relative to 1995–2014. Assessed observational trends and very likely range are from the SROCC assessment, and span 1970–2010 centred on 2005. (d)
Global mean surface pH. Assessed observational change spans 1985–2019, from the CMEMS SOCAT-based reconstruction (shading around the global mean shows the 90%
confidence interval). (e), (f): Ice sheet mass changes. Projected ice-sheet changes are shown as median, 5–95% range (light shading), and 17–83% range (dark shading) of
cumulative mass loss and sea level equivalent from ISMIP6 emulation under SSP1-2.6 and SSP5-8.5 (shading and bold line), with individual emulated projections as thin lines.
Median (dot), 17–83% range (thick vertical bar), and 5–95% range (thin vertical bar) in 2100 are shown as vertical bars on the right-hand side of each panel, from ISMIP6,
ISMIP6 emulation, and LARMIP-2. Observation-based estimates: For Greenland (e), for 1972–2018 (Mouginot), for 1992–2016 (Bamber), for 1992–2020 (IMBIE) and total
estimated mass loss range for 1840–1972 (Box). For Antarctica (f), estimates based on satellite data combined with simulated surface mass balance and glacial isostatic
adjustment for 1992–2020 (IMBIE), 1992–2016 (Bamber), and 1979–2017 (Rignot). Left inset maps: mean Greenland elevation changes 2010–2017 derived from CryoSat-2
radar altimetry (e) and mean Antarctica elevation changes 1978–2017 derived from restored analogue radar records (f). Right inset maps: ISMIP6 model mean (2093–2100)
projected changes under the MIROC5 climate model for the RCP8.5 scenario. {2.3.3; 2.3.4; 3.5.4; 4.3.2; 5.3.2; 5.3.3; 5.6.3; 9.2.3; 9.4.1; 9.4.2; Box 9.2; Box 9.2, Figure 1;
Figures 9.10, 9.17 and 9.18}

75
 Technical Summary

It is virtually certain that surface pH has declined globally over the become practically sea ice-free in late summer under high
last 40 years and that the main driver is uptake of anthropogenic CO2 emissions scenarios by the end of the 21st century (high
CO2. Ocean acidification and associated reductions in the saturation confidence). It is virtually certain that further warming will
state of calcium carbonate – a constituent of skeletons or shells of lead to further reductions of Northern Hemisphere snow
a variety of marine organisms – is expected to increase in the 21st cover, and there is high confidence that this is also the case
century under all emissions scenarios (high confidence). A long-term for near-surface permafrost volume.
increase in surface open ocean pH occurred over the past 50 million
years (high confidence), and surface ocean pH as low as recent Glaciers will continue to lose mass at least for several
times is uncommon in the last 2 million years (medium confidence). decades even if global temperature is stabilized (very
There is very high confidence that present-day surface pH values high confidence), and mass loss over the 21st century
are unprecedented for at least 26,000 years and current rates of pH is virtually certain for the Greenland Ice Sheet and likely
change are unprecedented since at least that time. Over the past 2–3 for the Antarctic Ice Sheet. Deep uncertainty persists with
decades, a pH decline in the ocean interior has been observed in all respect to the possible evolution of the Antarctic Ice Sheet
ocean basins (high confidence) (Figure TS.11d). {2.3.3, 2.3.4, 3.6.2, within the 21st century and beyond, in particular due to the
TS 4.3.2, 5.3.2, 5.3.3, 5.6.3, 12.4.8} potential instability of the West Antarctic Ice Sheet. {2.3,
3.4, 4.3, 8.3, 9.3–9.6, Box 9.4, 12.4}
Open-ocean deoxygenation and expansion of oxygen minimum zones
have been observed in many areas of the global ocean since the Current Arctic sea ice coverage levels (both annual and late summer)
mid 20th century (high confidence), in part due to human influence are at their lowest since at least 1850 (high confidence), and for
(medium confidence). Deoxygenation is projected to continue to late summer for the past 1000 years (medium confidence). Since the
increase with ocean warming (high confidence) (Figure TS.11c). late 1970s, Arctic sea ice area and thickness have decreased in both
Higher climate sensitivity and reduced ocean ventilation in CMIP6 summer and winter, with sea ice becoming younger, thinner and more
compared to CMIP5 results in substantially greater projections of dynamic (very high confidence). It is very likely that anthropogenic
subsurface (100–600 m) oxygen decline than reported in SROCC for forcing, mainly due to greenhouse gas increases, was the main driver
the period 2080–2099. {2.3.3, 2.3.4, Cross-Chapter Box 2.4, 3.6.2, of this loss, although new evidence suggests that anthropogenic
5.3.3, 12.4.8} aerosol forcing has offset part of the greenhouse gas-induced losses
since the 1950s (medium confidence). The annual Arctic sea ice area
Over at least the last two decades, the geographic range of many minimum will likely fall below 1 million km2 at least once before
marine organisms has shifted towards the poles and towards greater 2050 under all assessed SSP scenarios. This practically sea ice-free
depths (high confidence), indicative of shifts towards cooler waters. state will become the norm for late summer by the end of the 21st
The range of a smaller subset of organisms has shifted equatorward century in high CO2 emissions scenarios (high confidence). Arctic
and to shallower depths (high confidence). Phenological metrics summer sea ice varies approximately linearly with global surface
associated with the life cycles of many organisms have also temperature, implying that there is no tipping point and observed/
changed over the last two decades or longer (high confidence). projected losses are potentially reversible (high confidence). {2.3.2,
Since the changes in the geographical range of organisms and their 3.4.1, 4.3.2, 9.3.1, 12.4.9}
phenological metrics have been observed to differ with species
and location, there is the possibility of disruption to major marine For Antarctic sea ice, there is no significant trend in satellite-observed
ecosystems. {2.3.4} sea ice area from 1979 to 2020 in both winter and summer, due to
regionally opposing trends and large internal variability. Due to
mismatches between model simulations and observations, combined
TS.2.5 The Cryosphere with a lack of understanding of reasons for substantial inter-model
spread, there is low confidence in model projections of future Antarctic
Over recent decades, widespread loss of snow and ice has sea ice changes, particularly at the regional level. {2.3.2, 3.4.1, 9.3.2}
been observed, and several elements of the cryosphere
are now in states unseen in centuries (high confidence). In permafrost regions, increases in ground temperatures in the upper
Human influence was very likely the main driver of 30 m over the past three to four decades have been widespread (high
observed reductions in Arctic sea ice since the late 1970s confidence). For each additional 1°C of warming (up to 4°C above the
(with late-summer sea ice loss likely unprecedented for at 1850–1900 level), the global volume of perennially frozen ground to
least 1000 years) and the widespread retreat of glaciers 3 m below the surface is projected to decrease by about 25% relative
(unprecedented in at least the last 2,000 years, medium to the present volume (medium confidence). However, these decreases
confidence). Furthermore, human influence very likely may be underestimated due to an incomplete representation of
contributed to the observed Northern Hemisphere spring relevant physical processes in ESMs (low confidence). Seasonal snow
snow cover decrease since 1950. cover is treated in Section TS.2.6. {2.3.2, 9.5.2, 12.4.9}

By contrast, Antarctic sea ice area experienced no significant There is very high confidence that, with few exceptions, glaciers
net change since 1979, and there is only low confidence have retreated since the second half of the 19th century; this
in its projected changes. The Arctic Ocean is projected to behaviour is unprecedented in at least the last 2000 years (medium

76
 Technical Summary

confidence). Mountain glaciers very likely contributed 67.2 [41.8 to of the Antarctic Ice Sheet, is assessed in Box TS.9. {2.3.2, 3.4.3, 9.4.1,
92.6] mm to the observed GMSL change between 1901 and 2018. 9.4.2, 9.6.3, Atlas.11.2}
This retreat has occurred at increased rates since the 1990s, with
human influence very likely being the main driver. Under RCP2.6 and It is likely that the Antarctic Ice Sheet has lost 2670 ± 530 Gt,
RCP8.5, respectively, glaciers are projected to lose 18% ± 13% and contributing 7.4 ± 1.5 mm to GMSL rise over 1992–2020. The total
36% ± 20% of their current mass over the 21st century (medium Antarctic ice mass losses were dominated by the West Antarctic Ice
confidence). {2.3.2, 3.4.3, 9.5.1, 9.6.1} Sheet, with combined West Antarctic and Peninsula annual loss rates
increasing since about 2000 (very high confidence). Furthermore, it
The Greenland Ice Sheet was smaller than at present during the is very likely that parts of the East Antarctic Ice Sheet have lost mass
Last Interglacial period (roughly 125,000 years ago) and the mid- since 1979. Since the 1970s, snowfall has likely increased over the
Holocene (roughly 6,000 years ago) (high confidence). After reaching western Antarctic Peninsula and eastern West Antarctica, with large
a recent maximum ice mass at some point between 1450 and 1850, spatial and interannual variability over the rest of Antarctica. Mass
the ice sheet retreated overall, with some decades likely close to losses from West Antarctic outlet glaciers, mainly induced by ice shelf
equilibrium (i.e., mass loss approximately equalling mass gained). It basal melt (high confidence), outpace mass gain from increased
is virtually certain that the Greenland Ice Sheet has lost mass since snow accumulation on the continent (very high confidence). TS
the 1990s, with human influence a contributing factor (medium However, there is only limited evidence, with medium agreement,
confidence). There is high confidence that annual mass changes have of anthropogenic forcing of the observed Antarctic mass loss since
been consistently negative since the early 2000s. Over the period 1992 (with low confidence in process attribution). Increasing mass
1992–2020, Greenland likely lost 4890 ± 460 Gt of ice, contributing loss from ice shelves and inland discharge will likely continue to
13.5 ± 1.3 mm to GMSL rise. There is high confidence that Greenland outpace increasing snowfall over the 21st century (Figure TS.11f).
ice mass losses are increasingly dominated by surface melting and Deep uncertainty persists with respect to the possible evolution of
runoff, with large interannual variability arising from changes in the Antarctic Ice Sheet along high-end mass-loss storylines within
surface mass balance. Projections of future Greenland ice-mass loss the 21st century and beyond, primarily related to the abrupt and
(Box TS.4, Table 1; Figure TS.11e) are dominated by increased surface widespread onset of marine ice sheet instability and marine ice cliff
melt under all emissions scenarios (high confidence). Potential instability. (See also Boxes TS.3 and TS.4). {2.3.2, 3.4.3, 9.4.2, 9.6.3,
irreversible long-term loss of the Greenland Ice Sheet, and of parts Box 9.4, Atlas.11.1}

Box TS.4 | Sea Level

Global mean sea level (GMSL) increased by 0.20 [0.15 to 0.25] m over the period 1901 to 2018, with a rate of rise
that has accelerated since the 1960s to 3.7 [3.2 to 4.2] mm yr –1 for the period 2006–2018 (high confidence). Human
activities were very likely the main driver of observed GMSL rise since 1971, and new observational evidence leads
to an assessed sea level rise over the period 1901 to 2018 that is consistent with the sum of individual components
contributing to sea level rise, including expansion due to ocean warming and melting of glaciers and ice sheets (high
confidence). It is virtually certain that GMSL will continue to rise over the 21st century in response to continued
warming of the climate system (Box TS.4, Figure 1). Sea level responds to greenhouse gas (GHG) emissions more
slowly than global surface temperature, leading to weaker scenario dependence over the 21st century than for global
surface temperature (high confidence). This slow response also leads to long-term committed sea level rise, associated
with ongoing ocean heat uptake and the slow adjustment of the ice sheets, that will continue over the centuries and
millennia following cessation of emissions (high confidence) (Box TS.9). By 2100, GMSL is projected to rise by 0.28–
0.55 m (likely range) under SSP1-1.9 and 0.63–1.01 m (likely range) under SSP5-8.5 relative to the 1995–2014 average
(medium confidence). Under the higher CO2 emissions scenarios, there is deep uncertainty in sea level projections for
2100 and beyond associated with the ice-sheet responses to warming. In a low-likelihood, high-impact storyline and a
high CO2 emissions scenario, ice-sheet processes characterized by deep uncertainty could drive GMSL rise up to about
5 m by 2150. Given the long-term commitment, uncertainty in the timing of reaching different GMSL rise levels is an
important consideration for adaptation planning. {2.3, 3.4, 3.5, 9.6, Box 9.4, Cross-Chapter Box 9.1, Table 9.5}

GMSL change is driven by warming or cooling of the ocean (and the associated expansion/contraction) and changes in the amount
of ice and water stored on land. Paleo-evidence shows that GMSL has been about 70 m higher and 130 m lower than present within
the past 55 million years and was likely 5 to 10 m higher during the Last Interglacial (Box TS.2, Figure 1). Sea level observations show
that GMSL rose by 0.20 [0.15 to 0.25] m over the period 1901–2018 at an average rate of 1.7 [1.3 to 2.2] mm yr –1. New analyses
and paleo-evidence since AR5 show this rate is very likely faster than during any century over at least the last three millennia (high
confidence). Since AR5, there is strengthened evidence for an increase in the rate of GMSL rise since the mid-20th century, with an
average rate of 2.3 [1.6–3.1] mm yr –1 over the period 1971–2018 increasing to 3.7 [3.2–4.2] mm yr –1 for the period 2006–2018 (high
confidence). {2.3.3, 9.6.1, 9.6.2}

77
 Technical Summary

Box TS.4 (continued)

(a) Global mean sea level rise from 1900–2150

2.5
SSP1-1.9 SSP1-2.6 SSP2-4.5 SSP3-7.0 SSP5-8.5
Median (medium confidence)
2 Likely range (medium confidence)
SSP3-7.0
SSP5-8.5 Low confidence 83rd percentile
GMSL rise (m)

1.5 SSP5-8.5 Low confidence 95th percentile

0.5
Observations SSP1-2.6 2150 medium & low
TS 0 confidence projections
(see caption)
1900 1950 2000 2050 2100 2150

(b) Committed sea level rise by warming level and time scale (c) Projected timing of sea level rise milestones
Paleo ranges

35
Year by which a rise of 2.0 m
above 1995–2014 is expected
30 10,000-yr

Mid-Pliocene
Warm Period
25
GMSL rise (m)

1.5 m
20

15 2,000-yr
Last
Interglacial 1.0 m
10
Low−confidence
processes included SSP5−8.5
SSP3−7.0
5 SSP2−4.5
SSP1−2.6
SSP1−1.9
100-yr 0.5 m
0
1 1.5 2 3 4 5
2000 2100 2200 2300+
Peak global surface temperature (°C)

Box TS.4, Figure 1 | Global mean sea level (GMSL) change on different time scales and under different scenarios. The intent of this figure is to
(i) show the century-scale GMSL projections in the context of the 20th century observations, (ii) illustrate ‘deep uncertainty’ in projections by considering the timing
of GMSL rise milestones, and (iii) show the long-term commitment associated with different warming levels, including the paleo evidence to support this. (a) GMSL
change from 1900 to 2150, observed (1900–2018) and projected under the SSP scenarios (2000–2150), relative to a 1995–2014 baseline. Solid lines show median
projections. Shaded regions show likely ranges for SSP1-2.6 and SSP3-7.0. Dotted and dashed lines show respectively the 83rd and 95th percentile low confidence
projections for SSP5-8.5. Bars at right show likely ranges for SSP1-1.9, SSP1-2.6, SSP2-4.5, SSP3-7.0 and SSP5-8.5 in 2150. Lightly shaded thick/thin bars show
17th–83rd/5th–95th percentile low-confidence ranges in 2150 for SSP1-2.6 and SSP5-8.5, based upon projection methods incorporating structured expert judgement
and marine ice cliff instability. Low confidence range for SSP5-8.5 in 2150 extends to 4.8/5.4 m at the 83rd/95th percentile. (b) GMSL change on 100- (blue), 2000-
(green) and 10,000-year (magenta) time scales as a function of global surface temperature, relative to 1850–1900. For 100-year projections, GMSL is projected for the
year 2100, relative to a 1995–2014 baseline, and temperature anomalies are average values over 2081–2100. For longer-term commitments, warming is indexed by
peak warming above 1850–1900 reached after cessation of emissions. Shaded regions show paleo-constraints on global surface temperature and GMSL for the Last
Interglacial and mid-Pliocene Warm Period. Lightly shaded thick/thin blue bars show 17th–83rd/5th–95th percentile low confidence ranges for SSP1-2.6 and SSP5-8.5
in 2100, plotted at 2°C and 5°C. (c) Timing of exceedance of GMSL thresholds of 0.5, 1.0, 1.5 and 2.0 m, under different SSPs. Lightly shaded thick/thin bars show
17th–83rd/5th–95th percentile low-confidence ranges for SSP1-2.6 and SSP5-8.5. {4.3.2, 9.6.1, 9.6.2, 9.6.3, Box 9.4}

78
 Technical Summary

Box TS.4 (continued)

GMSL will continue to rise throughout the 21st century (Box TS.4, Figure 1a). Considering only those processes in whose projections
we have at least medium confidence, relative to the period 1995–2014, GMSL is projected to rise between 0.18 m (0.15–0.23 m,
likely range; SSP1-1.9) and 0.23 m (0.20–0.30 m, likely range; SSP5-8.5) by 2050. By 2100, the projected rise is between 0.38 m
(0.28–0.55 m, likely range; SSP1-1.9) and 0.77 m (0.63–1.01 m, likely range; SSP5-8.5) {Table 9.9}. The methods, models and scenarios
used for sea level projections in the AR6 are updated from those employed by SROCC, with contributions informed by the latest
model projections described in the ocean and cryosphere Sections (Sections TS.2.4 and TS.2.5). Despite these differences, the sea level
projections are broadly consistent with those of SROCC. {4.3.2, 9.6.3}

Importantly, likely range projections do not include those ice-sheet-related processes whose quantification is highly uncertain
or that are characterized by deep uncertainty. Higher amounts of GMSL rise before 2100 could be caused by earlier-than-projected
disintegration of marine ice shelves, the abrupt, widespread onset of marine ice sheet instability (MISI) and marine ice cliff instability
(MICI) around Antarctica, and faster-than-projected changes in the surface mass balance and dynamical ice loss from Greenland TS
(Box TS.4, Figure 1). In a low-likelihood, high-impact storyline and a high CO2 emissions scenario, such processes could in combination
contribute more than one additional meter of sea level rise by 2100 (Box TS.3). {4.3.2, 9.6.3, Box 9.4}

Beyond 2100, GMSL will continue to rise for centuries to millennia due to continuing deep ocean heat uptake and mass loss from
ice sheets, and will remain elevated for thousands of years (high confidence). By 2150, considering only those processes in whose
projections we have at least medium confidence and assuming no acceleration in ice-mass flux after 2100, GMSL is projected to rise
between 0.6 m (0.4–0.9 m, likely range, SSP1-1.9) and 1.3 m (1.0–1.9 m, likely range) (SSP5-8.5), relative to the period 1995–2014
based on the SSP scenario extensions. Under high CO2 emissions, processes in which there is low confidence, such as MICI, could drive
GMSL rise up to about 5 m by 2150 (Box TS.4, Figure 1a). By 2300, GMSL will rise 0.3–3.1 m under low CO2 emissions (SSP1-2.6) (low
confidence). Under high CO2 emissions (SSP5-8.5), projected GMSL rise is between 1.7 and 6.8 m by 2300 in the absence of MICI and
by up to 16 m considering MICI (low confidence). Over 2000 years, there is medium agreement and limited evidence that committed
GMSL rise is projected to be about 2–3 m with 1.5°C peak warming, 2–6 m with 2°C of peak warming, 4–10 m with 3°C of peak
warming, 12–16 m with 4°C of peak warming, and 19–22 m with 5°C of peak warming. {9.6.3}

Looking at uncertainty in time provides an alternative perspective on uncertainty in future sea level rise (Box TS.4, Figure 1c). For
example, considering only medium confidence processes, GMSL rise is likely to exceed 0.5 m between about 2080 and 2170 under
SSP1-2.6 and between about 2070 and 2090 under SSP5-8.5. Given the long-term commitment, uncertainty in the timing of reaching
different levels of GMSL rise is an important consideration for adaptation planning. {9.6.3}

At regional scales, additional processes come into play that modify the local sea level change relative to GMSL, including vertical land
motion, ocean circulation and density changes, and gravitational, rotational, and deformational effects arising from the redistribution
of water and ice mass between land and the ocean. These processes give rise to a spatial pattern that tends to increase sea level
rise at the low latitudes and reduce sea level rise at high latitudes. However, over the 21st century, the majority of coastal locations
have a median projected regional sea level rise within ±20% of the projected GMSL change (medium confidence). Further details on
regional sea level change and extremes are provided in Section TS.4. {9.6.3}

Box TS.5 | The Carbon Cycle

The continued growth of atmospheric CO2 concentrations over the industrial era is unequivocally due to emissions from
human activities. Ocean and land carbon sinks slow the rise of CO2 in the atmosphere. Projections show that while
land and ocean sinks absorb more CO2 under high emissions scenarios than low emissions scenarios, the fraction of
emissions removed from the atmosphere by natural sinks decreases with higher concentrations (high confidence).
Projected ocean and land sinks show similar responses for a given scenario, but the land sink has a much higher
interannual variability and wider model spread. The slowed growth rates of the carbon sinks projected for the second
half of this century are linked to strengthening carbon–climate feedbacks and stabilization of atmospheric CO2 under
medium-to-no-mitigation and high-mitigation scenarios, respectively (see FAQ 5.1). {5.2, 5.4}

79
 Technical Summary

Box TS.5 (continued)

Carbon sinks for anthropogenic CO2 are associated with mainly physical ocean and biospheric land processes that drive the exchange
of carbon between multiple land, ocean and atmospheric reservoirs. These exchanges are driven by increasing atmospheric CO2, but
are modulated by changes in climate (Box TS.5, Figure 1c,d). The Northern and Southern Hemispheres dominate the land and ocean
sinks, respectively (Box TS.5, Figure 1). Ocean circulation and thermodynamic processes also play a critical role in coupling the global
carbon and energy (heat) cycles. There is high confidence that this ocean carbon–heat nexus is an important basis for one of the
most important carbon–climate metrics, the transient climate response to cumulative CO2 emissions (TCRE; Section TS.3.2.1) used to
determine the remaining carbon budget. {5.1, 5.2, 5.5, 9.2, Cross-Chapter Box 5.3}

Based on multiple lines of evidence using interhemispheric gradients of CO2 concentrations, isotopes, and inventory data, it is
unequivocal that the growth in CO2 in the atmosphere since 1750 (see Section TS.2.2) is due to the direct emissions from human
activities. The combustion of fossil fuels and land-use change for the period 1750–2019 resulted in the release of 700 ± 75 PgC (likely
TS range, 1 PgC = 1015 g of carbon) to the atmosphere, of which about 41% ± 11% remains in the atmosphere today (high confidence).
Of the total anthropogenic CO2 emissions, the combustion of fossil fuels was responsible for about 64% ± 15%, growing to an 86% ±
14% contribution over the past 10 years. The remainder resulted from land-use change. During the last decade (2010–2019), average
annual anthropogenic CO2 emissions reached the highest levels in human history at 10.9 ± 0.9 PgC yr –1 (high confidence). Of these
emissions, 46% accumulated in the atmosphere (5.1 ± 0.02 PgC yr –1), 23% (2.5 ± 0.6 PgC yr –1) was taken up by the ocean and 31%
(3.4 ± 0.9 PgC yr –1) was removed by terrestrial ecosystems (high confidence). {5.2.1, 5.2.2, 5.2.3}

The ocean (high confidence) and land (medium confidence) sinks of CO2 have increased with anthropogenic emissions over the past
six decades (Box TS.5, Figure 1). This coherence between emissions and the growth in ocean and land sinks has resulted in the airborne
fraction of anthropogenic CO2 remaining at 44 ± 10% over the past 60 years (high confidence). Interannual and decadal variability of
the ocean and land sinks indicate that they are sensitive to changes in the growth rate of emissions as well as climate variability and
are therefore also sensitive to climate change (high confidence). {5.2.1}

The land CO2 sink is driven by carbon uptake by vegetation, with large interannual variability, for example, linked to the El Niño–
Southern Oscillation (ENSO). Since the 1980s, carbon fertilization from rising atmospheric CO2 has increased the strength of the net
land CO2 sink (medium confidence). During the historical period, the growth of the ocean sink has been primarily determined by
the growth rate of atmospheric CO2. However, there is medium confidence that changes to physical and chemical processes in the
ocean and in the land biosphere, which govern carbon feedbacks, are already modifying the characteristics of variability, particularly
the seasonal cycle of CO2, in both the ocean and land. However, changes to the multi-decadal trends in the sinks have not yet been
observed. {2.3.4, 3.6.1, 5.2.1}

In AR6, ESM projections are assessed with CO2 concentrations by 2100 from about 400 ppm (SSP1-1.9) to above 1100 ppm (SSP5-8.5).
Most simulations are performed with prescribed atmospheric CO2 concentrations, which already account for a central estimate of
climate–carbon feedback effects. Carbon dioxide emissions-driven simulations account for uncertainty in these feedbacks, but do
not significantly change the projected global surface temperature changes (high confidence). Although land and ocean sinks absorb
more CO2 under high emissions than low emissions scenarios, the fraction of emissions removed from the atmosphere decreases
(high confidence). This means that the more CO2 that is emitted, the less efficient the ocean and land sinks become (high confidence),
an effect which compensates for the logarithmic relationship between CO2 and its radiative forcing, which means that for each unit
increase in additional atmospheric CO2 the effect on global temperature decreases. (Box TS.5, Figure 1f,g). {4.3.1, 5.4.5, 5.5.1.2}

Ocean and land sinks show similar responses for a given scenario, but the land sink has a much higher interannual variability
and wider model spread. Under SSP3-7.0 and SSP5-8.5, the initial growth of both sinks in response to increasing atmospheric
concentrations of CO2 is subsequently limited by emerging carbon–climate feedbacks (high confidence) (Box TS.5, Figure 1f).
Projections show that the ocean and land sinks will stop growing from the second part of the 21st century under all emissions
scenarios, but with different drivers for different emissions scenarios. Under SSP3-7.0 and SSP5-8.5, the weakening growth rate
of the ocean CO2 sink in the second half of the century is primarily linked to the strengthening positive feedback from reduced
carbonate buffering capacity, ocean warming and altered ocean circulation (e.g., AMOC changes). In contrast, for SSP1-1.9, SSP1-2.6
and SSP2-4.5, the weakening growth rate of the ocean carbon sink is a response to the stabilizing or declining atmospheric
CO2 concentrations. Under SSP1-1.9, models project that combined land and ocean sinks will turn into a weak source by 2100
(medium confidence). Under high CO2 emissions scenarios, it is very likely that the land carbon sink will grow more slowly due to
warming and drying from the mid-21st century, but it is very unlikely that it will switch from being a sink to a source before 2100.

80
 Technical Summary

Box TS.5 (continued)

    

Latitute
Colour High model agreement Low model agreement

TS
  

number of model simulations

shades: uncertainty range

   
)

   

Box TS.5, Figure 1 | Carbon cycle processes and projections.

81
 Technical Summary

Box TS.5 (continued)

Box TS.5, Figure 1 (continued): The intent of this figure is to show the response of the carbon cycle to carbon dioxide (CO2) emissions and climate and its
role in determining future CO2 levels through projected changes to sinks and sink fractions. The figure shows changes in carbon storage in response to elevated
CO2 (a, b) and the response to climate warming (c, d). Maps show spatial patterns of changes in carbon uptake during simulations with 1% per year increase in CO2
(Section 5.4.5.5), and zonal mean plots show distribution of carbon changes is dominated by the land (green lines) in the tropics and Northern Hemisphere and ocean
(blue lines) in the Southern Hemisphere. Hatching indicates regions where fewer than 80% of models agree on the sign of response. (e) Future CO2 projections:
projected CO2 concentrations in the Shared Socio-economic Pathway (SSP) scenarios in response to anthropogenic emissions, results from coupled Earth system
models for SSP5-8.5 and from the MAGICC7 emulator for other scenarios (Section 4.3.1). (f) Future carbon fluxes: projected combined land and ocean fluxes (positive
downward) up to 2100 for the SSP scenarios, and extended to 2300 for available scenarios, 5–95% uncertainty plumes shown for SSP1-2.6 and SSP3-7.0 (Sections
4.3.2.4, 5.4.5.4 and 5.4.10). The numbers near the top show the number of model simulations used. (g) Sink fraction: the fraction of cumulative emissions of CO2
removed by land and ocean sinks. The sink fraction is smaller under conditions of higher emissions. {Figure 4.3; 5.4.5; Figures 5.25, 5.27 and 5.30}

Climate change alone is expected to increase land carbon accumulation in the high latitudes (not including permafrost, which is
assessed in Sections TS.2.5 and TS.3.2.2), but also to lead to a counteracting loss of land carbon in the tropics (medium confidence).
Earth system model projections show that the overall uncertainty of atmospheric CO2 by 2100 is still dominated by the emissions
TS pathway, but carbon–climate feedbacks (see Section TS.3.3.2) are important, with increasing uncertainties in high emissions pathways
(Box TS.5, Figure 1e). {4.3.2, 5.4.1, 5.4.2, 5.4.4, 5.4.5, 11.6, 11.9, Cross-Chapter Box 5.1, Cross-Chapter Box 5.3}

Under three SSP scenarios with long-term extensions until 2300 (SSP5-8.5, SSP5-3.4-OS, SSP1-2.6), ESMs project a change of the
land from a sink to a source (medium confidence). The scenarios make simplified assumptions about emissions reductions, with
SSP1-2.6 and SSP5-3.4-OS reaching about 400 ppm by 2300, while SSP5-8.5 exceeds 2000 ppm. Under high emissions, the transition
is warming-driven, whereas it is linked to the decline in atmospheric CO2 under net negative CO2 emissions. The ocean remains a sink
throughout the period to 2300 except under very large net negative emissions. The response of the natural aspects of the carbon cycle
to carbon dioxide removal is further developed in Section TS.3.3.2. {5.4.9}

TS.2.6 Land Climate, Including Biosphere and Extremes and about 80% larger than warming of the ocean surface. Warming of
the land surface during the period 1971–2018 contributed about 5%
Land surface air temperatures have risen faster than the of the increase in the global energy inventory (Section TS.3.1), nearly
global surface temperature since the 1850s, and it is virtually twice the estimate in AR5 (high confidence). It is virtually certain that
certain that this differential warming will persist into the the average surface warming over land will continue to be higher than
future. It is virtually certain that the frequency and intensity over the ocean throughout the 21st century. The warming pattern
of hot extremes and the intensity and duration of heatwaves will likely vary seasonally, with northern high latitudes warming more
have increased since 1950 and will further increase in the during winter than summer (medium confidence). {2.3.1, 4.3.1, 4.5.1,
future even if global warming is stabilized at 1.5°C. The 7.2.2, Box 7.2, Cross-Chapter Box 9.1, 11.3, Atlas 11.2}
frequency and intensity of heavy precipitation events have
increased over a majority of those land regions with good The frequency and intensity of hot extremes (warm days and nights)
observational coverage (high confidence) and will extremely and the intensity and duration of heatwaves have increased globally
likely increase over most land regions with additional global and in most regions since 1950, while the frequency and intensity
warming. of cold extremes have decreased (virtually certain). There is high
confidence that the increases in frequency and severity of hot
Over the past half century, key aspects of the biosphere extremes are due to human-induced climate change. Some recent
have changed in ways that are consistent with large- extreme events would have been extremely unlikely to occur without
scale warming: climate zones have shifted poleward, and human influence on the climate system. It is virtually certain that
the growing season length in the Northern Hemisphere further changes in hot and cold extremes will occur throughout the
extratropics has increased (high confidence). The amplitude 21st century in nearly all inhabited regions, even if global warming
of the seasonal cycle of atmospheric CO2 poleward of 45°N is stabilized at 1.5°C (Table TS.2, Figure TS.12a). {1.3, Cross-Chapter
has increased since the 1960s (very high confidence), with Box 3.2, 11.1.4, 11.3.2, 11.3.4, 11.3.5, 11.9, 12.4}
increasing productivity of the land biosphere due to the
increasing atmospheric CO2 concentration as the main driver Greater warming over land alters key water cycle characteristics
(medium confidence). Global-scale vegetation greenness has (Box TS.6). The rates of change in mean precipitation and runoff,
increased since the 1980s (high confidence). {2.3, 3.6, 4.3, and their variability, increase with global warming (Figure TS.12e,f).
4.5, 5.2, 11.3, 11.4, 11.9, 12.4} Human-induced climate change has contributed to increases in
agricultural and ecological droughts in some regions due to increases
Observed temperatures over land have increased by 1.59 [1.34– in evapotranspiration (medium confidence). More regions are
1.83] °C between the period 1850–1900 and 2011–2020. Warming affected by increases in agricultural and ecological droughts with
of the land is about 45% larger than for global surface temperature increasing global warming (high confidence; see also Figure TS.12c).

82
 Technical Summary

(a) Hot extreme events (b) Heavy precipitation events

40
10-year event, frequency
6 6 10-year event, frequency
40
Relative frequency (pre-industrial = 1)

Relative frequency (pre-industrial = 1)

50-year event, frequency 50-year event, frequency 35
35 10-year event, intensity 5 10-year event, intensity
50-year event, intensity 5 50-year event, intensity 30

Intensity change (°C)

Intensity change (%)

30
4 25
25 4
20 20
3
3
15 15
2
10
2 10
5 1
5
0 1
1.0 1.5 2.0 3.0 4.0 1.0 1.5 2.0 3.0 4.0
Global surface temperature change since 1850–1900 (°C) Global surface temperature change since 1850–1900 (°C)
TS

(c) Droughts in drought-prone regions (d) Northern Hemisphere March–May snow cover
8

Amplitude of annual soil moisture decrease (-)

20 SSP5-8.5
10-year event, frequency
SSP3-7.0
Relative frequency (pre-industrial = 1)

7 10-year event, intensity 2 SSP2-4.5

Snow cover extent change (%)

0 SSP1-2.6
6 SSP1-1.9
1.5
5 - 20

4 1
- 40
3
0.5
- 60
2
0
1 - 80

1.0 1.5 2.0 3.0 4.0 0 1 2 3 4 5 6 7

Global surface temperature change since 1850–1900 (°C) Global surface temperature change since 1850–1900 (°C)

(e) Hydrological change over tropical land (f) Hydrological change over extratropical land
50 Precipitable water annual mean Precipitable water annual mean
Precipitation 60 Precipitation
% change - 15 models mean - SSP5-8.5

% change - 15 models mean - SSP5-8.5

annual mean annual mean

40
interannual variability 50 interannual variability
Runoff Runoff
30 annual mean annual mean
interannual variability 40 interannual variability

20 30

10 20

10
0
17–83% range 0 17–83% range
-10
2.0 3.0 4.0 5.0 2.0 3.0 4.0 5.0
Global surface temperature change since 1850–1900 (°C) Global surface temperature change since 1850–1900 (°C)

Figure TS.12 | Land-related changes relative to the 1850-1900 as a function of global warming levels. The intent of this figure is to show that extremes and
mean land variables change consistently with warming levels and to show the changes with global warming levels of water cycle indicators (i.e., precipitation and runoff) over
tropical and extratropical land in terms of mean and interannual variability (interannual variability increases at a faster rate than the mean). (a) Changes in the frequency (left
scale) and intensity (in °C, right scale) of daily hot extremes occurring every 10 and 50 years. (b) as (a), but for daily heavy precipitation extremes, with intensity change in %.
(c) Changes in 10-year droughts aggregated over drought-prone regions (WNA, CNA, NCA, SCA, NSA, NES, SAM, SWS, SSA, WCE, MED, WSAF, ESAF, MDG, SAU, and EAU;
for definitions of these regions, see Figure Atlas.2), with drought intensity (right scale) represented by the change of annual mean soil moisture, normalized with respect to
interannual variability. Limits of the 5%−95% confidence interval are shown in panels (a–c). (d) Changes in Northern Hemisphere spring (March–April–May) snow cover extent
relative to 1850–1900; (e,f) Relative change (%) in annual mean of total precipitable water (grey line), precipitation (red solid lines), runoff (blue solid lines) and in standard
deviation (i.e., variability) of precipitation (red dashed lines) and runoff (blue dashed lines) averaged over (e) tropical and (f) extratropical land as function of global warming
levels. Coupled Model Intercomparison Project Phase 6 (CMIP6) models that reached a 5°C warming level above the 1850–1900 average in the 21st century in SSP5-8.5 have
been used. Precipitation and runoff variability are estimated by respective standard deviation after removing linear trends. Error bars show the 17–83% confidence interval for
the warmest +5°C global warming level. {Figures 8.16, 9.24, 11.6, 11.7, 11.12, 11.15, 11.18 and Atlas.2}

83
 Technical Summary

There is low confidence that the increase of plant water-use efficiency Hemisphere extratropics since the mid-20th century (high
due to higher atmospheric CO2 concentration alleviates extreme confidence), are consistent with large-scale warming. At the
agricultural and ecological droughts in conditions characterized same time an increase in the amplitude of the seasonal cycle of
by limited soil moisture and increased atmospheric evaporative atmospheric CO2 poleward of 45°N since the early 1960s (high
demand. {2.3.1, Cross-Chapter Box 5.1, 8.2.3, 8.4.1, 11.2.4, 11.4, confidence) and a global-scale increase in vegetation greenness
11.6, Box 11.1} of the terrestrial surface since the early 1980s (high confidence)
have been observed. Increasing atmospheric CO2, warming at high
Northern Hemisphere spring snow cover has decreased since at latitudes, and land management interventions have contributed
least 1978 (very high confidence), and there is high confidence to the observed greening trend, but there is low confidence in
that trends in snow cover loss extend back to 1950. It is very likely their relative roles. There is medium confidence that increased
that human influence contributed to these reductions. Earlier onset plant growth associated with CO2 fertilization is the main driver
of snowmelt has contributed to seasonally dependent changes of the observed increase in amplitude of the seasonal cycle of
in streamflow (high confidence). A further decrease of Northern atmospheric CO2 in the Northern Hemisphere. Reactive nitrogen,
Hemisphere seasonal snow cover extent is virtually certain under ozone and aerosols affect terrestrial vegetation and carbon cycle
TS further global warming (Figure TS.12d). {2.3.2, 3.4.2, 8.3.2. 9.5.3, through deposition and effects on large-scale radiation (high
12.4, 9.2, 11.2, Atlas 8.2} confidence), but the magnitude of these effects on the land carbon
sink, ecosystem productivity and indirect CO2 forcing remains
The frequency and intensity of heavy precipitation events have uncertain. {2.3.4, 3.6.1, 5.2.1, 6.4.5, 12.3.7, 12.4}
increased over a majority of land regions with good observational
coverage since 1950 (high confidence, Box TS.6, Table TS.2). Human Over the last century, there has been a poleward and upslope shift
influence is likely the main driver of this change (Table TS.2). It is in the distribution of many land species (very high confidence)
extremely likely that on most land regions heavy precipitation will as well as increases in species turnover within many ecosystems
become more frequent and more intense with additional global (high confidence). There is high confidence that the geographical
warming (Table TS.2, Figure TS.12b). The projected increase in heavy distribution of climate zones has shifted in many parts of the
precipitation extremes translates to an increase in the frequency and world in the last half century. The SRCCL concluded that continued
magnitude of pluvial floods (high confidence) (Table TS.2). {Cross- warming will exacerbate desertification processes (medium
Chapter Box 3.2, 8.4.1, 11.4.2, 11.4.4, 11.5.5, 12.4} confidence) and that ecosystems will become increasingly exposed
to climates beyond those that they are currently adapted to (high
The probability of compound extreme events has likely increased confidence). There is medium confidence that climate change will
due to human-induced climate change. Concurrent heatwaves and increase disturbance by, for example, fire and tree mortality, across
droughts have become more frequent over the last century, and this several ecosystems. Increases are projected in drought, aridity and
trend will continue with higher global warming (high confidence). fire weather in some regions (Section TS.4.3; high confidence).
The probability of compound flooding (storm surge, extreme There is low confidence in the magnitude of these changes, but
rainfall and/or river flow) has increased in some locations and will the probability of crossing uncertain regional thresholds (e.g., fires,
continue to increase due to both sea level rise and increases in heavy forest dieback) increases with further warming (high confidence).
precipitation, including changes in precipitation intensity associated The response of biogeochemical cycles to the anthropogenic
with tropical cyclones (high confidence). {11.8.1, 11.8.2, 11.8.3} perturbation can be abrupt at regional scales, and irreversible on
decadal to century time scales (high confidence). {2.3.4, 5.4.3,
Changes in key aspects of the terrestrial biosphere, such as an 5.4.9, 11.6, 11.8, 12.5, SRCCL 2.2, SRCCL 2.5, SR1.5 3.4}
increase of the growing season length in much of the Northern

84
 Technical Summary

Box TS.6 | Water Cycle

Human-caused climate change has driven detectable changes in the global water cycle since the mid-20th century (high
confidence), and it is projected to cause substantial further changes at both global and regional scales (high confidence).

Global land precipitation has likely increased since 1950, with a faster increase since the 1980s (medium confidence).
Atmospheric water vapour has increased throughout the troposphere since at least the 1980s (likely). Annual global
land precipitation will increase over the 21st century as global surface temperature increases (high confidence).
Human influence has been detected in amplified surface salinity and precipitation minus evaporation (P–E) patterns
over the ocean (high confidence).

The severity of very wet and very dry events increase in a warming climate (high confidence), but changes in
atmospheric circulation patterns affect where and how often these extremes occur. Water cycle variability and related
extremes are projected to increase faster than mean changes in most regions of the world and under all emissions TS
scenarios (high confidence).

Over the 21st century, the total land area subject to drought will increase and droughts will become more frequent
and severe (high confidence). Near-term projected changes in precipitation are uncertain mainly because of internal
variability, model uncertainty and uncertainty in forcings from natural and anthropogenic aerosols (medium
confidence).

Over the 21st century and beyond, abrupt human-caused changes to the water cycle cannot be excluded (medium
confidence). {2.3, 3.3, 4.3, 4.4, 4.5, 4.6, 8.2, 8.3, 8.4, 8.5, 8.6, 11.4, 11.6, 11.9}

There is high confidence that the global water cycle has intensified since at least 1980 expressed by, for example, increased atmospheric
moisture fluxes and amplified precipitation minus evaporation patterns. Global land precipitation has likely increased since 1950,
with a faster increase since the 1980s (medium confidence), and a likely human contribution to patterns of change, particularly for
increases in high-latitude precipitation over the Northern Hemisphere. Increases in global mean precipitation are determined by a
robust response to global surface temperature (very likely 2–3% per °C) that is partly offset by fast atmospheric adjustments to
atmospheric heating by greenhouse gases (GHGs) and aerosols (Section TS.3.2.2). The overall effect of anthropogenic aerosols is
to reduce global precipitation through surface radiative cooling effects (high confidence). Over much of the 20th century, opposing
effects of GHGs and aerosols on precipitation have been observed for some regional monsoons (high confidence) (Box TS.13). Global
annual precipitation over land is projected to increase on average by 2.4% (–0.2% to +4.7% likely range) under SSP1-1.9, 4.6% (1.5%
to 8.3% likely range) under SSP2-4.5, and 8.3% (0.9% to 12.9% likely range) under SSP5-8.5 by 2081–2100 relative to 1995–2014
(Box TS.6, Figure 1). Inter-model differences and internal variability contribute to a substantial range in projections of large-scale and
regional water cycle changes (high confidence). The occurrence of volcanic eruptions can alter the water cycle for several years (high
confidence). Projected patterns of precipitation change exhibit substantial regional differences and seasonal contrast as global surface
temperature increases over the 21st century (Box TS.6, Figure 1). {2.3.1, 3.3.2, 3.3.3, 3.5.2, 4.3.1, 4.4.1, 4.5.1, 4.6.1, Cross-Chapter
Box 4.1, 8.2.1, 8.2.2, 8.2.3, Box 8.1, 8.3.2.4, 8.4.1, 8.5.2, 10.4.2}

Global total column water vapour content has very likely increased since the 1980s, and it is likely that human influence has contributed
to tropical upper tropospheric moistening. Near-surface specific humidity has increased over the ocean (likely) and land (very likely)
since at least the 1970s, with a detectable human influence (medium confidence). Human influence has been detected in amplified
surface salinity and precipitation minus evaporation (P–E) patterns over the ocean (high confidence). It is virtually certain that
evaporation will increase over the ocean and very likely that evapotranspiration will increase over land, with regional variations under
future surface warming (Box TS.6, Figure 1). There is high confidence that projected increases in precipitation amount and intensity will
be associated with increased runoff in northern high latitudes (Box TS.6, Figure 1). In response to cryosphere changes (Section TS.2.5),
there have been changes in streamflow seasonality, including an earlier occurrence of peak streamflow in high-latitude and mountain
catchments (high confidence). Projected runoff (Box TS.6, Figure 1c) is typically decreased by contributions from small glaciers because
of glacier mass loss, while runoff from larger glaciers will generally increase with increasing global warming levels until their mass
becomes depleted (high confidence). {2.3.1, 3.3.2, 3.3.3, 3.5.2, 8.2.3, 8.4.1, 11.5}

85
 Technical Summary

Box TS.6 (continued)

Warming over land drives an increase in atmospheric evaporative demand and in the severity of drought events (high confidence). Greater
warming over land than over the ocean alters atmospheric circulation patterns and reduces continental near-surface relative humidity,
which contributes to regional drying (high confidence). A very likely decrease in relative humidity has occurred over much of the global
land area since 2000. Projected increases in evapotranspiration due to growing atmospheric water demand will decrease soil moisture
over the Mediterranean region, south-western North America, South Africa, South-Western South America and south-western Australia
(high confidence) (Box TS.6, Figure 1). Some tropical regions are also projected to experience enhanced aridity, including the Amazon
basin and Central America (high confidence). The total land area subject to increasing drought frequency and severity will expand (high
confidence), and in the Mediterranean, South-Western South America, and Western North America, future aridification will far exceed the
magnitude of change seen in the last millennium (high confidence). {4.5.1, 8.2.2, 8.2.3, 8.4.1, Box 8.2, 11.6, 11.9}

Land-use change and water extraction for irrigation have influenced local and regional responses in the water cycle (high confidence).
TS Large-scale deforestation likely decreases evapotranspiration and precipitation and increases runoff over the deforested regions
relative to the regional effects of climate change (medium confidence). Urbanization increases local precipitation (medium confidence)
and runoff intensity (high confidence) (Box TS.14). Increased precipitation intensities have enhanced groundwater recharge, most
notably in tropical regions (medium confidence). There is high confidence that groundwater depletion has occurred since at least the
start of the 21st century, as a consequence of groundwater withdrawals for irrigation in agricultural areas in drylands. {8.2.3, 8.3.1,
11.1.6, 11.4, 11.6, FAQ 8.1}

Water cycle variability and related extremes are projected to increase faster than mean changes in most regions of the world
and under all emissions scenarios (high confidence). A warmer climate increases moisture transport into weather systems, which
intensifies wet seasons and events (high confidence). The magnitudes of projected precipitation increases and related extreme
events depend on model resolution and the representation of convective processes (high confidence). Increases in near-surface
atmospheric moisture capacity of about 7% per 1ºC of warming lead to a similar response in the intensification of heavy precipitation
from sub-daily up to seasonal time scales, increasing the severity of flood hazards (high confidence). The average and maximum
rain-rates associated with tropical and extratropical cyclones, atmospheric rivers and severe convective storms will therefore also
increase with future warming (high confidence). For some regions, there is medium confidence that peak tropical cyclone rain-
rates will increase by more than 7% per 1°C of warming due to increased low-level moisture convergence caused by increases
in wind intensity. In the tropics year-round and in the summer season elsewhere, interannual variability of precipitation and
runoff over land is projected to increase at a faster rate than changes in seasonal mean precipitation (Figure TS.12e,f) (medium
confidence). Sub-seasonal precipitation variability is also projected to increase, with fewer rainy days but increased daily mean
precipitation intensity over many land regions (high confidence). {4.5.3, 8.2.3, 8.4.1, 8.4.2, 8.5.1, 8.5.2, 11.4, 11.5, 11.7, 11.9}

86
 Technical Summary

Box TS.6 (continued)

Long-term water cycle variables changes for SSP2-4.5 (2081–2100 vs 1995–2014)

(a) Precipitation (b) Evapotranspiration

TS

(c) Runoff (d) Surface soil moisture

Box TS.6, Figure 1 | Projected water cycle changes. The intent of this figure is to give a geographical overview of changes in multiple components of the global
water cycle using an intermediate emissions scenario. Important key message: without drastic reductions in greenhouse gas emissions, human-induced global warming
will be associated with widespread changes in all components of the water cycle. Long-term (2081–2100) projected annual mean changes (%) relative to present-day
(1995–2014) in the SSP2-4.5 emissions scenario for (a) precipitation, (b) surface evapotranspiration, (c) total runoff and (d) surface soil moisture. Numbers in top
right of each panel indicate indicate the number of Coupled Model Intercomparison Project Phase 6 (CMIP6) models used for estimating the ensemble mean. For other
scenarios, please refer to relevant figures in Chapter 8. Uncertainty is represented using the simple approach: No overlay indicates regions with high model agreement,
where ≥80% of models agree on sign of change; diagonal lines indicate regions with low model agreement, where <80% of models agree on sign of change. For
more information on the simple approach, please refer to the Cross-Chapter Box Atlas.1. {8.4.1; Figures 8.14, 8.17, 8.18, and 8.19}

87
 Technical Summary

Infographic TS.1 | Climate Futures

Climate futures
The climate change that people will experience this century and beyond depends on our
greenhouse gases emissions, how much global warming this will cause and the
response of the climate system to this warming.
Emissions pathways
Different social and economic developments can lead to substantially different future emissions of
carbon dioxide (CO 2 ), other greenhouse gases and air pollutants for the rest of the century.

Today
CO2 peak
TS 140
Very high
CO2 emissions (billion tonnes CO2 per year)

120

100 CO2 doubled

CO2 doubled
High
80

60
CO2 peak
40
CO2 halved

20 CO2 peak
CO2 peak Medium
0
Low
Net CO2 removal Very low
-20
2000 2020 2040 2060 2080 2100

Effect on surface temperature

For temperature to stabilize, CO2 emissions need to reach net zero.
Global warming since 1850–1900 (°C)

0 0.5 1 1.5 2 3 4

Very high
High
Medium
Low
Very low
1980 2000 2020 2040 2060 2080

Today

Short-term effect: Natural variability

Over short time scales (typically a decade), natural variability can temporarily dampen or accentuate global warming trends
resulting from emissions.

Infographic TS.1 | Climate Futures. The intent of this figure is to show possible climate futures: The climate change that people will experience this century and beyond
depends on our greenhouse gas emissions, how much global warming this will cause and the response of the climate system to this warming.
(top left) Annual emissions of CO2 for the five core Shared Socio-economic Pathway (SSP) scenarios (very low: SSP1-1.9, low: SSP1-2.6, intermediate: SSP2-4.5, high: SSP3-7.0,
very high: SSP5-8.5).(bottom left) Projected warming for each of these emissions scenarios.

88
 Technical Summary

Climate futures

Response of the climate system relative to 1850–1900

Many aspects of the climate system react quickly to temperature changes.
At progressively higher levels of global warming there are greater consequences (min/max range shown).

+1.1°C +1.5°C +2°C +4°C

Today TS
Temperature +1.9°C +2.6°C +5.1°C
Hottest day in (+1.3 to 2.3°C) (+1.8 to 3.1°C) (+4.3 to 5.8°C)
a decade (+°C) +1.2°C
(+0.7 to 1.5°C)

Drought
A drought that used to occur x2.0 x2.4 x4.1
once in a decade now (x1.0 to 5.1) (x1.3 to 5.8) (x1.7 to 7.2)
happens x times more
x1.7
(x0.7 to 4.1)

Precipitation
What used to be a wettest x1.5 x1.7 x2.7
day in a decade now (x1.4 to 1.7) (x1.6 to 2.0) (x2.3 to 3.6)
happens x times more x1.3
(x1.2 to1.4)

Snow
Snow cover extent -5% -9% -26%
change (%)
-1% (-13 to 2)
(-3 to 1) (-7 to 2) (-35 to -15)

Tropical cyclones
Proportion of intense tropical +10% +13% +30%
cyclones (%)

Long-term consequences: Sea level rise

Today, sea level has already

+1.5°C +2°C +4°C
Metre rise Metre rise Metre rise
increased by 20 cm and 33m
will increase an additional
30 cm to 1 m or more by 2100,
depending on future emissions.
16m 19m
13m
Sea level reacts very slowly to 7m
12m
6m
global warming so, once started, 3m 6m
8m

the rise continues for thousands 2m 2m

2000-year 10,000-year 2000-year 10,000-year 2000-year 10,000-year
of years. commitment commitment commitment commitment commitment commitment

The future...
The climate we and the young generations will experience depends on future emissions.
Reducing emissions rapidly will limit further changes, but continued emissions will trigger larger,
faster changes that will increasingly affect all regions. Some changes will persist for hundreds or
thousands of years, so today’s choices will have long-lasting consequences.
(top right) Response of some selected climate variables to four levels of global warming (°C). Changes in the ‘Today’ column are based on a global warming level of 1°C.
(bottom right) The long-term effect of each global warming level on sea level. See Section TS.1.3.1 for more detail on the SSP climate change scenarios.
This infographic builds from several figures in the Technical Summary: Figure TS.4 (for top left panel), Figure TS.6 (bottom left), Figure TS.12 (top right) and Box TS.4, Figure 1b
(bottom right).

89
 Technical Summary

TS.3 Understanding the Climate System TS.3.1 Radiative Forcing and Energy Budget
Response and Implications for Limiting
Global Warming Since AR5, the accumulation of energy in the Earth system,
quantified by observations of warming of the ocean,
This section summarizes advances in our knowledge of Earth’s atmosphere, and land and melting of ice, has become
energy budget, including the time evolution of forcings and climate established as a robust measure of the rate of global climate
feedbacks that lead to the climate system responses summarized change on interannual-to-decadal time scales. Compared to
in Section TS.2. It assesses advances since AR5 and SR1.5 in the changes in global surface temperature, the increase in the
estimation of remaining carbon budgets, the Earth system response to global energy inventory exhibits less variability, and thus
carbon dioxide removal, and the quantification of metrics that allow better indicates underlying climate trends.
comparisons of the relative effects of different forcing agents. The
section also highlights: future climate and air pollution responses due The global energy inventory increased by 282 [177 to
to projected changes in short-lived climate forcers (SLCFs); the state 387] zettajoules (ZJ, equal to 1021 Joules) for the period
of understanding of the climate response to potential interventions 1971–2006 and 152 [100 to 205] ZJ for the period 2006–
TS related to solar radiation modification (SRM); and irreversibility, 2018 (Figure TS.13), with more than 90% accounted for
tipping points and abrupt changes in the climate system.

(a) Global energy inventory (b) Integrated radiative forcing (c) Integrated radiative response

2018 2018 2018

(d) Energy inventory components (e) Radiative forcing components (f ) Energy budget 1971–2018

–
–

2018 2018

Figure TS.13 | Estimates of the net cumulative energy change (ZJ = 1021 Joules) for the period 1971–2018 associated with (a) observations of changes in the global
energy inventory, (b) integrated radiative forcing, and (c) integrated radiative response. The intent is to show assessed changes in energy budget and effective radiative forcings
(ERFs). Black dotted lines indicate the central estimate with likely and very likely ranges as indicated in the legend. The grey dotted lines indicate the energy change associated with an estimated
1850–1900 Earth energy imbalance of 0.2 W m–2 (panel a) and an illustration of an assumed pattern effect of –0.5 W m–2 °C–1 (panel c). Background grey lines indicate equivalent heating
rates in W m–2 per unit area of Earth’s surface. Panels (d) and (e) show the breakdown of components, as indicated in the legend, for the global energy inventory and integrated radiative
forcing, respectively. Panel (f) shows the global energy budget assessed for the period 1971–2018, that is, the consistency between the change in the global energy inventory relative to
1850–1900 and the implied energy change from integrated radiative forcing plus integrated radiative response under a number of different assumptions, as indicated in the figure legend,
including assumptions of correlated and uncorrelated uncertainties in forcing plus response. Shading represents the very likely range for observed energy change relative to 1850–1900 and
likely range for all other quantities. Forcing and response time series are expressed relative to a baseline period of 1850–1900. {Box 7.2, Figure 1}

90
 Technical Summary

by ocean warming. To put these numbers in context, the estimates of the global energy imbalance, and closure of the global
2006–2018 average Earth energy imbalance is equivalent sea level budget have led to a strengthened assessment relative
to approximately 20 times the annual rate of global energy to AR5. (high confidence) {7.2.2, 7.5.2.3, Box 7.2, Table 7.1, 9.6.1,
consumption in 2018. The accumulation of energy is driven Cross-Chapter Box 9.1, Table 9.5}
by a positive total anthropogenic effective radiative forcing
(ERF) relative to 1750. As in AR5, the perturbations to Earth’s top-of-atmosphere energy
budget are quantified using ERFs (see also Section TS.2.2). These
The best estimate ERF of 2.72 W m−2 has increased by 0.43 include any consequent adjustments to the climate system (e.g.,
W m–2 relative to that given in AR5 (for 1750–2014) due from changes in atmospheric temperatures, clouds and water vapour
to an increase in the greenhouse gas ERF that is partly as shown in Figure TS.14), but exclude any surface temperature
compensated by a more negative aerosol ERF compared response. Since AR5, ERFs have been estimated for a larger number
to AR5. The greenhouse gas ERF has been revised due to of forcing agents and shown to be more closely related to the
changes in atmospheric concentrations and updates to temperature response than the stratospheric-temperature-adjusted
forcing efficiencies, while the revision to aerosol ERF is due radiative forcing. (high confidence) {7.3.1}
to increased understanding of aerosol–cloud interactions TS
and is supported by improved agreement between Improved quantifications of ERF, the climate system radiative
different lines of evidence. Improved quantifications response, and the observed energy increase in the Earth system
of ERF, the climate system radiative response, and the for the period 1971–2018 demonstrate improved closure of the
observed energy increase in the Earth system for the global energy budget relative to AR5 (Figure TS.13). Combining
period 1971–2018 demonstrate improved closure of the the likely range of ERF over this period with the central estimate
global energy budget (i.e., the extent to which the sum of radiative response gives an expected energy gain of 340 [47 to
of the integrated forcing and the integrated radiative 662] ZJ. Both estimates are consistent with an independent
response equals the energy gain of the Earth system) observation-based assessment of the global energy increase of
compared to AR5 (high confidence). (See FAQ 7.1). {7.2.2, 284 [96 to 471] ZJ (very likely range), expressed relative to the
7.3.5, 7.5.2, Box 7.2, Table 7.1} estimated 1850–1900 Earth energy imbalance. (high confidence)
{7.2.2, 7.3.5, Box 7.2}
The global energy inventory change for the period 1971–2006
corresponds to an Earth energy imbalance (Box TS.1) of 0.50 [0.32 The assessed greenhouse gas ERF over the 1750–2019 period
to 0.69] W m–2, increasing to 0.79 [0.52 to 1.06] W m–2 for the period (Section TS.2.2) has increased by +0.59 W m−2 over AR5 estimates
2006–2018. Ocean heat uptake is by far the largest contribution and for 1750–2011. This increase includes +0.34 W m–2 from increases in
accounts for 91% of the total energy change. Land warming, melting atmospheric concentrations of well-mixed greenhouse gases (including
of ice and warming of the atmosphere account for about 5%, 3% and halogenated species) since 2011, +0.15 W m –2 from upwards revisions
1% of the total change, respectively. More comprehensive analysis of of their radiative efficiencies and +0.10 W m –2 from re-evaluation of
inventory components, cross-validation of satellite and in situ-based the ozone and stratospheric water vapour ERF. {7.3.2, 7.3.4, 7.3.5}

Atmospheric energy budget Instantaneous changes Adjustments Feedback responses

perturbation to radiative balance to meteorology and composition
Baseline 1 2 3
Adjustments in Biogeochemical O
chemical composition O
feedback O O
O O
Higher clouds

Absorbed Cloud
sunlight feedback
Adjustments
in clouds, water
vapour, temperature
fewer low clouds
Longwave Increased greenhouse Increased greenhouse Increased greenhouse
radiative gases or aerosols gases or aerosols gases or aerosols
emission Surface and Surface albedo and
vegetation biogeochemical
response feedback

Increased
water
vapour
No change in meteorology or surface No change in surface temperature Increased surface temperature
temperature

Figure TS.14 | Schematic representation of changes in the top-of-atmosphere (TOA) radiation budget following a perturbation. The intent of this figure is to
illustrate the concept of adjustments in the climate system following a perturbation in the radiation budget. The baseline TOA energy budget (a) responds instantaneously to
perturbations (b), leading to adjustments in the atmospheric meteorology and composition and land surface that are independent of changes in surface temperature (c). Surface
temperature changes (here using an increase as an example) lead to physical, biogeophysical and biogeochemical feedback processes (d). Long-term feedback processes, such
as those involving ice sheets, are not shown here. {adapted from Figure 7.2; FAQ 7.2, Figure 1; and Figure 8.3}

91
 Technical Summary

(a) Effective radiative forcing (b) Change in global surface temperature

1750 to 2019 1750 to 2019
Emitted Components

Climate effect through:

TS

(c) Aerosol effective radiative forcing

Figure TS.15 | Contribution to (a) effective radiative forcing (ERF) and (b) global surface temperature change from component emissions for 1750–2019
based on Coupled Model Intercomparison Project Phase 6 (CMIP6) models and (c) net aerosol ERF for 1750–2014 from different lines of evidence. The
intent of this figure is to show advances since AR5 in the understanding of (a) emissions-based ERF, (b) global surface temperature response for short-lived climate forcers as
estimated in Chapter 6, and (c) aerosol ERF from different lines of evidence as assessed in Chapter 7. In panel (a), ERFs for well-mixed greenhouse gases (WMGHGs) are
from the analytical formulae. ERFs for other components are multi-model means based on Earth system model simulations that quantify the effect of individual components.
The derived emissions-based ERFs are rescaled to match the concentration-based ERFs in Figure 7.6. Error bars are 5–95% and for the ERF account for uncertainty in radiative
efficiencies and multi-model error in the means. In panel (b), the global mean temperature response is calculated from the ERF time series using an impulse response function.
In panel (c), the AR6 assessment is based on energy balance constraints, observational evidence from satellite retrievals, and climate model-based evidence. For each line of
evidence, the assessed best-estimate contributions from ERF due to aerosol–radiation interactions (ERFari) and aerosol–cloud interactions (ERFaci) are shown with darker and
paler shading, respectively. Estimates from individual CMIP Phase 5 (CMIP5) and CMIP6 models are depicted by blue and red crosses, respectively. The observational assessment
for ERFari is taken from the instantaneous forcing due to aerosol–radiation interactions (IRFari). Uncertainty ranges are given in black bars for the total aerosol ERF and depict
very likely ranges. {6.4.2, Figure 6.12, 7.3.3, Cross-Chapter Box 7.1, Table 7.8, Figure 7.5}

For CO2, CH4, N2O, and chlorofluorocarbons, there is now evidence to emissions make the dominant contribution to the ERF from aerosol–
quantify the effect on ERF of tropospheric adjustments. The assessed cloud interactions (high confidence). Over the 1750–2019 period,
ERF for a doubling of CO2 compared to 1750 levels (3.9 ± 0.5 Wm–2) is the contributions from the emitted compounds to global surface
larger than in AR5. For CO2, the adjustments include the physiological temperature changes broadly match their contributions to the ERF
effects on vegetation. The reactive well-mixed greenhouse gases (high confidence) (Figure TS.15b). Since a peak in emissions-induced
(CH4, N2O, and halocarbons) cause additional chemical adjustments SO2 ERF has already occurred recently (Section TS.2.2) and since there
to the atmosphere through changes in ozone and aerosols (Figure is a delay in the full global surface temperature response owing to
TS.15a). The ERF due to CH4 emissions is 1.19 [0.81 to 1.58] W m–2, of the thermal inertia in the climate system, changes in SO2 emissions
which 0.35 [0.16 to 0.54] W m–2 is attributed to chemical adjustments have a slightly larger contribution to global surface temperature
mainly via ozone. These chemical adjustments also affect the change compared with changes in CO2 emissions, relative to their
emissions metrics (Section TS.3.3.3). Changes in sulphur dioxide (SO2) respective contributions to ERF. {6.4.2, 7.3.2}

92
 Technical Summary

Aerosols contributed an ERF of –1.3 [–2.0 to –0.6] W m–2 over the evolves and global surface temperature increases, leading to an ECS
period 1750 to 2014 (medium confidence). The ERF due to aerosol– that is higher than was inferred in AR5 based on warming over the
cloud interactions (ERFaci) contributes most to the magnitude of instrumental record (high confidence). Historical surface temperature
the total aerosol ERF (high confidence) and is assessed to be –1.0 change since 1870 has shown relatively little warming in several key
[–1.7 to –0.3] W m–2 (medium confidence), with the remainder due regions of positive feedbacks, including the eastern equatorial Pacific
to aerosol–radiation interactions (ERFari), assessed to be –0.3 [–0.6 Ocean and the Southern Ocean, while showing greater warming
to 0.0] W m–2 (medium confidence). There has been an increase in in key regions of negative feedbacks, including the western Pacific
the estimated magnitude – but a reduction in the uncertainty – of warm pool. Based on process understanding, climate modelling, and
the total aerosol ERF relative to AR5, supported by a combination paleoclimate reconstructions of past warm periods, it is expected
of increased process-understanding and progress in modelling and that future warming will become enhanced over the eastern Pacific
observational analyses (Figure TS.15c). Effective radiative forcing Ocean (medium confidence) and Southern Ocean (high confidence)
estimates from these separate lines of evidence are now consistent on centennial time scales. This new understanding, along with
with each other, in contrast to AR5, and support the assessment that updated estimates of historical temperature change, ERF, and energy
it is virtually certain that the total aerosol ERF is negative. Compared imbalance, reconciles previously disparate ECS estimates. {7.4.4,
to AR5, the assessed magnitude of ERFaci has increased, while that 7.5.2, 7.5.3} TS
of ERFari has decreased. {7.3.3, 7.3.5}
The AR6 best estimate of ECS is 3°C, the likely range is 2.5°C to
4°C and the very likely range is 2°C to 5°C. There is a high level
TS.3.2 Climate Sensitivity and Earth System Feedbacks of agreement among the four main lines of evidence listed above
(Figure TS.16b), and altogether it is virtually certain that ECS is larger
TS.3.2.1 Equilibrium Climate Sensitivity, Transient Climate than 1.5°C, but currently it is not possible to rule out ECS values
Response, and Transient Climate Response to above 5°C. Therefore, the 5°C upper end of the very likely range is
Cumulative Carbon-dioxide Emissions assessed with medium confidence and the other bounds with high
confidence. {7.5.5}
Since AR5, substantial quantitative progress has been
made in combining new evidence of Earth’s climate Based on process understanding, warming over the instrumental
sensitivity with improvements in the understanding and record, and emergent constraints, the best estimate of TCR is 1.8°C,
quantification of Earth’s energy imbalance, the instrumental the likely range is 1.4°C to 2.2°C and the very likely range is 1.2°C to
record of global surface temperature change, paleoclimate 2.4°C. There is a high level of agreement among the different lines of
change from proxy records, climate feedbacks and their evidence (Figure TS.16c) (high confidence). {7.5.5}
dependence on time scale and climate state. A key advance
is the broad agreement across these multiple lines of On average, CMIP6 models have higher mean ECS and TCR values
evidence, supporting a best estimate of equilibrium climate than the CMIP5 generation of models and also have higher mean
sensitivity of 3°C, with a very likely range of 2°C to 5°C. The values and wider spreads than the assessed best estimates and very
likely range of 2.5°C to 4°C is narrower than the AR5 likely likely ranges within this Report. These higher mean ECS and TCR
range of 1.5°C to 4.5°C. {7.4, 7.5} values can be traced to a positive net cloud feedback that is larger in
CMIP6 by about 20%. The broader ECS and TCR ranges from CMIP6
Constraints on equilibrium climate sensitivity (ECS) and transient also lead the models to project a range of future warming that
climate response (TCR) (see Glossary) are based on four main lines is wider than the assessed future warming range, which is based
of evidence: feedback process understanding, climate change and on multiple lines of evidence (Cross-Section Box TS.1). However,
variability seen within the instrumental record, paleoclimate evidence, some of the high-sensitivity CMIP6 models (Section TS.1.2.2) are
and so-called ‘emergent constraints’, whereby a relationship less consistent with observed recent changes in global warming
between an observable quantity and either ECS or TCR established and with paleoclimate proxy records than models with ECS within
within an ensemble of models is combined with observations to the very likely range. Similarly, some of the low-sensitivity models
derive a constraint on ECS or TCR. In reports up to and including are less consistent with the paleoclimate data. The CMIP6 models
the IPCC Third Assessment Report, ECS and TCR derived directly with the highest ECS and TCRs values provide insights into low-
from ESMs were the primary line of evidence. However, since AR4, likelihood, high-impact futures, which cannot be excluded based on
historical warming and paleoclimates provided useful additional currently available evidence (Cross-Section Box TS.1). {4.3.1, 4.3.4,
evidence (Figure TS.16a). This Report differs from previous reports in 7.4.2, 7.5.6}
not directly using climate model estimates of ECS and TCR in the
assessed ranges of climate sensitivity. {1.5, 7.5} Uncertainties regarding the true value of ECS and TCR are the dominant
source of uncertainty in global temperature projections over the
It is now clear that when estimating ECS and TCR, the dependence 21st century under moderate to high GHG concentrations scenarios.
of feedbacks on time scales and the climate state must be accounted For scenarios that reach net zero CO2 emissions (Section TS.3.3), the
for. Feedback processes are expected to become more positive uncertainty in the ERF values of aerosol and other SLCFs contribute
overall (more amplifying of global surface temperature changes) on substantial uncertainty in projected temperature. Global ocean heat
multi-decadal time scales as the spatial pattern of surface warming uptake is a smaller source of uncertainty in centennial warming. {7.5.7}

93
 Technical Summary

(a) Evolution of equilibrium climate sensitivity assessments from Charney to AR6

p < 10%
6 6

Equilibrium climate sensitivity (°C)

5 Very likely: 2–5°C 5

SAR

AR4

AR5
FAR
Charney

TAR
4 Likely: 2.5–4°C 4

Likely: p > 66%

AR6
3 Best estimate: 3°C 3

2 2
AR6 combines evidence from:
· Process understanding
· Instrumental record
1 Primarily model evidence p < 5%
·
· Paleoclimates
1
Emergent constraints
Also considers instrumental record and paleoclimates
TS
1980 1990 2000 2010 2020 2030
Year of assessment

(b) Equilibrium climate sensitivity (ºC) assessed in AR6 and (c) Transient climate response (ºC) assessed in AR6 and
simulated by CMIP6 ESMs simulated by CMIP6 ESMs

Process understanding Process understanding

Instrumental record Instrumental record

Paleoclimates Paleoclimates

Emergent constraints Emergent constraints

Combined assessment Combined assessment

CMIP6 ESMs CMIP6 ESMs

0 1 2 3 4 5 6 7 8 9 0 1 2 3 4
Best estimate range or value Likely range or limit Very likely range or limit Extremely likely limit

Figure TS.16 | (a) Evolution of equilibrium climate sensitivity (ECS) assessments from the Charney Report through a succession of IPCC Assessment
Reports to AR6, and lines of evidence and combined assessment for (b) ECS and (c) transient climate response (TCR) in AR6. The intent of this figure is to
show the progression in estimates of ECS, including uncertainty and the lines of evidence used for assessment, and to show the lines of assessment used to assess ECS and
TCR in AR6. In panel (a), the lines of evidence considered are listed below each assessment. Best estimates are marked by horizontal bars, likely ranges by vertical bars, and
very likely ranges by dotted vertical bars. In panel (b) and (c), assessed ranges are taken from Tables 7.13 and 7.14 for ECS and TCR respectively. Note that for the ECS
assessment based on both the instrumental record and paleoclimates, limits (i.e., one-sided distributions) are given, which have twice the probability of being outside the
maximum/minimum value at a given end, compared to ranges (i.e., two tailed distributions) which are given for the other lines of evidence. For example, the extremely likely limit
of greater than 95% probability corresponds to one side of the very likely (5% to 95%) range. Best estimates are given as either a single number or by a range represented by
grey box. Coupled Model Intercomparison Project Phase 6 (CMIP6) Earth system model (ESM) values are not directly used as a line of evidence but are presented on the figure
for comparison. {1.5, 7.5; Tables 7.13 and 7.14; Figure 7.18}

The transient climate response to cumulative CO2 emissions (TCRE) of TCR. Beyond this century, there is low confidence that the TCRE
is the ratio between globally averaged surface temperature increase alone remains an accurate predictor of temperature changes in
and cumulative CO2 emissions (see Glossary). This Report reaffirms scenarios of very low or net negative CO2 emissions because of
with high confidence the finding of AR5 that there is a near-linear uncertain Earth system feedbacks that can result in further changes
relationship between cumulative CO2 emissions and the increase in in temperature or a path dependency of warming as a function of
global average temperature caused by CO2 over the course of this cumulative CO2 emissions. {4.6.2, 5.4, 5.5.1}
century for global warming levels up to at least 2°C relative to 1850–
1900. The TCRE falls likely in the 1.0°C–2.3°C per 1000 PgC range, TS.3.2.2 Earth System Feedbacks
with a best estimate of 1.65°C per 1000 PgC. This is equivalent to a
0.27°C–0.63°C range with a best estimate of 0.45°C when expressed The combined effect of all climate feedback processes is to amplify
in units per 1000 GtCO2. This range is about 15% narrower than the the climate response to forcing (virtually certain). While major
0.8°–2.5°C per 1000 PgC assessment of AR5 because of a better advances in the understanding of cloud processes have increased
integration of evidence across chapters, in particular the assessment the level of confidence and decreased the uncertainty range for the

94
 Technical Summary

cloud feedback by about 50% compared to AR5, clouds remain the Natural sources and sinks of non-CO2 greenhouse gases such as
largest contribution to overall uncertainty in climate feedbacks (high methane (CH4) and nitrous oxide (N2O) respond both directly and
confidence). Uncertainties in the ECS and other climate sensitivity indirectly to atmospheric CO2 concentration and climate change,
metrics, such as the TCR and TCRE, are the dominant source of and thereby give rise to additional biogeochemical feedbacks
uncertainty in global temperature projections over the 21st century in the climate system. Many of these feedbacks are only partially
under moderate to high GHG emissions scenarios. CMIP6 models understood and are not yet fully included in ESMs. There is medium
have higher mean values and wider spreads in ECS and TCR than the confidence that the net response of natural ocean and land CH4 and
assessed best estimates and very likely ranges within this Report, N2O sources to future warming will be increased emissions, but the
leading the models to project a range of future warming that is magnitude and timing of the responses of each individual process is
wider than the assessed future warming range (Section TS.2.2). {7.1, known with low confidence. {5.4.7}
7.4.2, 7.5}
Non-CO2 biogeochemical feedbacks induced from changes in
Earth system feedbacks can be categorized into three broad groups: emissions, abundances or lifetimes of SLCFs mediated by natural
physical feedbacks, biogeophysical and biogeochemical feedbacks, processes or atmospheric chemistry are assessed to decrease
and feedbacks associated with ice sheets. In previous assessments, ECS (Figure TS.17b). These non-CO2 biogeochemical feedbacks TS
the ECS has been associated with a distinct set of physical feedbacks are estimated from ESMs, which since AR5 have advanced to
(Planck response, water vapour, lapse rate, surface albedo, and cloud include a consistent representation of biogeochemical cycles and
feedbacks). In this assessment, a more general definition of ECS is atmospheric chemistry. However, process-level understanding
adopted whereby all biogeophysical and biogeochemical feedbacks of many biogeochemical feedbacks involving SLCFs, particularly
that do not affect the atmospheric concentration of CO2 are included. natural emissions, is still emerging, resulting in low confidence in
These include changes in natural CH4 emissions, natural aerosol the magnitude and sign of the feedbacks. The central estimate of
emissions, N2O, ozone, and vegetation, which all act on time scales of the total biogeophysical and non-CO2 biogeochemical feedback is
years to decades and are therefore relevant for temperature change assessed to be −0.01 [–0.27 to +0.25] W m–2 °C–1 (Figure TS.17a).
over the 21st century. Because the total biogeophysical and non-CO2 {5.4.7, 5.4.8, 6.2.2, 6.4.5, 7.4, Table 7.10}
biogeochemical feedback is assessed to have a central value that is
near zero (low confidence), including it does not affect the assessed The combined effect of all known radiative feedbacks (physical,
ECS but does contribute to the net feedback uncertainty. The biogeophysical, and non-CO2 biogeochemical) is to amplify the
biogeochemical feedbacks that affect the atmospheric concentration base climate response (in the absence of feedbacks), also known
of CO2 are not included because ECS is defined as the response to as the Planck temperature response20 (virtually certain). Combining
a sustained doubling of CO2. Moreover, the long-term feedbacks these feedbacks with the Planck response, the net climate feedback
associated with ice sheets are not included in the ECS owing to their parameter is assessed to be –1.16 [–1.81 to –0.51] W m–2 °C–1,
long time scales of adjustment. {5.4, 6.4, 7.4, 7.5, Box 7.1} which is slightly less negative than that inferred from the overall ECS
assessment. The combined water vapour and lapse rate feedback
The net effect of changes in clouds in response to global warming is makes the largest single contribution to global warming, whereas
to amplify human-induced warming, that is, the net cloud feedback the cloud feedback remains the largest contribution to overall
is positive (high confidence). Compared to AR5, major advances uncertainty. Due to the state-dependence of feedbacks, as evidenced
in the understanding of cloud processes have increased the level from paleoclimate observations and from models, the net feedback
of confidence and decreased the uncertainty range in the cloud parameter will increase (become less negative) as global temperature
feedback by about 50% (Figure TS.17a). An assessment of the low- increases. Furthermore, on long time scales the ice-sheet feedback
altitude cloud feedback over the subtropical ocean, which was parameter is very likely positive, promoting additional warming on
previously the major source of uncertainty in the net cloud feedback, millennial time scales as ice sheets come into equilibrium with the
is improved owing to a combined use of climate model simulations, forcing. (high confidence) {7.4.2, 7.4.3, Figure 7.14, Table 7.10}
satellite observations, and explicit simulations of clouds, altogether
leading to strong evidence that this type of cloud amplifies global The carbon cycle provides for additional feedbacks on climate owing
warming. The net cloud feedback is assessed to be +0.42 [–0.10 to the sensitivity of land–atmosphere and ocean–atmosphere carbon
to 0.94] W m –2 °C–1. A net negative cloud feedback is very unlikely. fluxes and storage to changes in climate and in atmospheric CO2
The CMIP5 and CMIP6 ranges of cloud feedback are similar to this (Figure TS.17c). Because of the time scales associated with land and
assessed range, with CMIP6 having a slightly more positive median ocean carbon uptake, these feedbacks are known to be scenario
cloud feedback (high confidence). The surface albedo feedback and dependent. Feedback estimates deviate from linearity in scenarios
combined water vapour-lapse rate feedback are positive (Figure of stabilizing or reducing concentrations. With high confidence,
TS.17a), with high confidence in the estimated value of each based increased atmospheric CO2 will lead to increased land and ocean
on multiple lines of evidence, including observations, models and carbon uptake, acting as a negative feedback on climate change. It
theory (Box TS.6). {7.4.2, Figure 7.14, Table 7.10} is likely that a warmer climate will lead to reduced land and ocean
carbon uptake, acting as a positive feedback (Box TS.5). {4.3.2,
5.4.1–5}

20 For reference, the Planck temperature response for a doubling of atmospheric CO2 is approximately 1.2°C at equilibrium.

95
 Technical Summary

(a) Feedbacks in the climate system

Negative feedbacks diminish the Positive feedbacks amplify the
initial climate response to radiative forcing initial climate response to radiative forcing
Mean [very likely range]

Total −1.16 [−1.81 to −0.51]

Planck −3.22 [−3.39 to −3.05]

Water vapour and lapse rate 1.30 [ 1.13 to 1.47]

Surface albedo 0.35 [ 0.10 to 0.60]

Clouds 0.42 [−0.10 to 0.94]

Biogeophysical and −0.01 [−0.27 to 0.25]

non−CO2 biogeochemical
(Total from panel (b))
TS −3.5 −3.0 −2.5 −2.0 −1.5 −1.0 −0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
Climate feedback parameter (Wm-2 °C-1)

(b) Biogeophysical and non-CO2 biogeochemical climate feedbacks Mean [5−95% range]

Total (without permafrost feedback) −0.01 [−0.27 to 0.25]

Permafrost feedback
CH4 source response to climate 0.03 [ 0.01 to 0.05]

Atm. CH4 lifetime response to climate −0.03 [−0.12 to 0.06]

N2O source response to climate 0.01 [−0.01 to 0.02]

Other non−CO2 biogeochemical −0.17 [−0.36 to 0.02]

Biogeophysical 0.15 [ 0.00 to 0.30]

−0.4 −0.3 −0.2 −0.1 0.0 0.1 0.2 0.3 0.4

Climate feedback parameter (Wm-2 °C-1)

(c) Carbon-cycle climate feedbacks Mean [5−95% range]

Land carbon response to CO2 −0.78 [−1.28 to −0.28]

Ocean carbon response to CO2 Permafrost feedback −0.68 [−0.98 to −0.39]

Land carbon response to climate 0.25 [−0.03 to 0.54]

Ocean carbon response to climate 0.08 [0.04 to 0.12]

−3.5 −3.0 −2.5 −2.0 −1.5 −1.0 −0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
Climate feedback parameter (Wm-2 °C-1)

Figure TS.17 | An overview of physical and biogeochemical feedbacks in the climate system. The intent of this figure is to summarize assessed estimates of physical,
biogeophysical and biogeochemical feedbacks on global temperature based on Chapters 5, 6 and 7. (a) Synthesis of physical, biogeophysical and non-carbon dioxide (CO2)
biogeochemical feedbacks that are included in the definition of equilibrium climate sensitivity (ECS) assessed in this Technical Summary. These feedbacks have been assessed
using multiple lines of evidence including observations, models and theory. The net feedback is the sum of the Planck response, water vapour and lapse rate, surface albedo,
cloud, and biogeophysical and non-CO2 biogeochemical feedbacks. Bars denote the mean feedback values, and uncertainties represent very likely ranges; (b) Estimated values of
individual biogeophysical and non-CO2 biogeochemical feedbacks. The atmospheric methane (CH4) lifetime and other non-CO2 biogeochemical feedbacks have been calculated
using global Earth system model simulations from AerChemMIP, while the CH4 and nitrous oxide (N2O) source responses to climate have been assessed for the year 2100 using
a range of modelling approaches using simplified radiative forcing equations. The estimates represent the mean and 5–95% range. The level of confidence in these estimates
is low owing to the large model spread. (c) Carbon-cycle feedbacks as simulated by models participating in the C4MIP of the Coupled Model Intercomparison Project Phase 6
(CMIP6). An independent estimate of the additional positive carbon-cycle climate feedbacks from permafrost thaw, which is not considered in most C4MIP models, is added.
The estimates represent the mean and 5–95% range. Note that these feedbacks act through modifying the atmospheric concentration of CO2 and thus are not included in the
definition of ECS, which assumes a doubling of CO2, but are included in the definition and assessed range of the transient climate response to cumulative CO2 emissions (TCRE).
{5.4.7, 5.4.8, Box 5.1, Figure 5.29, 6.4.5, Table 6.9, 7.4.2, Table 7.10}

96
 Technical Summary

Thawing terrestrial permafrost will lead to carbon release (high TS.3.3 Temperature Stabilization, Net Zero Emissions
confidence), but there is low confidence in the timing, magnitude and Mitigation
and the relative roles of CO2 versus CH4 as feedback processes. An
ensemble of models projects CO2 release from permafrost to be TS.3.3.1 Remaining Carbon Budgets and Temperature
3–41 PgC per 1ºC of global warming by 2100, leading to warming Stabilization
strong enough that it must be included in estimates of the remaining
carbon budget but weaker than the warming from fossil fuel burning. The near-linear relationship between cumulative CO2
However, the incomplete representation of important processes, emissions and maximum global surface temperature increase
such as abrupt thaw, combined with weak observational constraints, caused by CO2 implies that stabilizing human-induced global
only allow low confidence in both the magnitude of these estimates temperature increase at any level requires net anthropogenic
and in how linearly proportional this feedback is to the amount of CO2 emissions to become zero. This near-linear relationship
global warming. There is emerging evidence that permafrost thaw further implies that mitigation requirements for limiting
and thermokarst give rise to increased CH4 and N2O emissions, which warming to specific levels can be quantified in terms of a
leads to the combined radiative forcing from permafrost thaw being carbon budget (high confidence). Remaining carbon budget
larger than from CO2 emissions only. However, the quantitative estimates have been updated since AR5 with methodological TS
understanding of these additional feedbacks is low, particularly for improvements, resulting in larger estimates that are
N2O. These feedbacks, as well as potential additional carbon losses consistent with SR1.5. Several factors, including estimates
due to climate-induced fire feedback are not routinely included in of historical warming, future emissions from thawing
Earth system models. {Box 5.1, 5.4.3, 5.4.7, 5.4.8} permafrost, variations in projected non-CO2 warming, and
the global surface temperature change after cessation of
CO2 emissions, affect the exact value of carbon budgets (high
confidence). {1.3.5, Box 1.2, 4.7.1, 5.5}

(a) Cumulative carbon emissions since 1850 (PgC)

(b)
0 200 400 600 800 1000 1200 1400 1600 1800 2000
7
Global mean temperature increase since 1850–1900 (°C)

Historical observations TCRE

Human-induced warming estimate global warming limit of interest (Section 5.5.1.4)
Temperature increase since pre-industrial levels (°C)

6 Human-induced warming assessment zero emissions commitment (ZEC, Section 5.5.2.2.4)

Assessed TCRE range
non-CO2 contribution (Section 5.5.2.2.3)
5 Scenarios:
SSP1-1.9
SSP1-2.6
4 SSP2-4.5
allowable warming

SSP3-7.0
SSP5-8.5
remaining

1
historical
0 human-induced
warming
0 (Section 5.5.2.2.2)
-1
0 1000 2000 3000 4000 5000 6000 7000 0 remaining unrepresented Earth system
carbon budget feedbacks (Section 5.5.2.2.5)
Cumulative carbon dioxide emissions since 1850 (GtCO2)
Cumulative CO2 emissions from today (GtCO2)

Figure TS.18 | Illustration of (a) relationship between cumulative emissions of carbon dioxide (CO2) and global mean surface air temperature increase
and (b) the assessment of the remaining carbon budget from its constituting components based on multiple lines of evidence. The intent of this figure is to
show (i) the proportionality between cumulative CO2 emissions and global surface air temperature in observations and models as well as the assessed range of the transient
climate response to cumulative CO2 emissions (TCRE), and (ii) how information is combined to derive remaining carbon budgets consistent with limiting warming to a specific
level. Carbon budgets consistent with various levels of additional warming are provided in Table 5.8 and should not be read from the illustrations in either panel. In panel (a)
thin black line shows historical CO2 emissions together with the assessed global surface temperature increase from 1850–1900 as assessed in Chapter 2 (Box 2.3). The orange-
brown range with its central line shows the estimated human-induced share of historical warming. The vertical orange-brown line shows the assessed range of historical human-
induced warming for the 2010–2019 period relative to 1850–1900 (Chapter 3). The grey cone shows the assessed likely range for the TCRE (Section 5.5.1.4), starting from
2015. Thin coloured lines show Coupled Model Intercomparison Project Phase 6 (CMIP6) simulations for the five scenarios of the WGI core set (SSP1-1.9, light blue; SSP1-2.6,
blue; SSP2-4.5, yellow; SSP3-7.0, red; SSP5-8.5, maroon), starting from 2015 and until 2100. Diagnosed carbon emissions are complemented with estimated land-use change
emissions for each respective scenario. Coloured areas show the Chapter 4 assessed very likely range of global surface temperature projections and thick coloured central lines
show the median estimate, for each respective scenario. These temperature projections are expressed relative to cumulative CO2 emissions that are available for emissions-
driven CMIP6 ScenarioMIP experiments for each respective scenario. For panel (b), the remaining allowable warming is estimated by combining the global warming limit
of interest with the assessed historical human-induced warming (Section 5.5.2.2.2), the assessed future potential non-CO2 warming contribution (Section 5.5.2.2.3) and the
zero emissions commitment (ZEC; Section 5.5.2.2.4). The remaining allowable warming (vertical blue bar) is subsequently combined with the assessed TCRE (Sections 5.5.1.4
and 5.5.2.2.1) and contribution of unrepresented Earth system feedbacks (Section 5.5.2.2.5) to provide an assessed estimate of the remaining carbon budget (horizontal blue
bar, Table 5.8). Note that contributions in panel (b) are illustrative and are not to scale. For example, the central ZEC estimate was assessed to be zero. {Box 2.3, 5.2.1, 5.2.2,
Figure 5.31}

97
 Technical Summary

Limiting further climate change would require substantial and but they may vary by an estimated ±220 GtCO2 depending on
sustained reductions of GHG emissions. Without net zero CO2 how deeply future non-CO2 emissions are assumed to be reduced
emissions, and a decrease in the net non-CO2 forcing (or sufficient (Table TS.3). {5.5.2, 5.6, Box 5.2, 7.6}
net negative CO2 emissions to offset any further warming from
net non-CO2 forcing), the climate system will continue to warm. There is high confidence that several factors, including estimates of
There is high confidence that mitigation requirements for limiting historical warming, future emissions from thawing permafrost, and
warming to specific levels over this century can be estimated using variations in projected non-CO2 warming, affect the value of carbon
a carbon budget that relates cumulative CO2 emissions to global budgets but do not change the conclusion that global CO2 emissions
mean temperature increase (Figure TS.18, Table TS.3). For the period would need to decline to net zero to halt global warming. Estimates
1850–2019, a total of 2390 ± 240 GtCO2 of anthropogenic CO2 has may vary by ±220 GtCO2 depending on the level of non-CO2 emissions
been emitted. Remaining carbon budgets (starting from 1 January at the time global anthropogenic CO2 emissions reach net zero levels.
2020) for limiting warming to 1.5°C, 1.7°C and 2.0°C are estimated This variation is referred to as non-CO2 scenario uncertainty and
at 500 GtCO2, 850 GtCO2 and 1350 GtCO2, respectively, based on will be further assessed in the AR6 Working Group III Contribution.
the 50th percentile of TCRE. For the 67th percentile, the respective Geophysical uncertainties surrounding the climate response to these
TS values are 400 GtCO2, 700 GtCO2 and 1150 GtCO2. The remaining non-CO2 emissions result in an additional uncertainty of at least
carbon budget estimates for different temperature limits assume ±220 GtCO2, and uncertainties in the level of historical warming
that non-CO2 emissions are mitigated consistent with the median result in a ±550 GtCO2 uncertainty. {5.4, 5.5.2}
reductions found in scenarios in the literature as assessed in SR1.5,

Table TS.3 | Estimates of remaining carbon budgets and their uncertainties. Assessed estimates are provided for additional human-induced warming, expressed as
global surface temperature, since the recent past (2010–2019), likely amounted to 0.8° to 1.3°C with a best estimate of 1.07°C relative to 1850–1900. Historical CO2 emissions
between 1850 and 2014 have been estimated at about 2180 ± 240 GtCO2 (1-sigma range), while since 1 January 2015, an additional 210 GtCO2 has been emitted until the
end of 2019. GtCO2 values to the nearest 50. {Table 3.1, 5.5.1, 5.5.2, Box 5.2, Table 5.1, Table 5.7, Table 5.8}

Global Global Estimated remaining carbon

surface surface budgets starting from 1 January
Scenario
temperature temperature 2020 and subject to variations Geophysical uncertaintiesd
variation
change since change since and uncertainties quantified in the
2010–2019 1850–1900a columns on the right
Non-CO2 Zero CO2
Non-CO2 Historical Recent
Percentiles of TCREb forcing and emissions
°C °C scenario temperature emissions
GtCO2 response commitment
variationc uncertaintya uncertaintye
uncertainty uncertainty

17th 33rd 50th 67th 83rd GtCO2 GtCO2 GtCO2 GtCO2 GtCO2

0.43 1.5 900 650 500 400 300 Values can

Values can
vary by at least
0.53 1.6 1200 850 650 550 400 vary by at least
±220 due to
0.63 1.7 1450 1050 850 700 550 ±220 due to
uncertainty in
choices related ±550 ±420 ±20
0.73 1.8 1750 1250 1000 850 650 the warming
to non-CO2
response to
0.83 1.9 2000 1450 1200 1000 800 emissions
future non-CO2
mitigation
0.93 2 2300 1700 1350 1150 900 emissions

a
Human-induced global surface temperature increase between 1850–1900 and 2010–2019 is assessed at 0.8–1.3°C (likely range; Cross-Section Box TS.1) with a best estimate
of 1.07°C. Combined with a central estimate of TCRE (1.65°C per 1000 PgC) this uncertainty in isolation results in a potential variation of remaining carbon budgets of ±550
GtCO2, which, however, is not independent of the assessed uncertainty of TCRE and thus not fully additional.
b
TCRE: transient climate response to cumulative emissions of carbon dioxide, assessed to fall likely between 1.0–2.3°C per 1000 PgC with a normal distribution, from which
the percentiles are taken. Additional Earth system feedbacks are included in the remaining carbon budget estimates as discussed in Section 5.5.2.2.5.
c
Estimates assume that non-CO2 emissions are mitigated consistent with the median reductions found in scenarios in the literature as assessed in SR1.5. Non-CO2 scenario
variations indicate how much remaining carbon budget estimates vary due to different scenario assumptions related to the future evolution of non-CO2 emissions in mitigation
scenarios from SR1.5 that reach net zero CO2 emissions. This variation is additional to the uncertainty in TCRE. The Working Group III Contribution to AR6 will reassess the
potential for non-CO2 mitigation based on literature since SR1.5.
d
Geophysical uncertainties reported in these columns and TCRE uncertainty are not statistically independent, as uncertainty in TCRE depends on uncertainty in the assessment
of historical temperature, non-CO2 versus CO2 forcing, and uncertainty in emissions estimates. These estimates cannot be formally combined, and these uncertainty variations
are not directly additional to the spread of remaining carbon budgets due to TCRE uncertainty reported in columns three to seven.
e
Recent emissions uncertainty reflects the ±10% uncertainty in the historical CO2 emissions estimate since 1 January 2015.

98
 Technical Summary

Methodological improvements and new evidence result in updated to be removed to compensate for an emission of a given
remaining carbon budget estimates. The assessment in AR6 applies magnitude to attain the same change in atmospheric CO2
the same methodological improvements as in SR1.5, which uses (medium confidence). CDR methods have wide-ranging side-
a recent observed baseline for historic temperature change and effects that can either weaken or strengthen the carbon
cumulative emissions. Changes compared to SR1.5 are therefore sequestration and cooling potential of these methods and
small: the assessment of new evidence results in updated median affect the achievement of sustainable development goals
remaining carbon budget estimates for limiting warming to 1.5°C (high confidence). {4.6.3, 5.6}
and 2°C being the same and about 60 GtCO2 smaller, respectively,
after accounting for emissions since SR1.5. Meanwhile, remaining Carbon dioxide removal (CDR) refers to anthropogenic activities
carbon budgets for limiting warming to 1.5°C would be about that deliberately remove CO2 from the atmosphere and durably
300–350 GtCO2 larger if evidence and methods available at the store it in geological, terrestrial or ocean reservoirs, or in products.
time of AR5 would be used. If a specific remaining carbon budget Carbon dioxide is removed from the atmosphere by enhancing
is exceeded, this results in a lower probability of keeping warming biological or geochemical carbon sinks or by direct capture of CO2
below a specified temperature level and higher irreversible global from air. Emissions pathways that limit global warming to 1.5°C or
warming over decades to centuries, or alternatively a need for net 2°C typically assume the use of CDR approaches in combination TS
negative CO2 emissions or further reductions in non-CO2 greenhouse with GHG emissions reductions. CDR approaches could be used to
gases after net zero CO2 is achieved to return warming to lower levels compensate for residual emissions from sectors that are difficult or
in the long term. {5.5.2, 5.6, Box 5.2} costly to decarbonize. CDR could also be implemented at a large scale
to generate global net negative CO2 emissions (i.e., anthropogenic
Based on idealized model simulations that explore the climate CO2 removals exceeding anthropogenic emissions), which could
response once CO2 emissions have been brought to zero, the compensate for earlier emissions as a way to meet long-term climate
magnitude of the zero CO2 emissions commitment (ZEC, see Glossary) stabilization goals after a temperature overshoot. This Report assesses
is assessed to be likely smaller than 0.3°C for time scales of about the effects of CDR on the carbon cycle and climate. Co-benefits and
half a century and cumulative CO2 emissions broadly consistent with trade-offs for biodiversity, water and food production are briefly
global warming of 2°C. However, there is low confidence about its discussed for completeness, but a comprehensive assessment of the
sign on time scales of about half a century. For lower cumulative CO2 ecological and socio-economic dimensions of CDR options is left to
emissions, the range would be smaller yet with equal uncertainty the WGII and WGIII reports. {4.6.3, 5.6}
about the sign. If the ZEC is positive on decadal time scales, additional
warming leads to a reduction in the estimates of remaining carbon CDR methods have the potential to sequester CO2 from the
budgets, and vice versa if it is negative. {4.7.1, 5.5.2} atmosphere (high confidence). In the same way part of current
anthropogenic net CO2 emissions are taken up by land and ocean
Permafrost thaw is included in estimates together with other carbon stores, net CO2 removal will be partially counteracted by CO2
feedbacks that are often not captured by models. Limitations in release from these stores, such that the amount of CO2 sequestered
modelling studies combined with weak observational constraints by CDR will not result in an equivalent drop in atmospheric CO2 (very
only allow low confidence in the magnitude of these estimates high confidence). The fraction of CO2 removed from the atmosphere
(Section TS.3.2.2). Despite the large uncertainties surrounding the that is not replaced by CO2 released from carbon stores – a measure
quantification of the effect of additional Earth system feedback of CDR effectiveness – decreases slightly with increasing amounts of
processes, such as emissions from wetlands and permafrost thaw, removal (medium confidence) and decreases strongly if CDR is applied
these feedbacks represent identified additional risk factors that at lower atmospheric CO2 concentrations (medium confidence). The
scale with additional warming and mostly increase the challenge of reduction in global surface temperature is approximately linearly
limiting warming to specific temperature levels. These uncertainties related to cumulative CO2 removal (high confidence). Because of this
do not change the basic conclusion that global CO2 emissions would near-linear relationship, the amount of cooling per unit CO2 removed
need to decline to net zero to halt global warming. {5.4.8, 5.5.2, is approximately independent of the rate and amount of removal
Box 5.1} (medium confidence). {4.6.3, 5.6.2.1, Figure 5.32, Figure 5.34}

TS.3.3.2 Carbon Dioxide Removal Due to non-linearities in the climate system, the century-scale
climate–carbon cycle response to a CO2 removal from the atmosphere
Deliberate carbon dioxide removal (CDR) from the is not always equal and opposite to its response to a simultaneous
atmosphere has the potential to compensate for residual CO2 emission (medium confidence). For CO2 emissions of 100 PgC
CO2 emissions to reach net zero CO2 emissions or to released from a state in equilibrium with pre-industrial atmospheric
generate net negative CO2 emissions. In the same way CO2 levels, CMIP6 models simulate that 27± 6% (mean ± 1 standard
that part of current anthropogenic net CO2 emissions deviation) of emissions remain in the atmosphere 80–100 years
are taken up by land and ocean carbon stores, net CO2 after the emissions, whereas for removals of 100 PgC only 23
removal will be partially counteracted by CO2 release from ± 6% of removals remain out of the atmosphere. This asymmetry
these stores (very high confidence). Asymmetry in the implies that an extra amount of CDR is required to compensate for
carbon cycle response to simultaneous CO2 emissions and a positive emission of a given magnitude to attain the same change
removals implies that a larger amount of CO2 would need in atmospheric CO2. Due to low agreement between models, there

99
 Technical Summary

is low confidence in the sign of the asymmetry of the temperature Carbon dioxide removal methods have a range of side effects that
response to CO2 emissions and removals. {4.6.3, 5.6.2.1, Figure 5.35} can either weaken or strengthen the carbon sequestration and
cooling potential of these methods and affect the achievement
Simulations with ESMs indicate that under scenarios where CO2 of sustainable development goals (high confidence). Biophysical
emissions gradually decline, reach net zero and become net negative and biogeochemical side-effects of CDR methods are associated
during the 21st century (e.g., SSP1-2.6), land and ocean carbon sinks with changes in surface albedo, the water cycle, emissions of CH4
begin to weaken in response to declining atmospheric CO2 and N2O, ocean acidification and marine ecosystem productivity
concentrations, and the land sink eventually turns into a source (high confidence). These side-effects and associated Earth system
(Figure TS.19). This sink-to-source transition occurs decades to a few feedbacks can decrease carbon uptake and/or change local and
centuries after CO2 emissions become net negative. The ocean regional climate and in turn limit the CO2 sequestration and cooling
remains a sink of CO2 for centuries after emissions become net potential of specific CDR methods (medium confidence). Deployment
negative. Under scenarios with large net negative CO2 emissions of CDR, particularly on land, can also affect water quality and quantity,
(e.g., SSP5-3.4-OS) and rapidly declining CO2 concentrations, the land food production and biodiversity (high confidence). These effects are
source is larger than for SSP1-2.6 and the ocean also switches to a often highly dependent on local context, management regime, prior
TS source. While the general response is robust across models, there is land use, and scale (high confidence). The largest co-benefits are
low confidence in the timing of the sink-to-source transition and the obtained with methods that seek to restore natural ecosystems or
magnitude of the CO2 source in scenarios with net negative CO2 improve soil carbon sequestration (medium confidence). The climate
emissions. Carbon dioxide removal could reverse some aspects and biogeochemical effects of terminating CDR are expected to be
climate change if CO2 emissions become net negative, but some small for most CDR methods (medium confidence). {4.6.3, 5.6.2.2,
changes would continue in their current direction for decades to Figure 5.36, 8.4.3, 8.6.3}
millennia. For instance, sea level rise due to ocean thermal expansion
would not reverse for several centuries to millennia (high confidence)
(Box TS.4). {4.6.3, 5.4.10, 5.6.2.1, Figure 5.30, Figure 5.33}

Different emission stages in SSP1-2.6 scenario, characterized by:

(a) Large net positive (b) Small net positive (c) Net negative (d) Net negative No (e) Net zero
CO2 emissions CO2 emissions CO2 emissions CO2 emissions significant CO2 emissions
(emissions exceed (emissions exceed (removals exceed (removals exceed change and removals
removals) removals) emissions) emissions) within
this period
CO2 in atmosphere

Net land flux

550
Net ocean flux
Carbon dioxide concentration (ppm)

203 –41

500
–61 33 –23
–33
and removals

450
Net CO2 emissions

–30 2
4 3 0
400 469
446 –7 –5 –7
411 –7
403 400 396
350 368
+101 ppm –23 ppm –35 ppm –8 ppm –4 ppm
300
2000 2050 2100 2150 2200 2250 2300

Figure TS.19 | Carbon sink response in a scenario with net carbon dioxide (CO2) removal from the atmosphere. The intent of this figure is to show how
atmospheric CO2 evolves under negative emissions and its dependence on the negative emissions technologies. It also shows the evolution of the ocean and land sinks. Shown
are CO2 flux components from concentration-driven Earth system model (ESM) simulations during different emissions stages of SSP1–2.6 and its long-term extension. (a) Large
net positive CO2 emissions, (b) small net positive CO2 emissions, (c–d) net negative CO2 emissions, and (e) net zero CO2 emissions. Positive flux components act to raise the
atmospheric CO2 concentration, whereas negative components act to lower the CO2 concentration. Net CO2 emissions and land and ocean CO2 fluxes represent the multi-model
mean and standard deviation (error bar) of four ESMs (CanESM5, UKESM1, CESM2-WACCM, IPSL-CM6a-LR) and one Earth system model of intermediate complexity (Uvic
ESCM). Net CO2 emissions are calculated from concentration-driven ESM simulations as the residual from the rate of increase in atmospheric CO2 and land and ocean CO2 fluxes.
Fluxes are accumulated over each 50-year period and converted to concentration units (parts per million, or ppm). {5.6.2.1, Figure 5.33}

100
 Technical Summary

TS.3.3.3 Relating Different Forcing Agents future global surface temperature outcomes (high confidence) {7.6.1,
Box 7.3}
When including other GHGs, the choice of emissions metric
affects the quantification of net zero GHG emissions and Emissions metrics are needed to aggregate baskets of gases to
their resulting temperature outcome (high confidence). determine net zero GHG emissions. Generally, achieving net zero CO2
Reaching and sustaining net zero GHG emissions typically emissions and declining non-CO2 radiative forcing would halt human-
leads to a peak and decline in temperatures when induced warming. Reaching net zero GHG emissions quantified by
quantified with the global warming potential over a 100- GWP-100 typically leads to declining temperatures after net zero
year period (GWP-100). Carbon-cycle responses are more GHGs emissions are achieved if the basket includes short-lived gases,
robustly accounted for in emissions metrics compared to such as CH4. Net zero GHG emissions defined by CGTP or GWP*
AR5 (high confidence). New emissions metric approaches imply net zero CO2 and other long-lived GHG emissions and constant
can be used to generate equivalent cumulative emissions (CGTP) or gradually declining (GWP*) emissions of short-lived gases.
of CO2 for short-lived greenhouse gases based on their rate The warming evolution resulting from net zero GHG emissions
of emissions. {7.6.2} defined in this way corresponds approximately to reaching net zero
CO2 emissions, and would thus not lead to declining temperatures TS
Over 10- to 20-year time scales, the temperature response after net zero GHG emissions are achieved but to an approximate
to a single year’s worth of current emissions of short- temperature stabilization (high confidence). The choice of emissions
lived climate forcers (SLCFs) is at least as large as that of metric hence affects the quantification of net zero GHG emissions,
CO2, but because the effects of SLCFs decay rapidly over and therefore the resulting temperature outcome of reaching and
the first few decades after emission, the net long-term sustaining net zero GHG emissions levels (high confidence). {7.6.1.4,
temperature response to a single year’s worth of emissions 7.6.2, 7.6.3}
is predominantly determined by cumulative CO2 emissions.
As pointed out in AR5, ultimately, it is a matter for policymakers
Emissions reductions in 2020 associated with COVID-19 to decide which emissions metric is most applicable to their needs.
containment led to small and positive global ERF; however, This Report does not recommend the use of any specific emissions
global and regional climate responses to the forcing are metric, as the most appropriate metric depends on the policy goal
undetectable above internal variability due to the temporary and context (see Chapter 7, Section 7.6). A detailed assessment of
nature of emissions reductions. {6.6, Cross-Chapter Box 6.1} GHG metrics to support climate change mitigation and associated
policy contexts is provided in the WGIII contribution to the AR6.
The relative climate effects of different forcing agents are typically
quantified using emissions metrics that compare the effects of The global surface temperature response following a climate change
an idealised pulse of 1 kg of some climate forcing agent against mitigation measure that affects emissions of both short- and long-
a reference climate forcing agent, almost always CO2. The two most lived climate forcers depends on their lifetimes, their ERFs, how
prominent pulse emissions metrics are the global warming potential fast and for how long the emissions are reduced, and the thermal
(GWP) and global temperature change potential (GTP) (see Glossary). inertia in the climate system. Mitigation, relying on emissions
The climate responses to CO2 emissions by convention include the reductions and implemented through new legislation or technology
effects of warming on the carbon cycle, so for consistency these also standards, implies that emissions reductions occur year after year.
need to be determined for non-CO2 emissions. The methodology Global temperature response to a year’s worth of current emissions
for doing this has been placed on a more robust scientific footing from different sectors informs about the mitigation potential (Figure
compared to AR5 (high confidence). Methane from fossil fuel sources TS.20). Over 10- to 20-year time scales, the influence of SLCFs is
has slightly higher emissions metric values than those from biogenic at least as large as that of CO2, with sectors producing the largest
sources since it leads to additional fossil CO2 in the atmosphere warming being fossil fuel production and distribution, agriculture,
(high confidence). Updates to the chemical adjustments for CH4 and and waste management. Because the effects of the SLCFs decay
N2O emissions (Section TS.3.1) and revisions in their lifetimes result rapidly over the first few decades after emission, the net long-term
in emissions metrics for GWP and GTP that are slightly lower than temperature effect from a single year’s worth of current emissions is
in AR5 (medium confidence). Emissions metrics for the entire suite predominantly determined by CO2. Fossil fuel combustion for energy,
of GHGs assessed in the AR6 have been calculated for various time industry and land transportation are the largest contributing sectors
horizons. {7.6.1, Table 7.15, Table 7.SM.7} on a 100-year time scale (high confidence). Current emissions of CO2,
N2O and SLCFs from East Asia and North America are the largest
New emissions metric approaches, such as GWP* and Combined- regional contributors to additional net future warming on both
GTP (CGTP), relate changes in the emissions rate of short-lived short (medium confidence) and long time scales (10 and 100 years,
greenhouse gases to equivalent cumulative emissions of CO2 respectively) (high confidence). {6.6.1, 6.6.2, Figure 6.16}
(CO2-e). Global surface temperature response from aggregated
emissions of short-lived greenhouse gases over time is determined COVID-19 restrictions led to detectable reductions in global
by multiplying these cumulative CO2-e by TCRE (see Section TS.3.2.1). anthropogenic emissions of nitrogen oxides (NOx) (about 35% in
When GHGs are aggregated using standard metrics such as GWP or April 2020) and fossil CO2 (7%, with estimates ranging from 5.8% to
GTP, cumulative CO2-e emissions are not necessarily proportional to 13.0%), driven largely by reduced emissions from the transportation

101
 Technical Summary

Effect of a one year pulse of present-day emissions on global surface temperature

Response after 10 years

Response after 100 years Net effect, 5% to 95% range

−0.06 −0.05 −0.04 −0.03 −0.02 −0.01 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07
Change in global surface temperature due to total anthropogenic emissions (°C)

By sector, response after 10 years By sector, response after 100 years

Fossil fuel production CO2
and distribution CH4
Agriculture N2O
BC
Waste management OC
TS SO2
Residential and commercial NOx
CO
Fossil fuel combustion VOC
for energy NH3
Residential and commercial Avia−contrail
(biofuel use only) Avia−stratH2O
HFCs
Land transportation
Open biomass burning
Industry
Shipping
Aviation

−0.010 −0.005 0 0.005 0.010 0.015 −0.010 −0.005 0 0.005 0.010 0.015
Change in global surface temperature (°C) Change in global surface temperature (°C)

Figure TS.20 | Global surface temperature change 10 and 100 years after a one-year pulse of present-day emissions. The intent of this figure is to show the
sectoral contribution to present-day climate change by specific climate forcers, including carbon dioxide (CO2) as well as short-lived climate forcers (SLCFs). The temperature
response is broken down by individual species and shown for total anthropogenic emissions (top), and sectoral emissions on 10-year (left) and 100-year time scales (right).
Sectors are sorted by (high-to-low) net temperature effect on the 10-year time scale. Error bars in the top panel show the 5–95% range in net temperature effect due to
uncertainty in radiative forcing only (calculated using a Monte Carlo approach and best estimate uncertainties from the literature). Emissions for 2014 are from the Coupled
Model Intercomparison Project Phase 6 (CMIP6) emissions dataset, except for hydrofluorocarbons (HFCs) and aviation H2O, which rely on other datasets (see Section 6.6.2 for
more details). CO2 emissions are excluded from open biomass burning and residential biofuel use. {6.6.2, Figure 6.16}

sector (medium confidence). There is high confidence that, with temporarily adding to the total anthropogenic climate influence,
the exception of surface ozone, reductions in pollutant precursors with positive forcing (warming influence) from aerosol changes
contributed to temporarily improved air quality in most regions of dominating over negative forcings (cooling influence) from CO2, NOx
the world. However, these reductions were lower than what would and contrail cirrus changes. Consistent with this small net radiative
be expected from sustained implementation of policies addressing forcing, and against a large component of internal variability, Earth
air quality and climate change (medium confidence). Overall, the net system models show no detectable effect on global or regional
global ERF from COVID-19 containment was likely small and positive surface temperature or precipitation (high confidence). {Cross
for 2020 (with a temporary peak value less than 0.2 W m–2), thus Chapter Box 6.1}

102
 Technical Summary

Box TS.7 | Climate and Air Quality Responses to Short-lived Climate Forcers in Shared
Socio-economic Pathways

Future changes in emissions of short-lived climate forcers (SLCFs) are expected to cause an additional global mean
warming, with a large diversity in the end-of-century response across the WGI core set of Shared Socio-economic
Pathways (SSPs), depending upon the level of climate change and air pollution mitigation (Box TS.7, Figure 1). This
additional warming is either due to reductions in cooling aerosols for air pollution regulation or due to increases in
methane (CH4), ozone and hydrofluorocarbons (HFCs). This additional warming is stable after 2040 in SSPs associated
with lower global air pollution as long as CH4 emissions are also mitigated, but the overall warming induced by SLCF
changes is higher in scenarios in which air quality continues to deteriorate (induced by growing fossil fuel use and
limited air pollution control) (high confidence).

Sustained CH4 mitigation reduces global surface ozone, contributing to air quality improvements, and also reduces
surface temperature in the longer term, but only sustained CO2 emissions reductions allow long-term climate TS
stabilization (high confidence). Future changes in air quality (near-surface ozone and particulate matter, or PM) at
global and local scales are predominantly driven by changes in ozone and aerosol precursor emissions rather than
climate (high confidence). Air quality improvements driven by rapid decarbonization strategies, as in SSP1-1.9 and
SSP1-2.6, are not sufficient in the near term to achieve air quality guidelines set by the World Health Organization in
some highly polluted regions (high confidence). Additional policies (e.g., access to clean energy, waste management)
envisaged to attain United Nations Sustainable Development Goals bring complementary SLCF reduction. {4.4.4, 6.6.3,
6.7.3, Box 6.2}

The net effect of SLCF emissions changes on temperature will depend on how emissions of warming and cooling SLCFs will evolve
in the future. The magnitude of the cooling effect of aerosols remains the largest uncertainty in the effect of SLCFs in future climate
projections. Since the SLCFs have undergone large changes over the past two decades, the temperature and air pollution responses
are estimated relative to the year 2019 instead of 1995–2014.

Temperature Response
In the next two decades, it is very likely that SLCF emissions changes will cause a warming relative to 2019, across the WGI core set of
SSPs (see Section TS.1.3.1), in addition to the warming from long-lived GHGs. The net effect of SLCF and HFC changes in global surface
temperature across the SSPs is a likely warming of 0.06°C–0.35°C in 2040 relative to 2019. This near-term global mean warming
linked to SLCFs is quite similar in magnitude across the SSPs due to competing effects of warming (CH4, ozone) and cooling (aerosols)
forcers (Box TS.7, Figure 1). There is greater diversity in the end-of-century response among the scenarios. SLCF changes in scenarios
with no climate change mitigation (SSP3-7.0 and SSP5-8.5) will cause a warming in the likely range of 0.4°C–0.9°C in 2100 relative to
2019 due to increases in CH4, tropospheric ozone and HFC levels. For the stringent climate change and pollution mitigation scenarios
(SSP1-1.9 and SSP1-2.6), the cooling from reductions in CH4, ozone and HFCs partially balances the warming from reduced aerosols,
primarily sulphate, and the overall SLCF effect is a likely increase in global surface temperature of 0.0°C–0.3°C in 2100, relative to
2019. With intermediate climate change and air pollution mitigations, SLCFs in SSP2-4.5 add a likely warming of 0.2°C–0.5°C to global
surface temperature change in 2100, with the largest warming resulting from reductions in aerosols. {4.4.4, 6.7.3}

Assuming implementation and efficient enforcement of both the Kigali Amendment to the Montreal Protocol on Substances that
Deplete the Ozone Layer and current national plans result in limiting emissions (as in SSP1-2.6), the effects of HFCs on global
surface temperature, relative to 2019, would remain below +0. 02°C from 2050 onwards versus about +0.04°C–0.08°C in 2050 and
+0.1°C–0.3°C in 2100 considering only national HFC regulations decided prior to the Kigali Amendment (as in SSP5-8.5) (medium
confidence). {6.6.3, 6.7.3}

Air Quality Responses

Air pollution projections range from strong reductions in global surface ozone and PM (e.g., SSP1-2.6, with stringent mitigation of both
air pollution and climate change) to no improvement and even degradation (e.g., SSP3-7.0 without climate change mitigation and
with only weak air pollution control) (high confidence). Under the SSP3-7.0 scenario, PM levels are projected to increase until 2050
over large parts of Asia, and surface ozone pollution is projected to worsen over all continental areas through 2100 (high confidence).
In SSP5-8.5, a scenario without climate change mitigation but with stringent air pollution control, PM levels decline through 2100, but
high CH4 levels hamper the decline in global surface ozone at least until 2080 (high confidence). {6.7.1}

103
 Technical Summary

Box TS.7 (continued)

TS

Box TS.7, Figure 1 | Effects of short-lived climate forcers (SLCFs) on global surface temperature and air pollution across the WGI core set of
Shared Socio-economic Pathways (SSPs). The intent of this figure is to show the climate and air quality (surface ozone and particulate matter smaller than
2.5 microns in diameter, or PM2.5 ) response to SLCFs in the SSP scenarios for the near and long-term. Effects of net aerosols, tropospheric ozone, hydrofluorocarbons
(HFCs; with lifetimes less than 50 years), and methane (CH4) are compared with those of total anthropogenic forcing for 2040 and 2100 relative to year 2019. The
global surface temperature changes are based on historical and future evolution of effective radiative forcing (ERF) as assessed in Chapter 7 of this Report. The
temperature responses to the ERFs are calculated with a common impulse response function (RT) for the climate response, consistent with the metric calculations in
Chapter 7 (Box 7.1). The RT has an equilibrium climate sensitivity of 3.0°C for a doubling of atmospheric CO2 concentration (feedback parameter of –1.31 W m–2 °C–1).
The scenario total (grey bar) includes all anthropogenic forcings (long- and short-lived climate forcers, and land-use changes). Uncertainties are 5–95% ranges. The
global changes in air pollutant concentrations (ozone and PM2.5) are based on multimodel Coupled Model Intercomparison Project Phase 6 (CMIP6) simulations and
represent changes in five-year mean surface continental concentrations for 2040 and 2098 relative to 2019. Uncertainty bars represent inter-model ±1 standard
deviation. {6.7.2, 6.7.3, Figure 6.24}

Box TS.8 | Earth System Response to Solar Radiation Modification

Since AR5, further modelling work has been conducted on aerosol-based solar radiation modification (SRM) options
such as stratospheric aerosol injection, marine cloud brightening, and cirrus cloud thinning21 and their climate and
biogeochemical effects. These investigations have consistently shown that SRM could offset some of the effects of
increasing greenhouse gases on global and regional climate, including the carbon and water cycles (high confidence).
However, there would be substantial residual or overcompensating climate change at the regional scales and seasonal
time scales (high confidence), and large uncertainties associated with aerosol–cloud–radiation interactions persist.
The cooling caused by SRM would increase the global land and ocean CO2 sinks (medium confidence), but this
would not stop CO2 from increasing in the atmosphere or affect the resulting ocean acidification under continued
anthropogenic emissions (high confidence). It is likely that abrupt water cycle changes will occur if SRM techniques
are implemented rapidly. A sudden and sustained termination of SRM in a high CO2 emissions scenario would cause
rapid climate change (high confidence). However, a gradual phase-out of SRM combined with emissions reduction and
carbon dioxide removal (CDR) would avoid these termination effects (medium confidence). {4.6.3, 5.6.3. 6.4.6, 8.6.3}.

21 Although cirrus cloud thinning aims to cool the planet by increasing longwave emissions to space, it is included in the portfolio of SRM options for consistency with AR5 and SR1.5. {4.6.3.3}

104
 Technical Summary

Box TS.8 (continued)

Solar radiation modification (SRM) refers to deliberate, large-scale climate intervention options that are studied as potential
supplements to deep mitigation, for example, in scenarios that overshoot climate stabilization goals. SRM options aim to offset some
of the warming effects of GHG emissions by modification of Earth’s shortwave radiation budget. Following SR1.5, the SRM assessed
in this Report also includes some options, such as cirrus cloud thinning, that alter the longwave radiation budget.

SRM contrasts with climate change mitigation activities, such as emissions reductions and CDR, as it introduces a ‘mask’ to the
climate change problem by altering Earth’s radiation budget, rather than attempting to address the root cause of the problem, which
is the increase in GHGs in the atmosphere. By masking only the climate effects of GHG emissions, SRM does not address other issues
related to atmospheric CO2 increase, such as ocean acidification. This Report assesses physical understanding of the Earth system
response to proposed SRM, and the assessment is based primarily on idealized climate model simulations. There are other important
considerations, such as risk to human and natural systems, perceptions, ethics, cost, governance, and trans-boundary issues and their
relationship to the United Nations Sustainable Development Goals – issues that the WGII (Chapter 16) and WGIII (Chapter 14) Reports TS
address. {4.6.3}

SRM options include those that increase surface albedo, brighten marine clouds by increasing the amount of cloud condensation
nuclei, or reduce the optical depth of cirrus clouds by seeding them with ice nucleating particles. However, the most commonly studied
approaches attempt to mimic the cooling effects of major volcanic eruptions by injecting reflective aerosols (e.g., sulphate aerosols)
or their precursors (e.g., sulphur dioxide) into the stratosphere. {4.6.3, 5.6.3, 6.4.6}

SRM could offset some effects of greenhouse gas-induced warming on global and regional climate, but there would be substantial
residual and overcompensating climate change at the regional scale and seasonal time scales (high confidence). Since AR5, more
modelling work has been conducted with more sophisticated treatment of aerosol-based SRM approaches, but the uncertainties in
cloud–aerosol–radiation interactions are still large (high confidence). Modelling studies suggest that it is possible to stabilize multiple
large-scale temperature indicators simultaneously by tailoring the deployment strategy of SRM options (medium confidence) but with
large residual or overcompensating regional and seasonal climate changes. {4.6.3}

SRM approaches targeting shortwave radiation are likely to reduce global mean precipitation, relative to future CO2 emissions
scenarios, if all global mean warming is offset. In contrast, cirrus cloud thinning, targeting longwave radiation, is expected to cause
an increase in global mean precipitation (medium confidence). If shortwave approaches are used to offset global mean warming,
the magnitude of reduction in regional precipitation minus evapotranspiration (P–E) (Box TS.5), which is more relevant to freshwater
availability, is smaller than precipitation decrease because of simultaneous reductions in both precipitation and evapotranspiration
(medium confidence). {4.6.3, 8.2.1, 8.6.3}.

If SRM is used to cool the planet, it would cause a reduction in plant and soil respiration and slow the reduction of ocean carbon
uptake due to warming (medium confidence). The result would be an enhancement of the global land and ocean CO2 sinks (medium
confidence) and a slight reduction in atmospheric CO2 concentration relative to unmitigated climate change. However, SRM would
not stop CO2 from increasing in the atmosphere or affect the resulting ocean acidification under continued anthropogenic emissions
(high confidence). {5.6.3}

The effect of stratospheric aerosol injection on global temperature and precipitation is projected by models to be detectable after one
to two decades, which is similar to the time scale for the emergence of the benefits of emissions reductions. A sudden and sustained
termination of SRM in a high GHG emissions scenario would cause rapid climate change and a reversal of the SRM effects on the
carbon sinks (high confidence). It is also likely that a termination of strong SRM would drive abrupt changes in the water cycle globally
and regionally, especially in the tropical regions by shifting the Inter-tropical Convergence Zone and Hadley cells. At the regional scale,
non-linear responses cannot be excluded, due to changes in evapotranspiration. However, a gradual phase-out of SRM combined with
emissions reductions and CDR would avoid larger rates of changes (medium confidence). {4.6.3, 5.6.3, 8.6.3}.

105
 Technical Summary

Box TS.9 | Irreversibility, Tipping Points and Abrupt Changes

The present rates of response of many aspects of the climate system are proportionate to the rate of recent temperature
change, but some aspects may respond disproportionately. Some climate system components are slow to respond,
such as the deep ocean overturning circulation and the ice sheets (Box TS.4). It is virtually certain that irreversible,
committed change is already underway for the slow-to-respond processes as they come into adjustment for past and
present emissions.

The paleoclimate record indicates that tipping elements exist in the climate system where processes undergo sudden
shifts toward a different sensitivity to forcing, such as during a major deglaciation, where 1°C degree of temperature
change might correspond to a large or small ice-sheet mass loss during different stages (Box TS.2). For global climate
indicators, evidence for abrupt change is limited, but deep ocean warming, acidification and sea level rise are committed
to ongoing change for millennia after global surface temperatures initially stabilize and are irreversible on human time
TS scales (very high confidence). At the regional scale, abrupt responses, tipping points and even reversals in the direction of
change cannot be excluded (high confidence). Some regional abrupt changes and tipping points could have severe local
impacts, such as unprecedented weather, extreme temperatures and increased frequency of droughts and forest fires.

Models that exhibit such tipping points are characterized by abrupt changes once the threshold is crossed, and even
a return to pre-threshold surface temperatures or to atmospheric carbon dioxide concentrations does not guarantee
that the tipping elements return to their pre-threshold state. Monitoring and early warning systems are being put into
place to observe tipping elements in the climate system. {1.3, 1.4.4, 1.5, 4.3.2, Table 4.10, 5.3.4, 5.4.9, 7.5.3, 9.2.2,
9.2.4, 9.4.1, 9.4.2, 9.6.3, Cross-chapter Box 12.1}

Understanding of multi-decadal reversibility (i.e., the system returns to the previous climate state within multiple decades after
the radiative forcing is removed) has improved since AR5 for many atmospheric, land surface and sea ice climate metrics following
sea surface temperature recovery. Some processes suspected of having tipping points, such as the Atlantic Meridional Overturning
Circulation (AMOC), have been found to often undergo recovery after temperature stabilization with a time delay (low confidence).
However, substantial irreversibility is further substantiated for some cryosphere changes, ocean warming, sea level rise, and ocean
acidification. {4.7.2, 5.3.3, 5.4.9, 9.2.2, 9.2.4, 9.4.1, 9.4.2, 9.6.3}

Some climate system components are slow to respond, such as the deep ocean overturning circulation and the ice sheets. It is likely that
under stabilization of global warming at 1.5°C, 2.0°C or 3.0°C relative to 1850–1900, the AMOC will continue to weaken for several
decades by about 15%, 20% and 30% of its strength and then recover to pre-decline values over several centuries (medium confidence).
At sustained warming levels between 2°C and 3°C, there is limited evidence that the Greenland and West Antarctic ice sheets will be lost
almost completely and irreversibly over multiple millennia; both the probability of their complete loss and the rate of mass loss increases
with higher surface temperatures (high confidence). At sustained warming levels between 3°C and 5°C, near-complete loss of the
Greenland Ice Sheet and complete loss of the West Antarctic Ice Sheet is projected to occur irreversibly over multiple millennia (medium
confidence); with substantial parts or all of Wilkes Subglacial Basin in East Antarctica lost over multiple millennia (low confidence). Early-
warning signals of accelerated sea level rise from Antarctica could possibly be observed within the next few decades. For other hazards
(e.g., ice-sheet behaviour, glacier mass loss and global mean sea level change, coastal floods, coastal erosion, air pollution, and ocean
acidification) the time and/or scenario dimensions remain critical, and a simple and robust relationship with global warming level cannot
be established (high confidence). {4.3.2, 4.7.2, 5.4.3, 5.4.5, 5.4.8, 8.6, 9.2, 9.4, Box 9.3, Cross-Chapter Box 12.1}

For global climate indicators, evidence for abrupt change is limited. For global warming up to 2°C above 1850–1900 levels, paleoclimate
records do not indicate abrupt changes in the carbon cycle (low confidence). Despite the wide range of model responses, uncertainty
in atmospheric CO2 by 2100 is dominated by future anthropogenic emissions rather than uncertainties related to carbon–climate
feedbacks (high confidence). There is no evidence of abrupt change in climate projections of global temperature for the next century:
there is a near-linear relationship between cumulative CO2 emissions and maximum global mean surface air temperature increase
caused by CO2 over the course of this century for global warming levels up to at least 2°C relative to 1850–1900. The increase in global
ocean heat content (Section TS.2.4) will likely continue until at least 2300 even for low emissions scenarios, and global mean sea level
will continue to rise for centuries to millennia following cessation of emissions (Box TS.4) due to continuing deep ocean heat uptake
and mass loss of the Greenland and Antarctic ice sheets (high confidence). {2.2.3; Cross-Chapter Box 2.1; 5.1.1; 5.4; Cross-Chapter
Box 5.1; Figures 5.3, 5.4, 5.25, and 5.26; 9.2.2; 9.2.4}

The response of biogeochemical cycles to anthropogenic perturbations can be abrupt at regional scales and irreversible on decadal to
century time scales (high confidence). The probability of crossing uncertain regional thresholds increases with climate change (high

106
 Technical Summary

Box TS.9 (continued)

confidence). It is very unlikely that gas clathrates (mostly methane) in deeper terrestrial permafrost and subsea clathrates will lead to a
detectable departure from the emissions trajectory during this century. Possible abrupt changes and tipping points in biogeochemical
cycles lead to additional uncertainty in 21st century atmospheric GHG concentrations, but future anthropogenic emissions remain the
dominant uncertainty (high confidence). There is potential for abrupt water cycle changes in some high emissions scenarios, but there
is no overall consistency regarding the magnitude and timing of such changes. Positive land surface feedbacks, including vegetation,
dust, and snow, can contribute to abrupt changes in aridity, but there is only low confidence that such changes will occur during
the 21st century. Continued Amazon deforestation, combined with a warming climate, raises the probability that this ecosystem will
cross a tipping point into a dry state during the 21st century (low confidence). (Section TS.3.2.2) {5.4.3, 5.4.5, 5.4.8, 5.4.9, 8.6.2, 8.6.3,
Cross-Chapter Box 12.1}

TS
TS.4 Regional Climate Change requirements, and demand (very high confidence). The
decision-making context, level of user engagement,
This section focuses on how to generate regional climate change and co-production between scientists, practitioners and
information and its relevance for climate services; the drivers of regional users are important determinants of the type of climate
climate variability and change and how they are being affected by service developed and its utility in supporting adaptation,
anthropogenic factors; and observed, attributed and projected changes mitigation and risk management decisions. {10.3, 10.6,
in climate, including extreme events and climatic impact-drivers (CIDs), Cross-Chapter Box 10.3, 12.6, Cross-Chapter Box 12.2}
across all regions of the world. There is a small set of CID changes
common to all land or ocean regions and a specific set of changes from
a broader range of CIDs seen in each region. This regional diversity TS.4.1.1 Sources and Methodologies for Generating Regional
results from regional climate being determined by a complex interplay Climate Information
between the seasonal-to-multi-decadal variation of large-scale modes
of climate variability, external natural and anthropogenic forcings, local Climate change information at regional scale is generated using a range
climate processes and related feedbacks. of data sources and methodologies (Section TS.1.4). Understanding
of observed regional climate change and variability is based on the
availability and analysis of multiple observational datasets that are
TS.4.1 Generation and Communication of Regional suitable for evaluating the phenomena of interest (e.g., extreme
Climate Change Information events), including accounting for observational uncertainty (Section
TS.1.2.1). These datasets are combined with climate model simulations
Climate change information at regional scale is generated of observed changes and events to attribute causes of those changes
using a range of data sources and methodologies. Multi- and events to large- and regional-scale anthropogenic and natural
model ensembles and models with a range of resolutions drivers and to assess the performance of the models. Future simulations
are important data sources, and discarding models that with many climate models (multi-model ensembles) are then used to
fundamentally misrepresent relevant processes improves generate and quantify ranges of projected regional climate responses
the credibility of ensemble information related to these (Section TS.4.2). Discarding models that fundamentally misrepresent
processes. A key methodology is distillation – combining relevant processes improves the credibility of regional climate
lines of evidence and accounting for stakeholder context information generated from these ensembles (high confidence).
and values – which helps ensure the information is relevant, However, multi-model mean and ensemble spread are not a full
useful and trusted for decision-making (see Core Concepts measure of the range of projection uncertainty and are not sufficient
Box) (high confidence). to characterize low-likelihood, high-impact changes (Box TS.3) or
situations where different models simulate substantially different
Since AR5, physical climate storylines have emerged as or even opposite changes (high confidence). Large single-model
a complementary approach to ensemble projections for ensembles are now available and provide a more comprehensive
generating more accessible climate information and spectrum of possible changes associated with internal variability (high
promoting a more comprehensive treatment of risk. They confidence) (Section TS.1.2.3). {1.5.1, 1.5.4, 10.2, 10.3.3, 10.3.4, 10.4.1,
have been used as part of the distillation process within 10.6.2, 11.2, Box 11.2, Cross-Chapter Box 11.1, 12.4, Atlas.1.4.1}
climate services to generate the required context-relevant,
credible and trusted climate information. Depending on the region of interest, representing regionally important
forcings (e.g., aerosols, land-use change and ozone concentrations)
Since AR5, climate change information produced for climate and feedbacks (e.g., between snow and albedo, soil moisture and
services has increased significantly due to scientific and temperature, or soil moisture and precipitation) in climate models is a
technological advancements and growing user awareness, prerequisite for them to reproduce past regional trends to underpin the

107
 Technical Summary

reliability of future projections (medium confidence) (Section TS.1.2.2). Methodologies such as statistical downscaling, bias adjustment and
In some cases, even the sign of a projected change in regional climate weather generators are beneficial as an interface between climate
cannot be trusted if relevant regional processes are not represented, for model projections and impact modelling and for deriving user-
example, for variables such as precipitation and wind speed (medium relevant indicators (high confidence). However, the performance of
confidence). In some regions, either geographical (e.g., Central Africa, these techniques depends on that of the driving climate model: in
Antarctica) or typological (e.g., mountainous areas, Small Islands and particular, bias adjustment cannot overcome all consequences of
cities), and for certain phenomena, fewer observational records are unresolved or strongly misrepresented physical processes, such as
available or accessible, which limits the assessment of regional climate large-scale circulation biases or local feedbacks (medium confidence).
change in these cases. {1.5.1, 1.5.3, 1.5.4, 8.5.1, 10.2, 10.3.3, 10.4.1, {10.3.3, Cross-Chapter Box 10.2, 12.2, Atlas.2.2}
11.1.6, 11.2, 12.4, Atlas.8.3, Atlas.11.1.5, Cross-Chapter Box Atlas.2}

Box TS.10 | Event Attribution

TS
The attribution of observed changes in extremes to human influence (including greenhouse gas and aerosol emissions
and land-use changes) has substantially advanced since AR5, in particular for extreme precipitation, droughts,
tropical cyclones, and compound extremes (high confidence). There is limited evidence for windstorms and convective
storms. Some recent hot extreme events would have been extremely unlikely to occur without human influence on the
climate system. (Section TS.1) {Cross-Working Group Box: Attribution in Chapter 1, 11.2, 11.3, 11.4, 11.6, 11.7, 11.8}

Since AR5, the attribution of extreme weather events has emerged as a growing field of climate research with an increasing body of
literature. It provides evidence that greenhouse gases and other external forcings have affected individual extreme weather events
by disentangling anthropogenic drivers from natural variability. Event attribution is now an important line of evidence for assessing
changes in extremes on regional scales. (Section TS.1) {Cross-Working Group Box: Attribution, 11.1.4}

The regional extremes and events that have been studied are geographically uneven (Section TS.4.1). A few events, for example,
extreme rainfall events in the United Kingdom, heatwaves in Australia, or Hurricane Harvey that hit Texas in 2017, have been heavily
studied. Many highly impactful extreme weather events have not been studied in the event attribution framework, particularly in
the developing world where studies are generally lacking. This is due to various reasons, including lack of observational data, lack of
reliable climate models, and lack of scientific capacity. While the events that have been studied are not representative of all extreme
events that have occurred, and results from these studies may also be subject to selection bias, the large number of event attribution
studies provide evidence that changes in the properties of these local and individual events are in line with expected consequences
of human influence on the climate and can be attributed to external drivers. {Cross-Working Group Box: Attribution, 11.1.4, 11.2.2}

It is very likely that human influence is the main contributor to the observed increase in the intensity and frequency of hot extremes and
the observed decrease in the intensity and frequency of cold extremes on continental scales. Some specific recent hot extreme events
would have been extremely unlikely to occur without human influence on the climate system. Changes in aerosol concentrations
have likely slowed the increase in hot extremes in some regions, in particular from 1950–1980. No-till farming, irrigation and crop
expansion have similarly attenuated increases in summer hot extremes in some regions, such as central North America (medium
confidence). {11.3.4}

Human influence has contributed to the intensification of heavy precipitation in three continents where observational data are most
abundant: North America, Europe and Asia (high confidence). On regional scales, evidence of human influence on extreme precipitation
is limited, but new evidence from attributing individual heavy precipitation events found that human influence was a significant driver
of the events. {11.4.4}

There is low confidence that human influence has affected trends in meteorological droughts in most regions, but medium confidence
that they have contributed to the severity of some specific events. There is medium confidence that human-induced climate change
has contributed to increasing trends in the probability or intensity of recent agricultural and ecological droughts, leading to an
increase of the affected land area. {11.6.4}

Event attribution studies of specific strong tropical cyclones provide limited evidence for anthropogenic effects on tropical cyclone
intensifications so far, but high confidence for increases in precipitation. There is high confidence that anthropogenic climate change
contributed to extreme rainfall amounts during Hurricane Harvey (in 2017) and other intense tropical cyclones. {11.7.3}

108
 Technical Summary

Box TS.10 (continued)

The number of evident attribution studies on compound events is limited. There is medium confidence that weather conditions that
promote wildfires have become more probable in southern Europe, northern Eurasia, the USA, and Australia over the last century. In
Australia a number of event attribution studies show that there is medium confidence of increase in fire weather conditions due to
human influence. {11.8.3, 12.4.3.2}

Climate change is already affecting every inhabited region across the globe,
with human influence contributing to many observed changes in weather
and climate extremes
(a) Synthesis of assessment of observed change in hot extremes and
confidence in human contribution to the observed changes in the world’s regions
Type of observed change
in hot extremes North
America GIC Europe
Increase (41) NWN NEN NEU RAR

WNA CNA ENA WCE EEU WSB ESB RFE Asia TS

Decrease (0)

NCA MED WCA ECA TIB EAS

Low agreement in the type of change (2) Small
Islands
Central SCA CAR SAH ARP SAS SEA
Limited data and/or literature (2) America PAC
NWS NSA WAF CAF NEAF
NAU
Confidence in human contribution Small
to the observed change SAM NES WSAF SEAF MDG Islands
CAU EAU
High
South SWS SES ESAF
Medium America Africa
Australasia
SAU NZ
Low due to limited agreement SSA
Low due to limited evidence
Type of observed change since the 1950s

(b) Synthesis of assessment of observed change in heavy precipitation and

confidence in human contribution to the observed changes in the world’s regions
Type of observed change
in heavy precipitation North
America GIC Europe
NWN NEN NEU RAR
Increase (19)
WNA CNA ENA WCE EEU WSB ESB RFE Asia
Decrease (0)
NCA MED WCA ECA TIB EAS
Small
Low agreement in the type of change (8) Islands
Central SCA CAR SAH ARP SAS SEA
Limited data and/or literature (18) America PAC
NWS NSA WAF CAF NEAF
NAU
Confidence in human contribution Small
SAM NES WSAF SEAF MDG Islands
to the observed change CAU EAU
High South SWS SES ESAF
Africa SAU
Medium America
Australasia NZ
Low due to limited agreement SSA
Low due to limited evidence
Type of observed change since the 1950s

(c) Synthesis of assessment of observed change in agricultural and ecological drought

and confidence in human contribution to the observed changes in the world’s regions
Type of observed change
in agricultural and ecological drought North
America GIC Europe
Increase (12) NWN NEN NEU RAR

WNA CNA ENA WCE EEU WSB ESB RFE Asia

Decrease (1)

NCA MED WCA ECA TIB EAS

Low agreement in the type of change (28) Small
Islands
Central SCA CAR SAH ARP SAS SEA
Limited data and/or literature (4) America PAC
NWS NSA WAF CAF NEAF
NAU
Confidence in human contribution Small
to the observed change SAM NES WSAF SEAF MDG Islands
CAU EAU
High
South SWS SES ESAF
Medium America Africa SAU NZ
Australasia
Low due to limited agreement SSA
Low due to limited evidence
Type of observed change since the 1950s

Each hexagon corresponds IPCC AR6 WGI reference regions: North America: NWN (North-Western North America, NEN (North-Eastern North America), WNA
to one of the IPCC AR6 (Western North America), CNA (Central North America), ENA (Eastern North America), Central America: NCA (Northern Central America),
WGI reference regions SCA (Southern Central America), CAR (Caribbean), South America: NWS (North-Western South America), NSA (Northern South America), NES
(North-Eastern South America), SAM (South American Monsoon), SWS (South-Western South America), SES (South-Eastern South America),
North-Western SSA (Southern South America), Europe: GIC (Greenland/Iceland), NEU (Northern Europe), WCE (Western and Central Europe), EEU (Eastern
NWN Europe), MED (Mediterranean), Africa: MED (Mediterranean), SAH (Sahara), WAF (Western Africa), CAF (Central Africa), NEAF (North Eastern
North America
Africa), SEAF (South Eastern Africa), WSAF (West Southern Africa), ESAF (East Southern Africa), MDG (Madagascar), Asia: RAR (Russian
Arctic), WSB (West Siberia), ESB (East Siberia), RFE (Russian Far East), WCA (West Central Asia), ECA (East Central Asia), TIB (Tibetan Plateau),
EAS (East Asia), ARP (Arabian Peninsula), SAS (South Asia), SEA (South East Asia), Australasia: NAU (Northern Australia), CAU (Central
Australia), EAU (Eastern Australia), SAU (Southern Australia), NZ (New Zealand), Small Islands: CAR (Caribbean), PAC (Pacific Small Islands)

Box TS.10, Figure 1 | Synthesis of assessed observed and attributable regional changes.

109
 Technical Summary

Box TS.10 (continued)

Box TS.10, Figure 1 (continued): The IPCC AR6 WGI inhabited regions are displayed as hexagons of identical sizes in their approximate geographical location (see
legend for regional acronyms). All assessments are made for each region as a whole and for the 1950s to the present. Assessments made on different time scales or
more local spatial scales might differ from what is shown in the figure. The colours in each panel represent the four outcomes of the assessment on observed changes.
Striped hexagons (white and light-grey) are used where there is low agreement in the type of change for the region as a whole, and grey hexagons are used when
there is limited data and/or literature that prevents an assessment of the region as a whole. Other colours indicate at least medium confidence in the observed change.
The confidence level for the human influence on these observed changes is based on assessing trend detection and attribution and event attribution literature,
and it is indicated by the number of dots: three dots for high confidence, two dots for medium confidence and one dot for low confidence (single, filled dot: limited
agreement; single, empty dot: limited evidence).
Panel (a) For hot extremes, the evidence is mostly drawn from changes in metrics based on daily maximum temperatures; regional studies using other indices
(heatwave duration, frequency and intensity) are used in addition. Red hexagons indicate regions where there is at least medium confidence in an observed increase
in hot extremes.
Panel (b) For heavy precipitation, the evidence is mostly drawn from changes in indices based on one-day or five-day precipitation amounts using global and
regional studies. Green hexagons indicate regions where there is at least medium confidence in an observed increase in heavy precipitation.
Panel (c) Agricultural and ecological droughts are assessed based on observed and simulated changes in total column soil moisture, complemented by evidence
TS on changes in surface soil moisture, water balance (precipitation minus evapotranspiration) and indices driven by precipitation and atmospheric evaporative demand.
Yellow hexagons indicate regions where there is at least medium confidence in an observed increase in this type of drought and green hexagons indicate regions
where there is at least medium confidence in an observed decrease in agricultural and ecological drought.
For all regions, Table TS.5 shows a broader range of observed changes besides the ones shown in this figure. Note that Southern South America (SSA) is the only
region that does not display observed changes in the metrics shown in this figure, but is affected by observed increases in mean temperature, decreases in frost and
increases in marine heatwaves.
(Table TS.5) {11.9, Atlas 1.3.3, Figure Atlas.2}

TS.4.1.2 Regional Climate Information Distillation and changes in a key mode of variability (the Southern Annular Mode)
Climate Services and drought in Cape Town among different observation periods and
in model simulations. {10.5.3, 10.6, 10.6.2, 10.6.4, Cross-Chapter
The construction of regional climate information involves people with Box 10.3, 12.4}
a variety of backgrounds, from various disciplines, who have different
sets of experiences, capabilities and values. The process of synthesizing Since AR5, physical climate storyline approaches have emerged as
climate information from different lines of evidence from a number a complementary instrument to provide a different perspective, or
of sources, taking into account the context of a user vulnerable to additional climate information, to facilitate communication of the
climate variability and change and the values of all relevant actors, is information or provide a more flexible consideration of risk. Storylines
called distillation. Distillation is conditioned by the sources available, that condition climatic events and processes on a set of plausible
the actors involved, and the context, which all depend heavily on the but distinct large-scale climatic changes enable the exploration of
regions considered, and is framed by the question being addressed. uncertainties in regional climate projections. For example, they can
Distilling regional climate information from multiple lines of evidence explicitly address low-likelihood, high-impact outcomes, which would
and taking the user context into account increases fitness, usefulness, be less emphasized in a probabilistic approach, and can be embedded
relevance and trust in that information for use in climate services in a user’s risk landscape, taking account of socio-economic factors
(Box TS.11) and decision-making (high confidence). {1.2.3, 10.1.4, as well as physical climate changes. Storylines can also be used to
10.5, Cross-Chapter Box 10.3, 12.6} communicate climate information by narrative elements describing
and contextualizing the main climatological features and the relevant
The distillation process can vary substantially, as it needs to consider consequences in the user context and, as such, can be used as part
multiple lines of evidence on all physically plausible outcomes of a climate information distillation process. {1.4.4., Box 10.2, 11.2,
(especially when they are contrasting) relevant to a specific decision Box 11.2, Cross-Chapter Box 12.2}
required in response to a changing climate. Confidence in the
distilled regional climate information is enhanced when there is
agreement across multiple lines of evidence, so the outcome can
be limited if these are inconsistent or contradictory. For example, in
the Mediterranean region the agreement between different lines of
evidence, such as observations, projections by regional and global
models, and understanding of the underlying mechanisms, provides
high confidence in summer warming that exceeds the global average
(see Box TS.12). In a less clear-cut case for Cape Town, South Africa,
despite consistency among global model future projections, there
is medium confidence in a projected future drier climate due to the
lack of consistency in links between increasing greenhouse gases,

110
 Technical Summary

Box TS.11 | Climate Services

Climate services involve providing climate information to assist decision-making, for example, about how extreme
rainfall will change to inform improvements in urban drainage. Since AR5, there has been a significant increase in the
range and diversity of climate service activities (very high confidence). The level of user-engagement, co-design and
co-production are factors determining the utility of climate services, while resource limitations for these activities
constrain their full potential. {12.6, Cross-Chapter Box 12.2}

Climate services include engagement from users and providers and an effective access mechanism; they are responsive to user needs
and based on integrating scientifically credible information and relevant expertise. Climate services are being developed across
regions, sectors, time scales and user-groups and include a range of knowledge brokerage and integration activities. These involve
identifying knowledge needs; compiling, translating and disseminating knowledge; coordinating networks and building capacity
through informed decision-making; analysis, evaluation and development of policy; and personal consultation.
TS
Since AR5, climate change information produced in climate service contexts has increased significantly due to scientific and
technological advancements and growing user awareness, requirements and demand (very high confidence). Climate services are
growing rapidly and are highly diverse in their practices and products. The decision-making context, level of user engagement and co-
production between scientists, practitioners and intended users are important determinants of the type of climate service developed
and their utility for supporting adaptation, mitigation and risk management decisions. They require different types of user–producer
engagement depending on what the service aims to deliver (high confidence), and these fall into three broad categories: website-
based services, interactive group activities and focused relationships.

Realization of the full potential of climate services is often hindered by limited resources for the co-design and co-production process,
including sustained engagement between scientists, service providers and users (high confidence). Further challenges relate to the
development and provision of climate services, generation of climate service products, communication with users, and evaluation of
their quality and socio-economic benefit. (Section TS.4.1) {1.2.3, 10.5.4, 12.6, Cross-Chapter Box 12.2, Glossary}

Box TS.12 | Multiple Lines of Evidence for Assessing Regional Climate Change and the
Interactive Atlas

A key novel element in the AR6 is the Working Group I Atlas, which includes the Interactive Atlas (https://interactive-
atlas.ipcc.ch/). The Interactive Atlas provides the ability to explore much of the observational and climate model data
used as lines of evidence in this assessment to generate regional climate information. {Atlas.2}

A significant innovation in the AR6 WGI Report is the Atlas. Part of its remit is to provide region-by-region assessment on changes in
mean climate and to link with other WGI chapters to generate climate change information for the regions. An important component
is the new online interactive tool, the Interactive Atlas, with flexible spatial and temporal analyses of much of the observed, simulated
past and projected future climate change data underpinning the WGI assessment. This includes the ability to generate global maps
and a number of regionally aggregated products (time series, scatter plots, tables, etc.) for a range of observations and ensemble
climate change projections of variables (such as changes in the climatic impact-drivers summarized in Table TS.5) from the Coupled
Model Intercomparison Project Phases 5 and 6 (CMIP5, CMIP6) and the Coordinated Regional Climate Downscaling Experiment
(CORDEX). The data can be displayed and summarized under a range of SSP-RCP scenarios and future time slices and also for different
global warming levels, relative to several different baseline periods. The maps and various statistics can be generated for annual mean
trends and changes or for any user-specified season. A new set of WGI reference regions is used for the regional summary statistics
and applied widely throughout the report (with the regions, along with aggregated datasets and the code to generate these, available
at the ATLAS GitHub: https://github.com/IPCC-WG1/Atlas).

Box TS.12, Figure 1 shows how the Interactive Atlas products, together with other lines of evidence, can be used to generate climate
information for an illustrative example of the Mediterranean summer warming. The lines of evidence include the understanding of
relevant mechanisms, dynamic and thermodynamic processes and the effect of aerosols in this case (Box TS.12, Figure 1a); trends
in observational datasets (which can have different spatial and temporal coverage; Box TS.12, Figure 1b, c); and attribution of these
trends and temperature projections from global and regional climate models at different resolutions, including single-model initial-
condition large ensembles (SMILEs; Box TS.12, Figure 1d, e). Taken together, this evidence shows there is high confidence that the

111
 Technical Summary

Box TS.12 (continued)

projected Mediterranean summer temperature increase will be larger than the global mean, with consistent results from CMIP5 and
CMIP6 (Box TS.12, Figure 1e). However, CMIP6 results project both more pronounced warming than CMIP5 for a given emissions
scenario and time period and a greater range of changes (Box TS.12, Figure 1d). {10.6.4, Atlas.2, Atlas.8.4}

The Interactive Atlas allows for flexible

spatial and temporal analyses of essential
climate variables, extreme indices and
climatic impact-drivers, including
multiple lines of evidence to support the
assessment of regional climate change:
- Observations
TS - CMIP5
- CMIP6
- CORDEX
Regional information
is displayed when CORDEX is available for 12
clicking on one or continent-wide domains.
several subregions
Regional (aggregated) information for
reference and typological regions:
Time series
Stripes
Annual cycle plots
Summary tabular information.
Scatter plots (e.g., precip. vs temp.)
Dimensions of analysis include time
periods across scenarios and global warming
levels (1ºC, 2ºC, 3ºC and 4ºC).
Available in the Interactive Atlas
Not available from the Interactive Atlas

(a) Mechanisms of enhanced Mediterranean warming (b) Station locations (c) Temperature trend distribution
Past period (1960–2014)
OBS
Warm Atlantic Ocean Reduction of Surface drying Lapse rate difference Ensemble means
aerosols Over land stronger CMIP5
Enhanced warming than over sea CMIP6
sensible HighResMIP
Reduced Sea Land CORDEX EUR-44
heat flux
latent CORDEX EUR-11
heat flux MIROC6
Altitude

Dynamic and CSIRO-Mk3-6-0

thermodynamic effects MPI-ESM
d4PDF
E-OBS Donat et al. 2013 0.2 0.4 0.6
Less dimming and Reduction of latent heat flux and
increase of solar radiation increase of sensible heat flux Temperature Trend (°C/decade)

(d) Mediterranean temperature anomalies (e) Mediterranean summer vs global warming

Baseline period is 1995–2014 CMIP6 long term Baseline period is 1861–1900
temperature change
10 CRU TS CMIP5 10 8 CMIP5 (N=21)
Mediterranean summer temperature (°C)

HadCRUT5 CMIP6 RCP8.5

7
8 Berkeley Earth HighResMIP 8 RCP6.0
CORDEX EUR-44 6
RCP4.5
6 CORDEX EUR-11 6 RCP2.6
5

4
4 4
(°C)

3
CMIP6 (N=31)
2 2
2 SSP5-8.5
1
SSP3-7.0
0 0
SSP1-2.6 SSP2-4.5
SSP2-4.5 0 SSP1-2.6
–2 Long term SSP3-7.0
SSP5-8.5 0 1 2 3 4
2000 2020 2040 2060 2080 2100 Global temperature (°C)

Box TS.12, Figure 1 | Example of generating regional climate information from multiple lines of evidence for the case of Mediterranean
summer warming.

112
 Technical Summary

Box TS.12 (continued)

Box TS.12, Figure 1 (continued): The intent of this figure is to provide an example of using different lines of evidence to assess the confidence in or likelihood
of a projected change in regional climate and which of these lines of evidence are available to view and explore in the Interactive Atlas. (a) Mechanisms and
feedbacks involved in enhanced Mediterranean summer warming. (b) Locations of observing stations from different datasets. (c) Distribution of 1960–2014 summer
temperature trends (°C per decade) for observations (black crosses), CMIP5 (blue circles), CMIP6 (red circles), HighResMIP (orange circles), CORDEX EUR-44 (light
blue circles), CORDEX EUR-11 (green circles), and selected single model initial-condition large ensembles (SMILEs; grey boxplots, MIROC6, CSIRO-Mk3-6-0, MPI-ESM
and d4PDF). (d) Time series of area averaged (25°N–50°N, 10°W–40°E) land point summer temperature anomalies (°C, baseline period is 1995–2014): the boxplot
shows long term (2081–2100) temperature changes of different CMIP6 scenarios in respect to the baseline period. (e) Projected Mediterranean summer warming
in comparison to global annual mean warming of CMIP5 (RCP2.6, RCP4.5, RCP6.0 and RCP8.5) and CMIP6 (SSP1-2.6, SSP2-4.5, SSP3-7.0 and SSP5-8.5) ensemble
means (lines) and spread (shading). {Figure 10.20, Figure 10.21, Figure Atlas.8}

TS.4.2 Drivers of Regional Climate Variability temperature is moderated or amplified by soil moisture feedback,
and Change snow/ice-albedo feedback, regional forcing from land-use/land-
cover changes, forcing from aerosol concentrations, or decadal/ TS
Anthropogenic forcing, including GHGs and aerosols, but multi-decadal natural variability. Changes in local and remote
also regional land use and irrigation have all affected aerosol forcings lead to south–north gradients of the effective
observed regional climate changes (high confidence) and radiative forcing (hemispherical asymmetry). Along latitudes,
will continue to do so in the future (high confidence), with it is more uniform, with strong amplification of the temperature
various degrees of influence and response times, depending response towards the Arctic (medium confidence). The decrease
on warming levels, the nature of the forcing and the relative of SO2 emissions since the 1980s reduces the damping effect of
importance of internal variability. aerosols, leading to a faster increase in surface air temperature
that is most pronounced at mid- and high latitudes of the Northern
Since the late 19th century, major modes of variability Hemisphere, where the largest emissions reductions have taken
(MoVs) exhibited fluctuations in frequency and magnitude place (medium confidence). {1.3, 3.4.1, 6.3.4, 6.4.1, 6.4.3, 8.3.1,
at multi-decadal time scales, but no sustained trends 8.3.2, Box 8.1, 10.4.2, 10.6, 11.1.6, 11.3}
outside the range of internal variability (Table TS.4). An
exception is the Southern Annular Mode (SAM), which has Multi-decadal dimming and brightening trends in incoming solar
become systematically more positive (high confidence) and radiation at Earth’s surface occurred at widespread locations (high
is projected to be more positive in all seasons, except for confidence). Multi-decadal variation in anthropogenic aerosol
December–January–February (DJF), in high CO2 emissions emissions are thought to be a major contributor (medium confidence),
scenarios (high confidence). The influence of stratospheric but multi-decadal variability in cloudiness may also have played a
ozone forcing on the SAM trend has been reduced since the role. Volcanic eruptions affect regional climate through their spatially
early 2000s compared to earlier decades, contributing to heterogeneous effect on the radiative budget as well as through
the weakening of its positive trend as observed over 2000– triggering dynamical responses by favouring a given phase from
2019 (medium confidence). some MoVs, for instance. {1.4.1, Cross-Chapter Box 1.2, 2.2.1, 2.2.2,
3.7.1, 3.7.3, 4.3.1, 4.4.1, 4.4.4, Cross-Chapter Box 4.1, 7.2.2, 8.5.2,
In the near term, projected changes in most of the MoVs 10.1.4, 11.1.6, 11.3.1}
and related teleconnections will likely be dominated by
internal variability. In the long term, it is very likely that Historical urbanization affects the observed warming trends in cities
the precipitation variance related to El Niño–Southern and their surroundings (very high confidence). Future urbanization
Oscillation will increase. Physical climate storylines, will amplify the projected air temperature under different background
including the complex interplay between climate drivers, climates, with a strong effect on minimum temperatures that could
MoVs, and local and remote forcing, increase confidence be as large as the global warming signal (very high confidence)
in the understanding and use of observed and projected (Box TS.14). Irrigation and crop expansion have attenuated increases
regional changes. {2.4, 3.7, 4.3, 4.4, 4.5, 6.4, 8.3, 8.4, 10.3, in summer hot extremes in some regions, such as central North
10.4, 11.3} America (medium confidence) (Box TS.6). {Box 10.3, 11.1.6, 11.3}

TS.4.2.2 Modes of Variability and Regional Teleconnections

TS.4.2.1 Regional Fingerprints of Anthropogenic
and Natural Forcing Modes of variability (Annex IV, Table TS.4) have existed for millennia
or longer (high confidence), but there is low confidence in detailed
While anthropogenic forcing has contributed to multi-decadal reconstructions of most of them prior to direct instrumental records.
mean precipitation changes in several regions, internal variability MoVs are treated as a main source of uncertainties associated with
can delay emergence of the anthropogenic signal in long-term internal dynamics, as they can either accentuate or dampen, even
precipitation changes in many land regions (high confidence). At the mask, the anthropogenically forced responses. {2.4, 8.5.2, 10.4, 10.6,
regional scale, the effect of human-induced GHG forcing on extreme 11.1.5, Atlas.3.1}

113
 Technical Summary

Since the late 19th century, major MoVs (Table TS.4) show contributed to observed temporal evolution in the Atlantic Multi-
no sustained trends, exhibiting fluctuations in frequency and decadal Variability (AMV) and associated regional teleconnections,
magnitude at multi-decadal time scales, except for the Southern especially since the 1960s, but there is low confidence in the
Annular Mode (SAM), which has become systematically more magnitude of this influence and the relative contributions of
positive (high confidence) (Table TS.4). It is very likely that human natural and anthropogenic forcings. Internal variability is the
influence has contributed to this trend from the 1970s to the main driver of Pacific Decadal Variability (PDV) observed since
1990s, and to the associated strengthening and southward shift the start of the instrumental records (high confidence), despite
of the Southern Hemispheric extratropical jet in austral summer. some modelling evidence for potential external influence. There is
The influence of stratospheric ozone forcing on the SAM trend has medium confidence that the AMV will undergo a shift towards a
been reduced since the early 2000s compared to earlier decades, negative phase in the near term. {2.4, 3.7.6, 3.7.7, 8.5.2, 4.4.3}
contributing to the weakening of its positive trend observed over
2000–2019 (medium confidence). By contrast, the cause of the
Northern Annular Mode (NAM) trend toward its positive phase
since the 1960s and associated northward shifts of Northern
TS Hemispheric extratropical jet and storm track in boreal winter
is not well understood. The evaluation of model performance on
simulating MoVs is assessed in Section TS.1.2.2. {2.3.3, 2.4, 3.3.3,
3.7.1, 3.7.2}

In the near term, the forced change in SAM in austral summer is likely
to be weaker than observed during the late 20th century under all
five SSPs assessed. This is because of the opposing influence in the
near to mid-term from stratospheric ozone recovery and increases in
other greenhouse gases on the Southern Hemisphere summertime
mid-latitude circulation (high confidence). In the near term, forced
changes in the SAM in austral summer are therefore likely to be
smaller than changes due to natural internal variability. In the long
term (2081–2100) under the SSP5-8.5 scenario, the SAM index is
likely to increase in all seasons relative to 1995–2014. The CMIP6
multi-model ensemble projects a long-term (2081–2100) increase
in the boreal wintertime NAM index under SSP3-7.0 and SSP5-8.5,
but regional associated changes may deviate from a simple shift
in the mid-latitude circulation due to a modified teleconnection
resulting from interaction with a modified mean background state.
{4.3.3, 4.4.3, 4.5.1, 4.5.3, 8.4.2}

Human influence has not affected the principal tropical modes of

interannual climate variability (Table TS.4) and their associated
regional teleconnections beyond the range of internal variability
(high confidence). It is virtually certain that the El Niño–Southern
Oscillation (ENSO) will remain the dominant mode of interannual
variability in a warmer world. There is no consensus from models for
a systematic change in amplitude of ENSO sea surface temperature
(SST) variability over the 21st century in any of the SSP scenarios
assessed (medium confidence). However, it is very likely that
rainfall variability related to ENSO will be enhanced significantly
by the latter half of the 21st century in the SSP2-4.5, SSP3-7.0
and SSP5-8.5 scenarios, regardless of the amplitude changes in
SST variability related to the mode. It is very likely that rainfall
variability related to changes in the strength and spatial extent of
ENSO teleconnections will lead to significant changes at regional
scale. {3.7.3, 3.7.4, 3.7.5, 4.3.3, 4.5.3, 8.4.2, 10.3.3}

Modes of decadal and multi-decadal variability over the Pacific

and Atlantic Ocean exhibit no significant changes in variance
over the period of observational records (high confidence). There
is medium confidence that anthropogenic and volcanic aerosols

114
 Technical Summary

Table TS.4 | Summary of the assessments on modes of variability (MoVs) and associated teleconnections. (a) Assessments on observed changes since the start
of instrumental records, Coupled Model Intercomparison Project Phases 5 and 6 (CMIP5 and CMIP6) model performance, human influence on the observed changes, and
near-term (2021–2040) and mid- to long-term (2041–2100) changes. Curves schematically illustrate the assessed overall changes, with the horizontal axis indicating time,
and are not intended to precisely represent the time evolution. (b) Fraction of surface air temperature (SAT) and precipitation (pr) variance explained at interannual time scale
by each MoV for each AR6 region (numbers in each cell; in percent). Values correspond to the average of significant explained variance fractions based on HadCRUT, GISTEMP,
BerkeleyEarth and CRU-TS (for SAT) and GPCC and CRU-TS (for precipitation). Significance is tested based on F-statistics at the 95% level confidence, and a slash indicates that
the value is not significant in more than half of the available data sets. The colour scale corresponds to the sign and values of the explained variance as shown at the bottom. The
corresponding anomaly maps are shown in Annex IV. DJF: December–January–February. MAM: March–April–May. JJA: June–July–August. SON: September–October–November.
In (b), Northern Annular Mode (NAM) and El Niño–Southern Oscillation (ENSO) teleconnections are evaluated for 1959–2019, Southern Annular Mode (SAM) for 1979–2019,
Indian Ocean Basin (IOB), Indian Ocean Dipole (IOD), Atlantic Zonal Mode (AZM) and Atlantic Meridional Mode (AMM) for 1958–2019, and Pacific Decadal Variability (PDV)
and Atlantic Multi-decadal Variability (AMV) for 1900–2019. All data are linearly detrended prior to computation. (Section TS.1.2.2) {2.4, 3.7, 4.3.3, 4.4.3, 4.5.3, Table Atlas.1,
Annex IV}
(a) Assessments on MoV.

NAM SAM ENSO IOB IOD AZM AMM PDV AMV

1990s 2000s
1960s
1950s Within proxy- Within proxy- Dominated Dominated
TS
inferred inferred Limited Limited by multi- by multi-
Past changes since the 2010s
variability variability evidence evidence decadal decadal
start of observations Boreal Austral
1400– Since range range fluctuations fluctuations
winter summer
1850 1950s
{2.4.1.1} {2.4.1.2} {2.4.2} {2.4.3} {2.4.3} {2.4.4} {2.4.4} {2.4.5} {2.4.6}

High High Medium Medium Medium Low Low Medium Medium

CMIP5 and CMIP6 performance performance performance performance performance performance performance performance performance
model performance
{3.7.1} {3.7.2} {3.7.3} {3.7.4} {3.7.4} {3.7.5} {3.7.5} {3.7.6} {3.7.7}

Contributed

through GHG Contributed

No robust Low No robust No robust No robust
Human influence on (all seasons) Not detected Not detected through
evidence agreement evidence evidence evidence
the observed changes & ozone aerosols
(DJF)

{3.7.1} {3.7.2} {3.7.3} {3.7.4} {3.7.4} {3.7.5} {3.7.5} {3.7.6} {3.7.7}

+
Internal Internal
No robust No robust No robust No robust Limited –
Near-term future variability variability
evidence evidence evidence evidence evidence Phase shift
changes (2021–2040) dominates All seasons dominates
except DJF from + to –

{4.4.3.1} {4.4.3.1} {4.4.3.2} {4.4.3.3} {4.4.3.3} {4.4.3.4} {4.4.3.4} {4.4.3.5} {4.4.3.6}

SSP5-8.5
SSP3-7.0
SSP5-8.5
SSP3-7.0
All
seasons No robust increase No robust No robust
Mid-to-long-term SSP1-1.9 No changes
DJF Increase in evidence in extreme evidence evidence
future changes SSP1-2.6 Decrease in
SSP1-1.9 precipitation positive
(2041–2100) variance
DJF variance events

{4.3.3.1; {4.3.3.1; {4.3.3.2;

{4.5.3.3} {4.5.3.3} {4.5.3.4} {4.5.3.4} {4.5.3.5} {4.5.3.6}
4.5.3.1} 4.5.3.1} 4.5.3.2}

low confidence medium confidence high confidence

more likely than not likely very likely

115
 Technical Summary

Table TS.4 (continued): (b) Regional climate anomalies associated with MoV.

Mode NAM SAM ENSO IOB IOD AZM AMM PDV AMV
Season DJF DJF DJF MAM SON JJA JJA annual annual
Variable SAT pr SAT pr SAT pr SAT pr SAT pr SAT pr SAT pr SAT pr SAT pr
Mediterranean 28 58 7 19
Sahara 58 14 10 19 12 9 12 25
Western Africa 25 15 45 21 10 6 6 23
Central Africa 19 8 10 14 50 13 10 14 11
Africa

North Eastern Africa 19 7 14 36 32 7 7

South Eastern Africa 14 22 36 57 10 4 9
West Southern Africa 49 26 27 16 8 4 12 5
East Southern Africa 13 75 34 35 7 4 6
Madagascar 24 24 7 11 10 9 5

TS West Siberia 45 7 9 11
East Siberia 52 3 11
Russian Far East 8 10 11 6 5 5
West Central Asia 15 21 4
East Central Asia 38
Asia

Tibetan Plateau 15 15 7 11 6 5 9
East Asia 7 20 23 9 9 13
South Asia 9 12 8 8 5
South East Asia 39 31 73 6 48 5 12 7
Arabian Peninsula 32 10 24 20 5 13 7
Northern Australia 21 13 38 19 7 7 7
Australasia

Central Australia 14 21 12 18 22 20 7 7 6 5
East Australia 22 20 11 18 9 8 7 7 8
Southern Australia 11 23 40 8 3
New Zealand 16
Southern Central America 21 16 33 10 11 17 6 6 7
Central & South America

North-Western South America 7 14 16 82 17 54 18 13 16 7 8

Northern South America 7 56 58 61 22 17 24 9 12 7
North-Eastern South America 25 58 19 9 12 8
South American Monsoon 54 31 22 7 6 7
South-Western South America 10 16 14 17 10 16 8
South-Eastern South America 21 13 21 10 12 5 6
Southern South America 23 13 7 9
Mediterranean 28 58 7 19
Europe

Western and Central Europe 28 18 13 10 4 8

Eastern Europe 35 7
Northern Europe 53 32 6
Northern Central America 10 26 13 27 18 7 12 15 12 6 19
Western North America 4 6 5
North America

Central North America 17 12 17 8 3 9 6

Eastern North America 12 11 9 4 9 4
North-Eastern North America 18 26 8 10 9 4
North-Western North America 14 10 8 17 8 4
Caribbean 10 15 18 26 8 10 17 12 7 5
Islands
Small

Pacific
Greenland/Iceland 42 8 7 44
Polar Terrestrial
Regions

Russian Arctic 25 10 6 11 8
West Antarctica 8 21
East Antarctica 38

colderwarmer drierwetter
Not significant in >50% of available data sets
40 30 20 0 20 30 40 40 30 20 0 20 30 40 Data unavailable in >50% of data sets
Temperature anomalies and Precipitation anomalies and
explained variance (%) explained variance (%)

116
 Technical Summary

TS.4.2.3 Interplay Between Drivers of Climate Variability annual to decadal time scales (high confidence). The anthropogenic
and Change at Regional Scales signal in regional sea level change will emerge in most regions by
2100 (medium confidence). {9.2.4, 9.6.1, 10.4.1, 10.4.2, 10.4.3}
Anthropogenic forcing has been a major driver of regional mean
temperature change since 1950 in many sub-continental regions of Regional climate change is subject to the complex interplay between
the world (virtually certain). At regional scales, internal variability multiple external forcings and internal variability. Time evolution
is stronger, and uncertainties in observations, models and external of mechanisms operating at different time scales can modify the
forcing are all larger than at the global scale, hindering a robust amplitude of the regional-scale response of temperature, and
assessment of the relative contributions of greenhouse gases, both the amplitude and sign of the response of precipitation, to
stratospheric ozone, and different aerosol species in most of the cases. anthropogenic forcing (high confidence). These mechanisms include
Multiple lines of evidence, combining multi-model ensemble global non-linear temperature, precipitation and soil moisture feedbacks;
projections with those coming from single-model initial-condition slow and fast responses of SST patterns; and atmospheric circulation
large ensembles, show that internal variability is largely contributing changes to increasing GHGs. Land-use and aerosol forcings and
to the delayed or absent emergence of the anthropogenic signal in land–atmosphere feedback play important roles in modulating
long-term regional mean precipitation changes (high confidence). regional changes, for instance in weather and climate extremes (high TS
Internal variability in ocean dynamics dominates regional patterns on confidence). These can also lead to a higher warming of extreme

Pathway to understanding past and assessing future climate changes at regional scale
The South-Eastern South America (SES) case study
(a) Identification of climate drivers and phenomena for interpreting (b) Models simulations/evaluation of SES DJF precipitation over
SES observed precipitation trend and variability in austral summer (DJF) the historical period and 21st century based on 7 large ensembles

Aerosol GPCC observations

Anomalies in mm/month w.r.t 1995-2014

AMV
Had
ley C
ell

Near Mid Long

Term Term Term
ENSO
SES
PDV
GHG
Rossby + O3 depletion
83.6% 53.3% 14.7% 6.3%
Waves
Internal variability
uncertainty
lar Model uncertainty
Precipitation trend (GPCC/1950-2014) Po tex 5.0%
r
Vo
(mm/month/decade)
Scenario uncertainty 1.4% 41.7% 78.9%
14.9%
Near term Mid term Long term

Figure TS.21 | Example of the interplay between drivers of climate variability and change at regional scale to understand past and projected changes. The
figure intent is to show an illustrative pathway for understanding past, and anticipating future, climate change at regional scale in the presence of uncertainties. (a) Identification
of the climate drivers and their influences on climate phenomena contributing through teleconnection to South-Eastern South America (SES) summer (December–January–
February; DJF) precipitation variability and trends observed over 1950–2014. Drivers (red squares) include modes of variability as well as external forcing. Observed precipitation
linear trend from GPCC is shown on continents (green-brown colour bar in mm month–1 per decade) and the SES AR6 WGI reference region is outlined with the thick black
contour. Climate phenomena leading to local effects on SES are schematically presented (blue ovals). (b) Time series of decadal precipitation anomalies for DJF SES simulated
from seven large ensembles of historical plus RCP8.5 simulations over 1950–2100. Shading corresponds to the 5–95th range of climate outcomes given from each large
ensemble for precipitation (in mm month –1) and thick coloured lines stand for their respective ensemble mean. The thick time series in white corresponds to the multi-model
multi-member ensemble mean, with model contribution being weighted according to their ensemble size. GPCC observation is shown in the light black line with squares over
1950–2014, and the 1995–2014 baseline period has been retained for calculation of anomalies in all datasets. (c) Quantification of the respective weight (in percent) between
the individual sources of uncertainties (internal in grey, model in magenta and scenario in green) at near-term, mid-term and long-term temporal windows defined in AR6 and
highlighted in (b) for SES DJF precipitation. All computations are done with respect to 1995–2014, taken as the reference period, and the scenario uncertainty is estimated
from Coupled Model Intercomparison Project Phase 5 (CMIP5) using the same set of models as for the large ensembles that have run different Representative Concentration
Pathway (RCP) scenarios. {Figure 10.12a}

117
 Technical Summary

temperatures compared to mean temperature (high confidence), which are relevant for the region. In fact, local changes over SES
and possibly cooling in some regions (medium confidence). The in terms of moisture convergence, ascending motion and storm-
soil moisture–temperature feedback was shown to be relevant for track locations depend on these climate phenomena, and they are
past and present-day heatwaves based on observations and model overall responsible for the observed precipitation trends. Projections
simulations. {10.4.3, 11.1.6, 11.3.1} suggest continuing positive trends in rainfall over SES in the near-
term in response to GHG emissions scenarios. Multi-model mean and
South-Eastern South America (SES) is one of the AR6 WGI reference ensemble spread are not sufficient to characterize situations where
regions (outlined with black thick contour in Figure TS.21a), and different models simulate substantially different or even opposite
it is used here as an illustrative example of the interplay between changes (high confidence). In such cases, physical climate storylines
drivers of climate variability and change at regional scale. Austral addressing possible outcomes for climate phenomena shown to
summer (DJF) precipitation positive trends have been observed play a role in the variability of the region of interest can aid the
over the region during 1950–2014. Drivers of this change include interpretation of projection uncertainties. In addition, single-model
MoVs, such as AMV, ENSO, and PDV, as well as external forcing, initial-condition large ensembles of many realizations of internal
like GHG increases and ozone depletion together with aerosols (as variability are required to separate internal variability from forced
TS illustrated in Figure TS.21a). Modes of variability and external forcing changes (high confidence) and to partition the different sources of
collectively affect climate phenomena, such as the Hadley cell width uncertainties as a function of future assessed periods. {10.3.4, 10.4.2,
and strength, Rossby waves activity emerging from the large-scale Figure 10.12a}
tropical SST anomalies, and the Southern Hemisphere polar vortex,

Box TS.13 | Monsoons

Global land monsoon precipitation decreased from the 1950s to the 1980s, partly due to anthropogenic aerosols, but
has increased since then in response to GHG forcing and large-scale multi-decadal variability (medium confidence).
Northern Hemispheric anthropogenic aerosols weakened the regional monsoon circulations in South Asia, East Asia
and West Africa during the second half of the 20th century, thereby offsetting the expected strengthening of monsoon
precipitation in response to GHG-induced warming (high confidence).

During the 21st century, global land monsoon precipitation is projected to increase in response to GHG warming in
all time horizons and scenarios (high confidence). Over South and South East Asia, East Asia and the central Sahel,
monsoon precipitation is projected to increase, whereas over North America and the far western Sahel it is projected
to decrease (medium confidence). There is low confidence in projected precipitation changes in the South American
and Australian-Maritime Continent monsoons. At global and regional scales, near-term monsoon changes will be
dominated by the effects of internal variability (medium confidence). {2.3, Cross-Chapter Box 2.4, 3.3, 4.4, 4.5, 8.2,
8.3, 8.4, 8.5, Box 8.1, Box 8.2, 10.6}

Global Monsoon
Paleoclimate records indicate that during warm climates, like the mid-Pliocene Warm Period, monsoon systems were stronger
(medium confidence). In the instrumental records, global summer monsoon precipitation intensity has likely increased since the
1980s, dominated by Northern Hemisphere summer trends and large multi-decadal variability. Contrary to the expected increase of
precipitation under global warming, the Northern Hemisphere monsoon regions experienced declining precipitation from the 1950s
to 1980s, which is partly attributable to the influence of anthropogenic aerosols (medium confidence) (Box TS.13, Figure 1). {2.3.1,
Cross-Chapter Box 2.4, 3.3.2, 3.3.3}

With continued global warming, it is likely that global land monsoon precipitation will increase during this century (Box TS.13, Figure 1),
particularly in the Northern Hemisphere, although the monsoon circulation is projected to weaken. A slowdown of the tropical circulation
with global warming can partly offset the warming-induced strengthening of precipitation in monsoon regions (high confidence). In the
near term, global monsoon changes are likely to be dominated by the effects of internal variability and model uncertainties (medium
confidence). In the long term, global monsoon rainfall change will feature a robust north–south asymmetry characterized by a greater
increase in the Northern Hemisphere than in the Southern Hemisphere and an east–west asymmetry characterized by enhanced Asian–
African monsoons and a weakened North American monsoon (medium confidence). {4.4.1, 4.5.1, 8.4.1}

Regional Monsoons
Paleoclimate reconstructions indicate stronger monsoons in the Northern Hemisphere but weaker ones in the Southern Hemisphere
during warm periods, particularly for the South and South East Asian, East Asian, and North and South American monsoons, with
the opposite occurring during cold periods (medium confidence). It is very likely that Northern Hemispheric anthropogenic aerosols
weakened the regional monsoon circulations in South Asia, East Asia and West Africa during the second half of the 20th century,

118
 Technical Summary

Box TS.13 (continued)

(a) Global and regional monsoon domains

2014

NAmerM SAsia M
EAsia M
WAfriM
EqAmer

SAfri
SAmer M AusMCM
TS

(b) Historical trend in monsoon precipitation

Trend in precipitation (mm/day/decade)

Simulations: NAT ALL GHG AER

0.12 Observations: x APHRO x CRU x GPCP
x
0.08 x
x

0.04 xx
x

0 x
x
xx
x
x
x x
–0.04 x
x

–0.08 x
x

–0.12
NAmerM WA fri M SAsiaM EAsiaM EqAmer SAmerM SAfri AusMCM GM

(c) Projected future change in monsoon precipitation (SSP2-4.5)

20
Change in precipitation (%)

10

0
Long term
Near term
Mid term

–10

–20
NAmerM WA fri M SAsiaM EAsiaM EqAmer SAmerM SAfri AusMCM GM

Box TS.13, Figure 1 | Global and regional monsoons: past trends and projected changes. The intent of this figure is to show changes in precipitation over
regional monsoon domains in terms of observed past trends, how greenhouse gases and aerosols relate to these changes, and in terms of future projections in one
intermediate emissions scenario in the near, medium and long term. (a) Global (black contour) and regional monsoons (colour shaded) domains. The global monsoon
(GM) is defined as the area with local summer-minus-winter precipitation rate exceeding 2.5 mm day–1 (see Annex V). The regional monsoon domains are defined
based on published literature and expert judgement (see Annex V) and accounting for the fact that the climatological summer monsoon rainy season varies across
the individual regions. Assessed regional monsoons are South and South East Asia (SAsiaM, Jun–July–August–September), East Asia (EAsiaM, June–July–August),
West Africa (WAfriM, June–July–August–September), North America (NAmerM, July–August–-September), South America (SAmerM, December–January–February),
Australia and Maritime Continent Monsoon (AusMCM, December–January–February). Equatorial South America (EqSAmer) and South Africa (SAfri) regions are also
shown, as they receive unimodal summer seasonal rainfall although their qualification as monsoons is subject to discussion. (b) Global and regional monsoons
precipitation trends based on DAMIP CMIP6 simulations with both natural and anthropogenic (ALL), greenhouse gas only (GHG), aerosols only (AER) and natural
only (NAT) radiative forcing. Weighted ensemble means are based on nine Coupled model Intercomparison Project Phase 6 (CMIP6) models contributing to the MIP
(with at least three members). Observed trends computed from CRU, GPCP and APHRO (only for SAsiaM and EAsiaM) datasets are shown as well. (c) Percentage
change in projected seasonal mean precipitation over global and regional monsoons domain in the near term (2021–2040), mid-term (2041–2060), and long term
(2081–2100) under SSP2-4.5 based on 24 CMIP6 models. {Figures 8.11 and 8.22}

119
 Technical Summary

Box TS.13 (continued)

thereby offsetting the expected strengthening of monsoon precipitation in response to GHG-induced warming (Box TS.13, Figure 1).
Multiple lines of evidence explain this contrast over South Asia, with the observed trends dominated by the effects of aerosols, while
future projections are mostly driven by GHG increases. The recent partial recovery and enhanced intensity of monsoon precipitation
over West Africa is related to the growing influence of GHGs with an additional contribution due to the reduced cooling effect of
anthropogenic aerosols, emitted largely from North America and Europe (medium confidence). For other regional monsoons, that
is, North and South America and Australia, there is low confidence in the attribution of recent changes in precipitation (Box TS.13,
Figure 1) and winds. {2.3.1, 8.3.1, 8.3.2, Box 8.1, 10.6.3}

Projections of regional monsoons during the 21st century indicate contrasting (region-dependent) and uncertain precipitation and
circulation changes. The annual contrast between the wettest and driest month of the year is likely to increase by 3–5% per degree
Celsius in most monsoon regions in terms of precipitation, precipitation minus evaporation, and runoff (medium confidence). For the
TS North American monsoon, projections indicate a decrease in precipitation, whereas increased monsoon rainfall is projected over South
and South East Asia and over East Asia (medium confidence) (Box TS.13, Figure 1). West African monsoon precipitation is projected
to increase over the central Sahel and decrease over the far western Sahel (medium confidence). There is low confidence in projected
precipitation changes in the South American and Australian-Maritime Continent regional monsoons (for both magnitude and sign)
(Box TS.13, Figure 1). There is medium confidence that the monsoon season will be delayed in the Sahel and high confidence that it
will be delayed in North and South America. {8.2.2, 8.4.2.4, Box 8.2}

Building the Assessment from Multiple Lines of Evidence

Large natural variability of monsoon precipitation across different time scales, found in both paleoclimate reconstructions and
instrumental measurements, poses an inherent challenge for robust quantification of future changes in precipitation at regional and
smaller spatial scales. At both global and regional scales, there is medium confidence that internal variability contributes the largest
uncertainty related to projected changes, at least in the near term (2021–2040). A collapse of the Atlantic Meridional Overturning
Circulation could weaken the African and Asian monsoons but strengthen the Southern Hemisphere monsoons (high confidence).
{4.4.4, 4.5.1, Cross-Chapter Box 4.1, 8.5.2, 8.6.1, 9.2.3, 10.6.3}

Overall, long-term (2081–2100) future changes in regional monsoons like the South and South East Asian monsoon are generally
consistent across global (including high-resolution) and regional climate models and are supported by theoretical arguments.
Uncertainties in simulating the observed characteristics of regional monsoon precipitation are related to varying complexities of
regional monsoon processes and their responses to external forcing, internal variability, and deficiencies in representing monsoon warm
rain processes, organized tropical convection, heavy orographic rainfall and cloud–aerosol interactions. {8.3.2, 8.5.1, 10.3.3, 10.6.3}

TS.4.3 Regional Climate Change and Implications for Relative sea level rise is very likely to virtually certain
Climate Extremes and Climatic Impact-Drivers (depending on the region) to continue during the 21st
century, contributing to increased coastal flooding in low-
Current climate in all regions is already distinct from the lying areas (high confidence) and coastal erosion along
climate of the early or mid-20th century with respect to most sandy coasts (high confidence). Sea level will continue
several climatic impact-drivers (CIDs), resulting in shifting to rise beyond 2100 (high confidence) (Box TS.4).
magnitude, frequency, duration, seasonality and spatial
extent of associated climate indices (high confidence). Every region of the world will experience concurrent
It is very likely that mean temperatures have increased changes in multiple CIDs by mid-century or at 2°C global
in all land regions and will continue to increase at rates warming and above (high confidence). Even for the current
greater than the global average (high confidence). The climate, climate change-induced shifts in CID distributions
frequency of heat and cold extremes have increased and and event probabilities, some of which have occurred over
decreased, respectively. These changes are attributed to recent decades, are relevant for risk assessments. {11.9,
human influence in almost all regions (medium to high 12.1, 12.2, 12.4, 12.5, Atlas.3–Atlas.11}
confidence) and will continue through the 21st century
(high confidence). In particular, extreme heat would exceed An overview of changes in regional CIDs (introduced in Section TS.1)
critical thresholds for health, agriculture and other sectors is given in Table TS.5, which summarizes multiple lines of evidence
more frequently by the mid 21st century with 2°C of global on regional climate change derived from observed trends, attribution
warming (high confidence). of these trends and future projections. The level of confidence and

120
 Technical Summary

the amplitude in the projected direction of change in CIDs at a given

time horizon depends on climate change mitigation efforts over the
21st century. It is evident from Table TS.5 that many heat, cold, snow
and ice, coastal, and oceanic CID changes are projected with high
confidence in most regions starting from a global warming level
(GWL) of 2°C, indicating worldwide challenges. Changes in many
other regional CIDs have higher confidence later in the 21st century
or at higher GWLs (high confidence), and another small subset are
projected with high confidence for the 1.5°C GWL. This section
focuses on the 2°C GWL and mid-century time period because the
signal emerges from natural variability for a wider range of CIDs
at this higher warming level. Figure TS.22 shows the geographical
location of regions belonging to one of five groups characterized
by a specific combination of changing CIDs. The Regional Synthesis
component of the Interactive Atlas provides comprehensive synthesis TS
information about changes in all of the individual CIDs across all of
the AR6 WGI reference regions. {10.5, Cross-Chapter Box 10.3, 11.1,
11.9, Box 11.1, 12.1, 12.2, 12.4, 12.5}

121
 TS
Table TS.5 | Summary of confidence for climatic impact-driver changes in each AR6 WGI reference region (illustrated in Figure TS.25) across multiple lines of evidence for observed, attributed and projected

Technical Summary
122

directional changes. The colours represent their projected aggregate characteristic changes for the mid-21st century, considering scenarios RCP4.5, SSP2-4.5, SRES A1B, or above (RCP6.0, RCP8.5, SSP3-7.0, SSP5-8.5, SRES A2), which
approximately encompasses global warming levels of 2.0°C to 2.4°C. Arrows indicate medium to high confidence trends derived from observations, and asterisks indicate medium and high confidence in attribution of observed changes. (North
Africa is not an AR6 WGI reference region, but assessment here is based upon the African portion of the Mediterranean reference region). {Tables 12.3–12.11 and Tables 11.4–11.21}

Climatic Impact-driver
Mean air temperature Heat and Cold Wet and Dry Wind Snow and Ice Coastal and Oceanic Other

Extreme heat

Cold spell

Frost

Mean precipitation

River flood

pluvial flood
Heavy precipitation and

Landslide

Aridity

Hydrological drought

Agricultural and ecological drought

Fire weather

Mean wind speed

Severe wind storm

Tropical cyclone

Sand and dust storm

Snow, glacier and ice sheet

Permafrost

Lake, river and sea ice

Heavy snowfall and ice storm

Hail

Snow avalanche

Relative sea level

Coastal flood

Coastal erosion

Marine heatwave

Ocean and lake acidity

Air pollution weather

Atmospheric CO2 at surface

Radiation at surface
Africa

North Africa ↗ ↗ *** ↘ *** ↗ ↗ ↗ 3 ↗ 4 ↗ ↗ ↗

Sahara ↗ ↗ ** ↘ ** ↗ 4 ↗ ↗ ↗

Western Africa ↗ ↗ ** ↘ ** 1 ↗ ↗1 ↗1 ↗1 ↗ 4 ↗ ↗ ↗

Central Africa ↗ ↘ 1,2 ↗ ↗ 4 ↗ ↗ ↗

North Eastern
Africa ↗ ↗ ** ↘ ** ↘ 1 1 1 ↘ ↗ 4 ↗ ↗ ↗
South Eastern
Africa ↗ ↗ ** 1 1 1 3 ↘ ↗ 4 ↗ ↗ ↗
West Southern
Africa ↗ ↗ *** ↘ *** ↘ ↗ ↗ ↗ ↗ 4 ↗ ↗ ↗
East Southern
Africa ↗ ↗ *** ↘ *** ↘ ↗ ↗ ↗ 3 ↗ 4,5 ↗ ↗ ↗

Madagascar ↗ ↗ ↘ 3 ↗ 4,5 ↗ ↗ ↗

Note: There are several region-specific qualifiers/exceptions attached to some of the directions of change/confidence levels indicated above. {12.4}
Key for observational trend evidence ↗ Past upward trend (medium or higher confidence) ↘ Past downward trend (medium or higher confidence)
Key for attribution evidence *** High confidence (or more) ** Medium confidence
High confidence Medium confidence Low confidence Medium confidence High confidence Not broadly
Key for level of confidence in future changes
of increase (or more) of increase (or more) in direction of change of decrease of decrease relevant
 Table TS.5 (continued)

Climatic Impact-driver

Heat and Cold Wet and Dry Wind Snow and Ice Coastal and Oceanic Other

Mean air temperature

Extreme heat

Cold spell

Frost

Mean precipitation

River flood

pluvial flood
Heavy precipitation and

Landslide

Aridity

Hydrological drought

Agricultural and ecological drought

Fire weather

Mean wind speed

Severe wind storm

Tropical cyclone

Sand and dust storm

Snow, glacier and ice sheet

Permafrost

Lake, river and sea ice

Heavy snowfall and ice storm

Hail

Snow avalanche

Relative sea level

Coastal flood

Coastal erosion

Marine heatwave

Ocean and lake acidity

Air pollution weather

Atmospheric CO2 at surface

Radiation at surface
Asia

Arabian Peninsula ↗ ↗ *** ↘ ** ↘ ↗ 1 ↗ ↗

West Central Asia ↗ ↗ *** ↘ *** ↘ 5 ↗ ↗ ↗ ↘ ↘ ↗ 1,2 ↗ ↗

West Siberia ↗ ↗ *** ↘ *** ↘ ↗ ↗ ↘ ↗ ↘ ↗

East Siberia ↗ ↗ *** ↘ *** ↘ ↗ ↗ ↘ ↗ ↘ ↗

Russian Far East ↗ ↗ *** ↘ *** ↘ ↗ ↗ ↗ ↘ ↗ 1,2 ↗ ↗ ↗

East Asia ↗ ↗ *** ↘ *** ↘ ↗ ↗ ↗ ↘ ↗3 ↘ ↗ 1,2 ↗ ↗

East Central Asia ↗ ↗ *** ↘ *** ↗ ↗ ↘ ↘ ↘

Tibetan Plateau ↗ ↗ *** ↘ *** ↘ ↗ ↘ ↘ ↗

South Asia ↗ ↗ *** ↘ *** ↘ ↘ ↗ ↘ ↘ ↘ ↗ 1 ↗

South East Asia ↗ ↗ *** ↘ *** 4 ↗ ↗3 ↗ 1,2 ↗ ↗

Technical Summary
Note: There are several region-specific qualifiers/exceptions attached to some of the directions of change/confidence levels indicated above. {12.4}
Key for observational trend evidence ↗ Past upward trend (medium or higher confidence) ↘ Past downward trend (medium or higher confidence)
Key for attribution evidence *** High confidence (or more) ** Medium confidence
High confidence Medium confidence Low confidence Medium confidence High confidence Not broadly
Key for level of confidence in future changes
of increase (or more) of increase (or more) in direction of change of decrease of decrease relevant
123

TS
 TS
Table TS.5 (continued)

Technical Summary
124

Climatic Impact-driver

Heat and Cold Wet and Dry Wind Snow and Ice Coastal and Oceanic Other

Mean air temperature

Extreme heat

Cold spell

Frost

Mean precipitation

River flood

pluvial flood
Heavy precipitation and

Landslide

Aridity

Hydrological drought

Agricultural and ecological drought

Fire weather

Mean wind speed

Severe wind storm

Tropical cyclone

Sand and dust storm

Snow, glacier and ice sheet

Permafrost

Lake, river and sea ice

Heavy snowfall and ice storm

Hail

Snow avalanche

Relative sea level

Coastal flood

Coastal erosion

Marine heatwave

Ocean and lake acidity

Air pollution weather

Atmospheric CO2 at surface

Radiation at surface
Australasia

Northern Australia ↗ *** ↘ *** ↘ ↗ ↘ ↘5 ↗ 7 ↗ ↗ ↗

Central Australia ↗ ↗ *** ↘ *** ↘ ↗ 7 ↗ ↗ ↗

Eastern Australia ↗ ↗ *** ↘ *** ↘ ↗ 7 ↗ ↗ ↗

Southern Australia ↗ ↗ *** ↘ *** ↘ 1 ↗3 ↗ ↗ ↗ ** 7 ↘ ↗ 7 ↗ ↗ ↗

New Zealand ↗ ↗ ↘ ** 2 4 8 ↘6 ↗ 7 ↗ ↗ ↗

Note: There are several region-specific qualifiers/exceptions attached to some of the directions of change/confidence levels indicated above. {12.4}
Key for observational trend evidence ↗ Past upward trend (medium or higher confidence) ↘ Past downward trend (medium or higher confidence)
Key for attribution evidence *** High confidence (or more) ** Medium confidence
High confidence Medium confidence Low confidence Medium confidence High confidence Not broadly
Key for level of confidence in future changes
of increase (or more) of increase (or more) in direction of change of decrease of decrease relevant
 Table TS.5 (continued)

Climatic Impact-driver

Heat and Cold Wet and Dry Wind Snow and Ice Coastal and Oceanic Other
Mean air temperature

Extreme heat

Cold spell

Frost

Mean precipitation

River flood

pluvial flood
Heavy precipitation and

Landslide

Aridity

Hydrological drought

Agricultural and ecological drought

Fire weather

Mean wind speed

Severe wind storm

Tropical cyclone

Sand and dust storm

Snow, glacier and ice sheet

Permafrost

Lake, river and sea ice

Heavy snowfall and ice storm

Hail

Snow avalanche

Relative sea level

Coastal flood

Coastal erosion

Marine heatwave

Ocean and lake acidity

Air pollution weather

Atmospheric CO2 at surface

Radiation at surface
Central and South America

Southern Central
America ↗ ↗ ** ↘ ** 2 ↗ 3 ↗ ↗ ↗
North-Western
South America ↗ ↗ *** ↘ *** ↗ 3,4 ↗ ↗ ↗
Northern South
America ↗ ↗ ** ↘ ** 2 ↗ 3,4 ↗ ↗ ↗
South American
Monsoon ↗ ↗ ** ↘ ** ↗1 ↗
North-Eastern
South America ↗ ↗ ** ↘ ** ↘ ↗ ↗ 3,4 ↗ ↗ ↗
South-Western
South America ↗ ↗ ** ↘ ** ↘ ↗ ↗ 3 ↗ ↗ ↗
South-Eastern
South America ↗ ↗ *** ↘ *** ↘ ↗ ↗ ↘ ↗ 3 ↗ ↗ ↗
Southern South
America ↗ ↘ ↗ 3 ↗ ↗ ↗

Note: There are several region-specific qualifiers/exceptions attached to some of the directions of change/confidence levels indicated above. {12.4}
Key for observational trend evidence ↗ Past upward trend (medium or higher confidence) ↘ Past downward trend (medium or higher confidence)
Key for attribution evidence *** High confidence (or more) ** Medium confidence

Technical Summary
High confidence Medium confidence Low confidence Medium confidence High confidence Not broadly
Key for level of confidence in future changes
of increase (or more) of increase (or more) in direction of change of decrease of decrease relevant
125

TS
 TS
Table TS.5 (continued)

Technical Summary
126

Climatic Impact-driver

Heat and Cold Wet and Dry Wind Snow and Ice Coastal and Oceanic Other

Mean air temperature

Extreme heat

Cold spell

Frost

Mean precipitation

River flood

pluvial flood
Heavy precipitation and

Landslide

Aridity

Hydrological drought

Agricultural and ecological drought

Fire weather

Mean wind speed

Severe wind storm

Tropical cyclone

Sand and dust storm

Snow, glacier and ice sheet

Permafrost

Lake, river and sea ice

Heavy snowfall and ice storm

Hail

Snow avalanche

Relative sea level

Coastal flood

Coastal erosion

Marine heatwave

Ocean and lake acidity

Air pollution weather

Atmospheric CO2 at surface

Radiation at surface
Europe

Mediterranean ↗ ↗ *** ↘ *** ↘ 5 ↗ ** ↗ ** ↘6 7 ↘ ↗ 2 ↗ ↗ ↗

Western and Central
Europe ↗ ↗ *** ↘ *** ↗ ↗ ↗ 4 ↗ ↘ ↘ ↘ ↗ 2 ↗ ↗ ↗

Eastern Europe ↗ ↗ *** ↘ *** ↗ ↗ ↘ ↘ ↗

Northern Europe ↗ ↗ *** ↘ *** ↗ ↘1 ↗ *** ↘ ↘ ↘ ↘ ↗ 8 2,3 ↗ ↗ ↗

North America

North Central
America ↗ ↗ ** ↘ ** ↗ ↗ ↗ ↗ 2 ↗ ↗ ↗
Western North 6,7
America ↗ ↗ ** ↘ ** 3 5 5 4,7 ↗ ** ↗ 6.7 ↘ 8 ↗6 ↘1 ↘ 1 1 ↗ ↗5 2 ↗ ↗ ↗
Central North
America ↗ ↗ ↗ ** 7 7 7 ↘ 8 4 ↘ ↘ ↗ ↗ 2 ↗ ↗ ↗
Eastern North
America ↗ ↗5 ↗ 7 ↘ 8 ↘1 ↘ 1 1 ↗ ↗ 2 ↗ ↗ ↗
North-Eastern
North America ↗ ↗ *** ↘ *** ↘ 5 5 6,7 6,7 8 ↘ 1,6 ↘ ↘ 1 4 ↗ 4,6 2,6 ↗ ↗ ↗
North-Western
North America ↗ ↗ *** ↘ *** ↘ 5 6 5 6,7 ↗ 6,7 8 ↘1 ↘ ↘ 1,6 ↗9 ↗ 2 ↗ ↗ ↗

Note: There are several region-specific qualifiers/exceptions attached to some of the directions of change/confidence levels indicated above. {12.4}
Key for observational trend evidence ↗ Past upward trend (medium or higher confidence) ↘ Past downward trend (medium or higher confidence)
Key for attribution evidence *** High confidence (or more) ** Medium confidence
High confidence Medium confidence Low confidence Medium confidence High confidence Not broadly
Key for level of confidence in future changes
of increase (or more) of increase (or more) in direction of change of decrease of decrease relevant
 Table TS.5 (continued)

Climatic Impact-driver

Heat and Cold Wet and Dry Wind Snow and Ice Coastal and Oceanic Other

Mean air temperature

Extreme heat

Cold spell

Frost

Mean precipitation

River flood

pluvial flood
Heavy precipitation and

Landslide

Aridity

Hydrological drought

Agricultural and ecological drought

Fire weather

Mean wind speed

Severe wind storm

Tropical cyclone

Sand and dust storm

Snow, glacier and ice sheet

Permafrost

Lake, river and sea ice

Heavy snowfall and ice storm

Hail

Snow avalanche

Relative sea level

Coastal flood

Coastal erosion

Marine heatwave

Ocean and lake acidity

Air pollution weather

Atmospheric CO2 at surface

Radiation at surface
Small Islands

Caribbean ↗ ↗ ** 5 ↗ 6 ↗ ↗ ↗

Pacific ↗ ↗ **1 2 3 4 5 ↗ 6 ↗ ↗ ↗

Polar Terrestrial Regions

Greenland and
Iceland ↗ ↗ ** ↘ ** ↗ ↗ ↘3 2,3 ↘1 ↘ ↘ ↗5 ↗ ↗
Arctic North
Europe ↗ ↗ ↗ ↘3 2,3 ↘1 ↘ ↘ ↗6 7 ↗ ↗

Russian Arctic ↗ ↗ ** ↘ ** ↗ ↘3 2,3 ↘ 1,4 ↘ 7 ↗ ↗

Arctic North-Western
North America ↗ ↗ ↘3 2,3 ↘ 1,4 ↘ ↘ 7 ↗ ↗
Arctic North-East
North America ↗ ↗ ↗ ↘3 2,3 ↘ 1,4 ↘ ↘ ↘ ↗ ↗

West Antarctica ↘ 1,4 ↗ ↗

East Antarctica ↗ ↗

Technical Summary
Note: There are several region-specific qualifiers/exceptions attached to some of the directions of change/confidence levels indicated above. {12.4}
Key for observational trend evidence ↗ Past upward trend (medium or higher confidence) ↘ Past downward trend (medium or higher confidence)
Key for attribution evidence *** High confidence (or more) ** Medium confidence

Key for level of confidence in future changes High confidence Medium confidence Low confidence Medium confidence High confidence Not broadly
127

of increase (or more) of increase (or more) in direction of change of decrease of decrease relevant
TS
 TS
Table TS.5 (continued)

Technical Summary
128

Climatic Impact-driver

Mean ocean temperature

Marine heatwave

Ocean acidity

Ocean salinity

Dissolved oxygen

Sea ice
Oceans

Arctic Ocean ↗ ↗ ↘ ***

South Pacific Ocean ↗ ↗ ↗

Equatorial Pacific Ocean ↗ ↗ ↗

North Pacific Ocean ↗ ↗ ↗

South Atlantic Ocean ↗ ↗ ↗

Equatorial Atlantic Ocean ↗ ↗ ↗

North Atlantic Ocean ↗ ↗ ↗

Equatorial Indian Ocean ↗ ↗ ↗

South Indian Ocean ↗ ↗ ↗

Arabian Sea ↗ ↗ ↗

Bay of Bengal ↗ ↗ ↗

Southern Ocean

Note: There are several region-specific qualifiers/exceptions attached to some of the directions of change/confidence levels indicated above. {12.4}
Key for observational trend evidence ↗ Past upward trend (medium or higher confidence) ↘ Past downward trend (medium or higher confidence)
Key for attribution evidence *** High confidence (or more) ** Medium confidence
High confidence Medium confidence Low confidence Medium confidence High confidence Not broadly
Key for level of confidence in future changes
of increase (or more) of increase (or more) in direction of change of decrease of decrease relevant
 Technical Summary

Notes:
Africa (projections)
1. Contrasted regional signal: drying in western portions and wetting in eastern portions
2. Likely increase over the Ethiopian Highlands
3. Medium confidence of decrease in frequency and increase in intensity
4. Along sandy coasts and in the absence of sufficient sediment supply from terrestrial or offshore sources
5. Substantial parts of the East Southern Africa and Madagascar coast are projected to prograde if present-day ambient shoreline change rates continue
Asia (projections)
1. Along sandy coasts and in the absence of additional sediment sinks/sources or any physical barriers to shoreline retreat.
2. Substantial parts of the coasts in these regions are projected to prograde if present-day ambient shoreline change rates continue
3. Tropical cyclones decrease in number but increase in intensity
4. High confidence of decrease in Indonesia (Atlas.5.4.5)
5. Medium confidence of decreasing in summer and increasing in winter
Australasia (projections)
1. High confidence of decrease in the south-west of the state of Western Australia
2. Medium confidence of decrease in north and east and increase in south and west TS
3. High confidence of increase in the south-west of the state of Western Australia
4. Medium confidence of increase in the north and east and decrease in south and west
5. Low confidence of increasing intensity, and high confidence of decreasing occurrence
6. High confidence of decrease in glacier volume, medium confidence of decrease in snow
7. Along sandy coasts and in the absence of additional sediment sinks/sources or any physical barriers to shoreline retreat
Central and South America (projections)
1. Increase in extreme flow in the Amazon basin
2. Tropical cyclones decrease in number but increase in intensity
3. Along sandy coasts and in the absence of additional sediment sinks/sources or any physical barriers to shoreline retreat.
4. Substantial parts of the North-Western South America, Northern South America and North-Eastern South America coasts are projected to prograde if present-day
ambient shoreline change rates continue
Europe (projections)
1. Excluding southern United Kingdom
2. Along sandy coasts and in the absence of additional sediment sinks/sources or any physical barriers to shoreline retreat
3. The Baltic Sea shoreline is projected prograde if present-day ambient shoreline change rates continue.
4. For the Alps, conditions conducive to landslides are expected to increase
5. Low confidence of decrease in the southernmost part of the region
6. General decrease except in Aegean Sea
7. Medium confidence of decrease in frequency and increase in intensities
8. Except in the Northern Baltic Sea region
North America (projections)
1. Snow may increase in some high elevations and during the cold season and decrease in other seasons and at lower elevations
2. Along sandy coasts and in the absence of additional sediment sinks/sources or any physical barriers to shoreline retreat.
3. Increasing in northern regions and decreasing toward the south
4. Decreasing in northern regions and increasing toward the south
5. Higher confidence in northern regions and lower toward the south
6. Higher confidence in southern regions and lower toward the north
7. Higher confidence in increase for some climatic impact-driver indices during summertime
8. Increase in convective conditions but decrease in winter extratropical cyclones
9. Relative sea level rise reduced given land uplift in Southern Alaska
Small Islands (projections)
1. Very high confidence in the direction of change, but low to medium confidence in the magnitude of change due to model uncertainty
2. Decrease in eastern Pacific and southern Pacific subtropics, but increase in parts of western and equatorial Pacific; with seasonal variation in future changes
3. High confidence in increase in extreme rain frequency and intensity in western tropical Pacific; low confidence in magnitude of change due to model bias
4. Increase in southern Pacific
5. Increase in intensity; decrease in frequency except over central North Pacific.
6. Along sandy coasts and in the absence of additional sediment sinks/sources or any physical barriers to shoreline retreat.
Polar Terrestrial Regions (projections)
1. Snow may increase in some high elevations and during the cold season and decrease in other seasons and at lower elevations
2. Higher confidence in southern regions and lower toward the north
3. Higher confidence in increase for some climatic impact-driver indices during summertime
4. Glaciers decline even as some regional snow climatic impact-driver indices increase
5. Decreasing in west and increasing in east
6. Except for Northern Baltic Sea coasts where relative sea levels fall
7. Along sandy coasts and in the absence of additional sediment sinks/sources or any physical barriers to shoreline retreat

129
 Technical Summary

While changes in climatic impact-drivers are projected everywhere, there is a

specific combination of changes each region would experience
(a) World regions grouped into five clusters, each one based on a combination of changes in climatic impact-drivers
Assessed future changes: Changes refer to a 20–30 year period centred around 2050 and/or consistent with 2°C global warming compared to a similar period
within 1960–2014 or 1850–1900.

GIC RAR 1) Hotter and drier

NWN NEN NEU
EEU WSB ESB RFE 2) Hotter and drier and in
WCE some regions wetter
WNA ENA ECA extremes
CNA MED WCA
TIB EAS
PAC NCA 3) Hotter and wetter
SAH ARP SAS extremes and in some
CAR
SCA regions more precipitation or
WAF NEAF
TS NSA SEA fire weather
CAF PAC
NWS SEAF
NES 4) Hotter and wetter and in
PAC SAM ESAF NAU some regions more flooding
MDG CAU
SES WSAF EAU
SWS
5) Hotter and in some
SAU NZ regions wetter extremes or
more precipitation
SSA
All coastal regions except North-East North America (NEN) and 6) Increase in Tropical
Greenland/Iceland (GIC) will be exposed to at least two among cyclones intensity or Severe
increases in relative sea level, coastal flood and coastal erosion winds

Combinations of future changes in climatic impact-drivers (CIDs)

LEGEND

Heat
increase in mean
temperature, extreme heat
CIDs changing in all the
regions of the cluster with
high confidence
Cold
decrease in cold spell, frost

CIDs changing in some

(1) (2) regions of the cluster with Drought
Hotter Hotter and more fire weather high and sometimes with increase in aridity, hydrological
and in some regions more and in some regions more pluvial medium confidence and
fire weather or drought or flooding or drought or less mean agricultural drought
both precipitation or snow and ice or
combinations of these

Snow and Ice

decrease in snow, glacier
and ice sheet

Mean Mean
precipitation precipitation
decrease increase

River Pluvial
flooding flooding
increase increase
(3) (4) (5)
Hotter, less snow and ice and Hotter and less snow/ice Hotter
more pluvial flooding and in some regions more and in some regions more
and in some regions more pluvial flooding or river pluvial flooding or mean Fire
precipitation or fire weather or flooding or mean precipita- precipitation or both weather
both tion or both increase

130
 Technical Summary

(b) Number of land & coastal regions (i) and open-ocean regions (ii) where each climatic impact-driver (CID) is projected to
increase or decrease with high confidence (dark shade) or medium confidence (light shade)

(i) (ii)
Heat & Cold Wet & Dry
Heavy precipitation and pluvial flood Wind Snow & Ice Other Coastal Open Ocean

Agricultural and ecological drought

& COASTAL REGIONS

Heavy snowfall and ice storm

NUMBER OF OPEN-OCEAN REGIONS

Atmospheric CO2 at surface
Snow, glacier and ice sheet
Mean surface temperature

Mean ocean temperature

TS

Lake, river and sea ice

Sand and dust storm
Hydrological drought

Air pollution weather

Radiation at surface
Severe wind storm
Mean precipitation

Mean wind speed

Relative sea level

Dissolved oxygen
Marine heatwave

Marine heatwave
Snow avalanche
Tropical cyclone

Coastal erosion

Ocean salinity
Ocean acidity

Ocean acidity
Extreme heat

Coastal flood
NUMBER OF LAND

Fire weather

Permafrost
River flood
Cold spell

Landslide
Aridity
Frost

Hail
55

45

35

25

15 15

5 5
5 5

15 15

25

35

45

55

BAR CHART LEGEND LIGHTER-SHADED ‘ENVELOPE’ LEGEND

Regions with high confidence increase The height of the lighter shaded ‘envelope’ behind each bar represents the
Regions with medium confidence increase maximum number of regions for which each CID is relevant. The envelope
Regions with high confidence decrease is symmetrical about the x-axis showing the maximum possible number of
Regions with medium confidence decrease relevant regions for CID increase (upper part) or decrease (lower part).

Climatic impact-drivers (CIDs) are physical climate system conditions (e.g., means, events, extremes) that affect an element of
society or ecosystems. Depending on system tolerance, CIDs and their changes can be detrimental, beneficial, neutral, or a
mixture of each across interacting system elements and regions. The CIDs are grouped into seven types, which are summarized
under the icons in sub-panels (i) and (ii). All regions are projected to experience changes in at least 5 CIDs. Almost all (96%) are
projected to experience changes in at least 10 CIDs and half in at least 15 CIDs. For many CID changes, there is wide
geographical variation, and so each region is projected to experience a specific set of CID changes. Each bar in the chart interactive-atlas.ipcc.ch
represents a specific geographical set of changes that can be explored in the WGI Interactive Atlas.

Figure TS.22 | Synthesis of the geographical distribution of climatic impact-drivers changes and the number of AR6 WGI reference regions where they
are projected to change.

131
 Technical Summary

Figure TS.22 (continued): Panel (a) shows the geographical location of regions belonging to one of five groups characterized by a specific combination of changing climatic
impact-drivers (CIDs). The five groups are represented by the five different colours, and the CID combinations associated with each group are represented in the corresponding
‘fingerprint’ and text below the map. Each fingerprint comprises a set of CIDs projected to change with high confidence in every region in the group and a second set of CIDs,
one or more of which are projected to change in each region with high or medium confidence. The CID combinations follow a progression from those becoming hotter and drier
(group 1) to those becoming hotter and wetter (group 5). In between (groups 2–4), the CIDs that change include some becoming drier and some wetter and always include
a set of CIDs which are getting hotter. Tropical cyclones and severe wind CID changes are represented on the map with black dots in the regions affected. Regions affected
by coastal CID changes are described by text on the map. The five groups are chosen to provide a reasonable level of detail for each region while not overwhelming the map
with a full summary of all aspects of the assessment, which is available in Table TS.5 and can be visualized in the Regional Synthesis component of the Interactive Atlas. The
CID changes summarized in the figure represent high and medium confidence changes for the mid-21st century, considering scenarios SSP2-4.5, RCP4.5, SRES A1B, or above
(SSP3-7.0, SSP5-8.5, RCP6.0, RCP8.5, SRES A2), which approximately encompasses global warming levels of 2.0°C to 2.4°C.
The bar chart in panel (b) shows the numbers of regions where each CID is increasing or decreasing with medium or high confidence for all land regions and ocean regions
listed in Table TS.5. The colours represent the direction of change and the level of confidence in the change: purple indicates an increase while brown indicates a decrease; darker
and lighter shades refer to high and medium confidence, respectively. Lighter background colours represent the maximum number of regions for which each CID is broadly
relevant. Sub-panel (i) shows the 30 CIDs relevant to the land and coastal regions while sub-panel (ii) shows the 5 CIDs relevant to the open ocean regions. Marine heatwaves
and ocean acidity are assessed for coastal ocean regions in panel (i) and for open ocean regions in panel (ii). Changes refer to a 20- to 30-year period centred around 2050
and/or consistent with 2°C global warming compared to a similar period within 1960–2014, except for hydrological drought and agricultural and ecological drought, which
is compared to 1850–1900. Definitions of the regions are provided in Atlas.1, the Interactive Atlas (https://interactive-atlas.ipcc.ch/) and Chapter 12. (Table TS.5, Figure TS.24)
{11.9, 12.2, 12.4, Atlas.1}
TS

TS.4.3.1 Common Regional Changes in Climatic Impact-Drivers regions (high confidence). In many tropical regions, the number of
days per year where a heat index of 41°C is exceeded would increase
Heat and cold: Changes in temperature-related CIDs such as mean by more than 100 days relative to the recent past under SSP5-8.5,
temperatures, growing season length, and extreme heat and frost have while this increase will be limited to less than 50 days under SSP1-2.6
already occurred (high confidence), and many of these changes have (high confidence) (Figure TS.6). The number of days per year where
been attributed to human activities (medium confidence). Over all temperature exceeds 35°C would increase by more than 150 days in
land regions with sufficient data (i.e., all except Antarctica), observed many tropical areas, such as the Amazon basin and South East Asia,
changes in temperature have already clearly emerged outside the by the end of century for the SSP5-8.5 scenario, while it is expected to
range of internal variability, relative to 1850–1900 (Figure TS.23). In increase by less than 60 days in these areas under SSP1-2.6 (except
tropical regions, recent past temperature distributions have already for the Amazon Basin) (high confidence) (Figure TS.24). {4.6.1, 11.3,
shifted to a range different to that of the early 20th century (high 11.9, 12.4, 12.5.2, Atlas}
confidence) (Section TS.1.2.4). Most land areas have very likely
warmed by at least 0.1°C per decade since 1960, and faster in recent Wet and dry: Compared to the global scale, precipitation internal
decades. On regional-to-continental scales, trends of increased variability is stronger at the regional scale while uncertainties in
frequency of hot extremes and decreased frequency of cold extremes observations, models and external forcing are all larger. However,
are generally consistent with the global-scale trends in mean GHG forcing has driven increased contrasts in precipitation amounts
temperature (high confidence). In a few regions, trends are difficult between wet and dry seasons and weather regimes over tropical land
to assess due to limited data availability. {2.3.1.1, 11.3, 11.9, 12.4, areas (medium confidence), with a detectable precipitation increase
Atlas.3.1} in the northern high latitudes (high confidence) (Box TS.6). The
frequency and intensity of heavy precipitation events have increased
Warming trends observed in recent decades are projected to continue over a majority of land regions with good observational coverage
over the 21st century and over most land regions at a rate higher (high confidence). A majority of land areas have experienced
than the global average (high confidence). For given global warming decreases in available water in dry seasons due to human-induced
levels, model projections from CMIP6 show future regional warming climate change associated with changes in evapotranspiration
changes that are similar to those projected by CMIP5. However, (medium confidence). Global hydrological models project a larger
projected regional warming in CMIP6 for given time periods and fraction of land areas to be affected by an increase rather than by
emissions scenarios has a wider range with a higher upper limit a decrease in river floods (medium confidence). Extreme precipitation
compared to CMIP5 because of the higher climate sensitivity in some and pluvial flooding will increase in many regions around the world
CMIP6 models and differences in the forcings. {Atlas.3–Atlas.11} on almost all continents (high confidence), but regional changes
in river floods are more uncertain than changes in pluvial floods
Under RCP8.5/SSP5-8.5, it is likely that most land areas will experience because complex hydrological processes, including land cover and
further warming of at least 4°C compared to a 1995–2014 baseline human water management, are involved. {8.2.2.1, 8.3.1, Box 8.2,
by the end of the 21st century, and in some areas significantly more. 10.4.1, 11.5, 11.6, 11.9, 12.4, 12.5.1, Atlas.3.1}
At increasing warming levels, extreme heat will exceed critical
thresholds for health, agriculture and other sectors more frequently Wind: Mean wind speed has decreased over most land areas with
(high confidence), and it is likely that cold spells will become less good observational coverage (medium confidence). It is likely that
frequent towards the end of the century. For example, by the end the global proportion of major tropical cyclone (TC) intensities
of the 21st century, dangerous humid heat thresholds, such as the (Categories 3–5) over the past four decades has increased. The
National Oceanic and Atmospheric Administration (NOAA) heat index proportion of intense TCs, average peak TC wind speeds, and peak
(HI) threshold of 41°C, will be exceeded much more frequently under wind speeds of the most intense TCs will increase on the global scale
the SSP5-8.5 scenario than under SSP1-2.6 and will affect many with increasing global warming (high confidence). {11.7.1}

132
 Technical Summary

Year of significant emergence of changes in temperature over land regions (S/N>2)

TS

Dataset: Berkeley Earth. Temperature changes relative to 1850–1900.

Before 1981 1981–1988 1989–1996 1997–2004 2005–2012 2013–2020

Year of significant emergence of changes in temperature over land regions (S/N>2)

Dataset: CRUTEM5. Temperature changes relative to 1850–1900. Grey: not enough data.

Before 1981 1981–1988 1989–1996 1997–2004 2005–2012 2013–2020

Figure TS.23 | Time period during which the signals of temperature change in observed data aggregated over the reference regions emerged from the
noise of annual variability in the respective aggregated data, using a signal-to-noise ratio of two as the threshold for emergence. The intent of this figure
is to show, for the AR6 WGI reference regions, when a signal of annual mean surface temperature change emerged from the noise of annual variability in two global datasets
and thus also provide some information on observational uncertainty. Emergence time is calculated for two global observational datasets: (a) Berkeley Earth and (b) CRUTEM5.
Regions in the CRUTEM5 map are shaded grey when data are available over less than 50% of the area of the region. (Section TS.1.2.4) {Figure Atlas.11}

133
 Technical Summary

CMIP5 CMIP6 CORDEX

(a) TX35 for 2041−2060 (RCP2.6) rel. to 1995−2014 (e) TX35 for 2041−2060 (SSP1-2.6) rel. to 1995−2014 (i) TX35 for 2041−2060 (SSP1-2.6) rel. to 1995−2014

(b) TX35 for 2081−2100 (RCP2.6) rel. to 1995−2014 (f) TX35 for 2081−2100 (SSP1-2.6) rel. to 1995−2014 (j) TX35 for 2081−2100 (SSP1-2.6) rel. to 1995−2014

TS
(c) TX35 for 2041−2060 (RCP8.5) rel. to 1995−2014 (g) TX35 for 2041−2060 (SSP5-8.5) rel. to 1995−2014 (k) TX35 for 2041−2060 (SSP5-8.5) rel. to 1995−2014

(d) TX35 for 2081−2100 (RCP8.5) rel. to 1995−2014 (h) TX35 for 2081−2100 (SSP5-8.5) rel. to 1995−2014 (l) TX35 for 2081−2100 (SSP5-8.5) rel. to 1995−2014

Change (days) Colour High model agreement No data

Low model agreement
-25 0 50 100 150 200 250

Figure TS.24 | Projected change in the mean number of days per year with maximum temperature exceeding 35°C for Coupled Model Intercomparison
Project Phase 5 (CMIP5; first column), Phase 6 (CMIP6; second column) and Coordinated Regional Climate Downscaling Experiment (CORDEX; third
column) ensembles. The intent of this figure is to show that there is a consistent message about the patterns of projected change in extreme daily temperatures from the
CMIP5, CMIP6 and CORDEX ensembles. The map shows the median change in the number of days per year between the mid-century (2041–2060) or end-century (2081–2100)
and historical (1995–2014) periods for the CMIP5 and CORDEX RCP8.5 and RCP2.6 scenario ensembles and the CMIP6 SSP5-8.5 and SSP1-2.6 scenario ensembles. Hatching
indicates areas where less than 80% of the models agree on the sign of change. {Interactive Atlas}

Snow and ice: Many aspects of the cryosphere either have seen Coastal and oceanic: There is high confidence that SST will
significant changes in the recent past or will see them during the increase in all oceanic regions except the North Atlantic. Regional
21st century (high confidence). Glaciers will continue to shrink sea level change has been the main driver of changes in extreme
and permafrost to thaw in all regions where they are present (high sea levels across the quasi-global tide gauge network over the
confidence). Also, it is virtually certain that snow cover will experience 20th century (high confidence). With the exception of a few regions
a decline over most land regions during the 21st century, in terms with substantial land uplift, relative sea level rise is very likely to
of water equivalent, extent and annual duration. There is high virtually certain (depending on the region) to continue during the
confidence that the global warming-induced earlier onset of spring 21st century, contributing to increased coastal flooding in low-lying
snowmelt and increased melting of glaciers have already contributed areas (high confidence) and coastal erosion along most sandy
to seasonal changes in streamflow in high-latitude and low-elevation coasts (high confidence) over the 21st century. In the open ocean,
mountain catchments. Nevertheless, it is very likely that some high- acidification, changes in sea ice, and deoxygenation have already
latitude regions will experience an increase in winter snow water emerged in many areas (high confidence). Marine heatwaves are
equivalent due to the effect of increased snowfall prevailing over also expected to increase around the globe over the 21st century
warming-induced increased snowmelt. (Section TS.2.5) {8.2.2.1, (high confidence). (Section TS.2.4) {Box 9.2, 9.2.1.1, 9.6, 9.6.4,
8.3.1, Box 8.2, 9.4, 9.5.1, 9.5.2, 12.4, Atlas.4–Atlas.9, Atlas.11} 9.6.4.2, 12.4}

134
 Technical Summary

Other variables and concurrent CID changes: It is virtually certain In addition to the main changes summarized above and in
that atmospheric CO2 and oceanic pH will increase in all climate Section TS.4.3.1, additional details per CID are given below.
scenarios, until net zero CO2 emissions are achieved (Section TS.2.2).
In nearly all regions, there is low confidence in changes in hail, ice Heat and cold: Observed and projected increases in mean
storms, severe storms, dust storms, heavy snowfall, and avalanches, temperature and a shift toward heat extreme characteristics are
although this does not indicate that these CIDs will not be affected broadly similar to the generic pattern described in Section TS.4.3.1.
by climate change. For such CIDs, observations are often short- {2.3.1.1.2, 11.3, 11.9, 12.4.1.1, Atlas.4.2, Atlas.4.4}
term or lack homogeneity, and models often do not have sufficient
resolution or accurate parametrizations to adequately simulate them Wet and dry: Mean precipitation changes have been observed
over climate change time scales. The probability of compound events over Africa, but the historical trends are not spatially coherent (high
has increased in the past due to human-induced climate change and confidence). North Eastern Africa, East Southern Africa and Central
will likely continue to increase with further global warming, including Africa have experienced a decline in rainfall since about 1980 and
for concurrent heatwaves and droughts, compound flooding, and the parts of West Africa an increase (high confidence). Increases in the
possibility of connected sectors experiencing multiple regional extreme frequency and/or the intensity of heavy rainfall have been observed in
events at the same time (for example, in multiple breadbaskets) (high East and West Southern Africa, and the eastern Mediterranean region TS
confidence). {5.3.4.2, 11.8, Box 11.3, Box 11.4, 12.4} (medium confidence). Increasing trends in river flood occurrence can
be identified beyond 1980 in East and West Southern Africa (medium
TS.4.3.2 Region-by-Region Changes in Climatic Impact-Drivers confidence) and Western Africa (high confidence). However, Northern
Africa and West Southern Africa are likely to have a reduction in
This section provides a continental synthesis of changes in CIDs, precipitation. Over West Africa, rainfall is projected to decrease
some examples of which are presented in Figure TS.25. in the western Sahel subregion and increase along the Guinea Coast
subregion (medium confidence). Rainfall is projected to increase over
With 2°C global warming, and as early as the mid-21st Eastern Africa (medium confidence). {8.3.1.6, 11.4, 11.9, 12.4.1.2,
century, a wide range of CIDs, particularly related to the Atlas.4.2, Atlas.4.4, Atlas.4.5}
water cycle and storms, are expected to show simultaneous
region-specific changes relative to the recent past with high Precipitation declines and aridity trends in Western Africa, Central
or medium confidence. In a number of regions (Southern Africa, Southern Africa and the Mediterranean co-occur with
Africa, the Mediterranean, North Central America, Western trends towards increased agricultural and ecological droughts in
North America, the Amazon regions, South-Western South the same regions (medium confidence). Trends towards increased
America, and Australia), increases in one or more of drought, hydrological droughts have been observed in the Mediterranean
aridity and fire weather (high confidence) will affect a wide (high confidence) and Western Africa (medium confidence). These
range of sectors, including agriculture, forestry, health and trends correspond with projected regional increases in aridity and
ecosystems. In another group of regions (North-Western, fire weather conditions (high confidence). {8.3.1.6, 8.4.1.6, 11.6,
Central and Eastern North America, Arctic regions, North- 11.9, 12.4.1.2}
Western South America, Northern, Western and Central
and Eastern Europe, Siberia, Central, South and East Asia, Wind: Mean wind, extreme winds and the wind energy potential in
Southern Australia and New Zealand), decreases in snow North Africa and the Mediterranean are projected to decrease across
and/or ice or increases in pluvial/river flooding (high all scenarios (high confidence). Over Western Africa and Southern
confidence) will affect sectors such as winter tourism, Africa, a future significant increase in wind speed and wind energy
energy production, river transportation and infrastructure. potential is projected (medium confidence). There is a projected
{11.9, 12.3, 12.4, 12.5, Table 12.2} decrease in the frequency of tropical cyclones making landfall
over Madagascar, East Southern Africa and East Africa (medium
TS.4.3.2.1 Africa confidence). {12.4.1.3}

Additional regional changes in Africa, besides those Snow and ice: There is high confidence that African glaciers and
described in Section TS.4.3.1, include a projected decrease snow have very significantly decreased in the last decades and that
in total precipitation in the northernmost and southernmost this trend will continue in the 21st century. {12.4.1.4}
regions (high confidence), with Western Africa having a west-
to-east pattern of decreasing-to-increasing precipitation Coastal and oceanic: Relative sea level has increased at a higher
(medium confidence). Increases in heavy precipitation that rate than GMSL around Africa over the last 3 decades. The present
can lead to pluvial floods (high confidence) are projected day 1-in-100-year extreme total water level (ETWL) is between
for most African regions, even as increasing dry CIDs 0.1 m and 1.2 m around Africa, with values around 1 m or above
(aridity; hydrological, agricultural and ecological droughts; along the East and West Southern and Central Eastern Africa coasts.
fire weather) are projected in the western part of Western Satellite-derived shoreline retreat rates up to 1 m yr –1 have been
Africa, Southern Africa and Northern Africa and the observed around the continent from 1984 to 2015, except in South
Mediterranean regions (medium to high confidence). {8.4, Eastern Africa, which has experienced a shoreline progradation
11.3, 11.6, 11.9, 12.4, Atlas.4} (growth) rate of 0.1 m yr –1 over the same period. {12.4.1.5}

135
 Technical Summary

(a) Change in number of days with SWE>100 mm

NWN NEN WNA CNA
100
Days with SWE>100 mm (days yr-1) 80
60
40
20
0
ENA
100
80
60
TS 40 NWN NEN
20
0
1.5 2 4 mid. long.
r.past.

GWL Time slices WNA CNA ENA

NOAA Heat Index exceeding 41°C
NCA SCA CAR NCA
300 CAR
SCA
days yr-1

200
100 NSA

NWS
0 NES
SAM
r.past

r.past

r.past
long.

long.

long.
mid.

mid.

mid.
1.5

1.5

1.5
2
4

2
4

2
4

GWL Time slices GWL Time slices GWL Time slices

100-yr return period stream flow SES

2.5
NWS NSA NES SWS
0.8
2 0.6 SSA
Recent past value (m3 s-1 km-2)

1.5 0.4
0.2
Change from recent past (m3 s-1 km-2)
1
0.5 0
0.2
0
SWS SAM SES SSA
2.5
0.8
2 0.6
1.5 0.4
1 0.2
0.5 0
0.2
0
1.5 2 4 mid. long. 1.5 2 4 mid. long. 1.5 2 4 mid. long. 1.5 2 4 mid. long.
r.past.

r.past.

r.past.

r.past.

GWL Time slices GWL Time slices GWL Time slices GWL Time slices

Legend GWL RCPs/SSPs 90 th p

+1.5 ° C +2 ° C +4 ° C r.past 2.6 8.5
CMIP6 median
CMIP5
CORDEX 10 th p

136
 Technical Summary

(b) 100-yr return period stream flow Maximum temperature exceeding 35°C
NEU WCE MED EEU WSB RAR
0.8 0.1
300 300

Change from past

Recent past value

days yr -1
0.6

(m3 s-1 km-2)

200 200
(m3 s-1 km-2)

0
0.4 100 100
0.2 0.1 0 0
ESB ECA TIB
0 1.5 2 4 mid. long. 1.5 2 4 mid. long. 1.5 2 4 mid. long. 300 300

days yr -1
r.past.

r.past.

r.past.
GWL Time slices GWL Time slices GWL Time slices 200
200
100 100
0 0
RAR

r.past

r.past

r.past
long.

long.

long.
mid.

mid.

mid.
1.5

1.5

1.5
2
4

2
4

2
4
NEU
EEU WSB ESB RFE GWL Time slices GWL Time slices GWL Time slices
WCE TS
ECA Shoreline position change
MED WCA
TIB EAS EAS RFE

position change (m)

Average shoreline
200
SAH ARP SAS
0
WAF NEAF –200
CAF SEA
–400
SEAF –600
NAU mid- long- mid- long-
term term term term
WSAF
ESAF

MDG CAU EAU

NOAA Heat Index exceeding 41°C
SAU NZ ARP WCA
1.5
SAH WAF 300
days yr -1

0.4 200
Change from recent past (m3 s-1 km-2)

1 0.2 100
0 0
0.5 SEA
SAS
–0.2
0 300
Recent past value (m3 s-1 km-2)

days yr -1

1.5
CAF NEAF SEAF 200
0.4 100
1 0.2 0
r.past

r.past
long.

long.
mid.

mid.
0
1.5

1.5

0.5
2
4

2
4

–0.2 GWL Time slices GWL Time slices

0
1.5
WSAF ESAF 2.5
MDG Shoreline position change
0.4 0.8
NAU CAU EAU SAU NZ
position change (m)

2
Average shoreline

1 0.6
0.2 1.5 200
0.4
0 0
0.5 1 0.2
0 –200
–0.2 0.5
0 –0.2 –400
1.5 2 4 mid. long. 1.5 2 4 mid. long. 0
1.5 2 4 mid. long. –600
r.past.
r.past.

GWL Time slices GWL Time slices mid- long- mid- long- mid- long- mid- long- mid- long-
r.past.

GWL Time slices term term term term term term term term term term

Legend GWL RCPs/SSPs 90 th p 95 p

th
+1.5 °C +2 °C +4 °C r.past 2.6 8.5 CMIP5 RCP8.5
CMIP6 median median
CMIP5 th th CMIP5 RCP4.5
CORDEX 10 p 5 p

Figure TS.25 | Distribution of projected changes in selected climatic impact-driver (CID) indices for selected regions for Coupled Model Intercomparison
Project Phases 5 and 6 (CMIP6, CMIP5) and Coordinated Regional Downscaling Experiment (CORDEX) model ensembles. The intent of this figure is to show
that many CID projections for multiple global warming levels and scenarios time slices are available for all the AR6 WGI reference regions and are based on both global (CMIP5,
CMIP6) and regional (CORDEX) model ensembles. Different indices are shown for different region: for Eastern Europe and North Asia, the mean number of days per year with
maximum temperature exceeding 35°C; for Central America, the Caribbean, South West Asia, South Asia and South East Asia, the mean number of days per year with the
National Oceanic and Atmospheric Administration (NOAA) Heat Index exceeding 41°C; for Australasia, East Asia and Russian Far East, the average shoreline position change;
for South America, Europe and Africa, the mean change in 1-in-100-year river discharge per unit catchment area (m3 s–1 km–2); and for North America, the median change in
the number of days with snow water equivalent (SWE) over 100 mm. For each box plot, the changes or the climatological values are reported with respect to, or compared to,
the recent past (1995–2014) period for 1.5°C, 2°C and 4°C global warming levels and for mid-century (2041–2060) or end-century (2081–2100) periods for the CMIP5 and
CORDEX RCP8.5 and RCP2.6 and CMIP6 SSP5-8.5 and SSP1-2.6 scenarios ensembles. {Figures 12.5, 12.6, 12.9, 12.SM.1, 12.SM.2, and 12.SM.6}

137
 Technical Summary

TS.4.3.2.2 Asia {2.3, 8.3, 8.4, 9.5, 9.6, 10.6, Box 10.4, 11.4, 11.5, 11.7, 11.9,
12.4.2, Atlas.3.1, Atlas.5, Atlas.5.2, Atlas.5.3, Atlas.5.4,
Due to the high climatological and geographical heterogeneity of Atlas.5.5}
Asia, some assessment findings below are summarized over five
sub-continental areas comprising one or more of the AR6 WGI In addition to the main changes summarized above and in Section
reference regions (Box TS.12): East Asia (EAS+ECA), North Asia TS.4.3.1, further details are given below.
(WSB+ESB+RFE), South Asia (SAS), South East Asia (SEA) and South
West Asia (ARP+WCA). Heat and cold: Over all regions of Asia, observed and projected
increases in mean temperature and a shift toward heat extreme
Additional regional changes in Asia, besides those features characteristics are broadly similar to the generic pattern described
described in Section TS.4.3.1, include historical trends in Section TS.4.3.1. Over South East Asia, annual mean surface
of annual precipitation that show considerable regional temperature will likely increase by a slightly smaller amount than the
differences (high confidence). East Asian Monsoon global average. {Atlas.5.4.4}
precipitation has changed, with drying in the north and
TS wetting in the south since the 1950s, and annual mean Wet and dry: Over East Asia, historical trends of annual precipitation
precipitation totals very likely have increased over show considerable regional differences but with increases over north-
most territories of North Asia since the mid-1970s (high west China and South Korea (high confidence). Daily precipitation
confidence). South Asian summer monsoon precipitation extremes have increased over part of the region (high confidence).
decreased over several areas since the mid-20th century Extreme hydrological drought frequency has increased in a region
(high confidence) but is likely to increase during the 21st extending from south-west to north-east China, with projected
century, with enhanced interannual variability. (Box TS.13) increases of agricultural and ecological drought for 4°C GWL and
fire weather for 2°C and above (medium confidence). {8.3.2, 8.4.2,
Increases in precipitation and river floods are projected 11.4.4, 11.4.5, 11.9, 12.4.2.2, Atlas.5.1.2}
over much of Asia: in the annual mean precipitation in
East, North, South and South East Asia (high confidence); Over North Asia, annual mean precipitation totals have very likely
for extremes in East, South, West Central, North and South increased, causing more intense flooding events, and there is
East Asia (high confidence) and Arabian Peninsula (medium medium confidence that the number of dry days has decreased.
confidence); and for river floods in East, South and South Concurrently, total soil moisture is projected to decline extensively
East Asia and East Siberia (medium confidence). Aridity (medium confidence). {8.3.1.3, 8.4.1.6, 11.4.5, 11.5.2, 11.5.5,
in East and West Central Asia is projected to increase, 12.4.2.2, Atlas.5.2.2}
especially beyond the middle of the 21st century and global
warming levels beyond 2°C (medium confidence). Fire Over South Asia, the summer monsoon precipitation decreased
weather seasons are projected to lengthen and intensify over several areas since the mid-20th century (high confidence),
everywhere except South East Asia, Tibetan Plateau and while it increased in parts of the western HKH and decreased over
Arabian Peninsula (medium confidence). eastern-central HKH (medium confidence). The frequency of heavy
precipitation and flood events has increased over several areas
Surface wind speeds have been decreasing in Asia (high during the last few decades (medium confidence). {8.3.1.3, 8.3.2.4.1,
confidence), but there is a large uncertainty in future 8.4.1.5, 8.4.2.4.1, 10.6.3.3, 10.6.3.5, 10.6.3.6, 10.6.3.8, Cross-
trends, with medium confidence that mean wind speeds Chapter Box 10.4, 11.4.1, 11.4.2, 11.4.5, 11.5.5, 12.4.2.2, Box 10.4,
will decrease in North Asia, East Asia and Tibetan Plateau Atlas 5.3.2}
and that tropical cyclones will have decreasing frequency
and increasing intensity overall in South East and East Over South East Asia, mean precipitation trends are not spatially
Asia. coherent or consistent across datasets and seasons (high confidence).
Most of the region has experienced an increase in rainfall intensity
Over North Asia, increases in permafrost temperature and but with a reduced number of wet days (medium confidence). Rainfall
its thawing have been observed over recent decades (high is projected to increase in the northern parts of South East Asia and
confidence). Future projections indicate continuing decline decrease in areas in the Maritime Continent (medium confidence).
in seasonal snow duration, glacial mass, and permafrost area {8.4.1, 11.4.2, 11.5.5, 11.9, 12.4.2.2, Atlas.3.1, Atlas.5.4.2, Atlas.5.4.4}
by mid-century (high confidence). Snow-covered areas and
snow volumes will decrease in most regions of the Hindu Over South West Asia, an observed annual precipitation decline
Kush Himalaya (HKH) during the 21st century, and snowline over the Arabian Peninsula since the 1980s of 6.3 mm per decade
elevations will rise (high confidence) and glacier volumes is contrasted with observed increases between 1.3 mm and 4.8 mm
are likely to decline with greater mass loss in higher CO2 per decade during 1960–2013 over the elevated part of eastern
emissions scenarios. Heavy snowfall is increasing in East West Central Asia (very high confidence), along with an increase of
Asia and North Asia (medium confidence) but with limited the frequency and intensity of extreme precipitation. {Figure 8.19,
evidence on future changes in hail and snow avalanches. Figure 8.20, 8.3.1.6, 8.4.1.6, 11.9, Table 11.2A, 12.4.2.2, Atlas.5.5}

138
 Technical Summary

Wind: Over East Asia, the terrestrial near-surface wind speed has TS.4.3.2.3 Australasia
decreased and is projected to decrease further in the future (medium
confidence). Since the mid 1980’s, there has been an increase in the Additional regional changes in Australasia, besides those
number and intensification rate of intense TCs (medium confidence), features described in Section TS.4.3.1, include a significant
with a significant north-westward shift in tracks and a northward decrease in April to October rainfall in the south-west of
shift in their average latitude, increasing exposure over East China, the state of Western Australia, observed from 1910 to 2019
the Korean Peninsula and the Japanese Archipelago (medium and attributable to human influence (high confidence),
confidence). {11.7.1, 12.4.2.3} which is very likely to continue in future. Agricultural
and ecological droughts and hydrological droughts have
Over North Asia, there is medium confidence for a decreasing trend increased over Southern Australia (medium confidence),
in wind speed during 1979–2018 and for projected continuing and meteorological droughts have decreased over Northern
decreases of terrestrial near-surface wind speed. {2.3.1.4.4, 12.4.2.3} and Central Australia (medium confidence). Relative sea
level has increased over the period 1993–2018 at a rate
Over South East Asia, although there is no significant long-term trend higher than GMSL around Australasia (high confidence).
in the number of TCs, fewer but more extreme TCs have affected the Sandy shorelines have retreated around the region, except TS
Philippines during 1951–2013. {11.7.4, 12.4.2.3} in Southern Australia, where a shoreline progradation rate
of 0.1 m yr –1 has been observed.
Snow and ice: Over East Asia, decreases have been observed in the
frequency, and increases in the mean intensity, of snowfall in north- In the future, heavy precipitation and pluvial flooding are
western, north-eastern and south-eastern China and the eastern very likely to increase over Northern Australia and Central
Tibetan Plateau since the 1960s. Heavy snowfall is projected to Australia, and they are likely to increase elsewhere in
occur more frequently in some parts of Japan (medium confidence). Australasia for global warming levels (GWLs) exceeding 2°C
{12.4.2.4, Atlas.5.1.2} and with medium confidence for a 2°C GWL. Agricultural and
ecological droughts are projected to increase in Southern
Over North Asia, seasonal snow duration and extent have decreased and Eastern Australia (medium confidence) for a 2°C GWL.
in recent decades (high confidence), and maximum snow depth likely Fire weather is projected to increase throughout Australia
has increased since the mid-1970s, particularly over the south of the (high confidence) and New Zealand (medium confidence).
Russian Far East. {2.3.2.5, 8.3.1.7.2, 9.5, 12.4.2.4, Atlas.5.2, Atlas.5.4} Snowfall is expected to decrease throughout the region
at high altitudes in both Australia (high confidence) and
Over South Asia, snow cover has reduced over most of the HKH since New Zealand (medium confidence), with glaciers receding
the early 21st century, and glaciers have thinned, retreated, and lost in New Zealand (high confidence). {11.4, Table 11.6, 12.3,
mass since the 1970s (high confidence), although the Karakoram 12.4.3, Atlas.6.4, Atlas.6.5}
glaciers have either slightly gained mass or are in an approximately
balanced state (medium confidence). {8.3.1.7.1, Cross-Chapter In addition to the main changes summarized above and in Section
Box 10.4} TS.4.3.1, further details are given below.

Over South West Asia, mountain permafrost degradation at high Heat and cold: Observed and projected increases in mean
altitudes has increased the instability of mountain slopes in the past temperature and a shift toward heat extreme characteristics are
decade (medium confidence). More than 60% of glacier mass in the broadly similar to the generic pattern described in Section TS.4.3.1.
Caucasus is projected to disappear under RCP8.5 emissions by the {11.9, 12.4.3.1, Atlas.6}
end of the 21st century (medium confidence). {9.5.1, 9.5.3, 12.4.2.4}
Wet and dry: There is medium confidence that heavy precipitation has
Coastal and oceanic: Over the last three decades, relative sea increased in Northern Australia since 1950. Annual mean precipitation
level has increased at a rate higher than GMSL around Asia (high is projected to increase in the south and west of New Zealand (medium
confidence). Gross coastal area loss and shoreline retreat has been confidence) and is projected to decrease in south-west Southern
observed over 1984–2015, but with localized shoreline progradation Australia (high confidence), Eastern Australia (medium confidence), and
in the Russian Far East, East and South East Asia. {12.4.2.5} in the north and east of New Zealand (medium confidence) for a GWL
of 2°C. There is medium confidence that river flooding will increase
Projections show that regional mean sea level continues to rise (high in New Zealand and Australia, with higher increases in Northern
confidence), ranging from 0.4–0.5 m under SSP1-2.6 to 0.8–1.0 m Australia. Aridity is projected to increase with medium confidence in
under SSP5-8.5 for 2081–2100 relative to 1995–2014 (median Southern Australia (high confidence in south-west Southern Australia),
values). This will contribute to more frequent coastal flooding and Eastern Australia (medium confidence) and in the north and east of
higher ETWL in low-lying areas and coastal erosion along sandy New Zealand (medium confidence) for GWLs around 2°C. {11.4, 11.9,
beaches (high confidence). There is high confidence that compound Table 11.6, 12.4.3.2, Atlas.6.2}
effects of climate change, land subsidence, and human factors will
lead to higher flood levels and prolonged inundation in the Mekong Wind: Mean wind speeds are projected to increase in parts of north-
Delta and other Asian coasts. {9.6.1, 9.6.3, 12.4.2.5} eastern Australia (medium confidence) by the end of the 21st century

139
 Technical Summary

under high CO2 emissions scenarios. TCs in north-eastern and north South America and Southern Central America (medium confidence).
Australia are projected to decrease in number (high confidence) but In Northern South America and Southern Central America, aridity
increase in intensity except for ‘east coast lows’ (low confidence). and agricultural and ecological droughts are increasing with
{12.4.3.3} medium confidence. Fire weather is projected to increase over
Southern Central America and Southern South America with medium
Snow and ice: Observations in Australia show that the snow season confidence. {8.3.1.3, 8.4.2.4.5, 11.4.2, 11.9, Table 11.14, Table 11.15,
length has decreased by 5% in the last five decades. Furthermore, the 12.4.4.2, Atlas.7.2.2, Atlas.7.2.4}
date of peak snowfall in Australia has advanced by 11 days over the
last 5 decades. Glacier ice volume in New Zealand has decreased by Wind: Climate projections indicate an increase in mean wind speed
33% from 1977 to 2018. {12.4.3.4, Atlas.6.2} and in wind power potential over the Amazonian region (Northern
South America, South American Monsoon, North-Eastern South
Coastal and oceanic: Observed changes in marine heatwaves America) (medium confidence). {12.4.4.3}
(MHWs) over the 20th century in the region show an increase in their
occurrence frequency, except along the south-east coast of New Snow and ice: Glacier volume loss and permafrost thawing will
TS Zealand, an increase in duration per event, and the total number of likely continue in the Andes Cordillera under all climate scenarios,
MHW days per decade, with the change being stronger in the Tasman causing important reductions in river flow and potentially high-
Sea than elsewhere. The present day 1-in-100-year ETWL is between magnitude glacial lake outburst floods. {9.5.1.1, 12.4.4.4}
0.5–2.5 m around most of Australia, except the north-western
coast where 1-in-100-year ETWL can be as high as 6–7 m. {Box 9.1, Coastal and oceanic: Around Central and South America, relative
12.3.1.5, 12.4.3.5} sea level has increased at a higher rate than GMSL in the South
Atlantic and the subtropical North Atlantic, and at a rate lower than
TS.4.3.2.4 Central and South America GMSL in the East Pacific over the last 3 decades. The present day
1-in-100-year ETWL is highest in Southern and South-Western South
Additional regional changes in Central and South America, America subregions, where it can be as large as 5 to 6 m. Satellite
besides those features described in Section TS.4.3.1, include observations for 1984–2015 show shoreline retreat rates along the
increases in mean and extreme precipitation in South- sandy coasts of Southern Central America, South-Eastern South
Eastern South America since the 1960s (high confidence) America and Southern South America, while shoreline progradation
(Section TS.4.2.3). Decreasing trends in mean precipitation rates have been observed in North-Western South America and
and increasing trends in agricultural and ecological drought Northern South America. Over the period 1982–2016, the coastlines
are observed over North-Eastern South America (medium experienced at least one MHW per year, and more along the Pacific
confidence). The intensity and frequency of extreme coast of North Central America and the Atlantic coast of South-
precipitation and pluvial floods is projected to increase Eastern South America. {12.4.4.5}
over South-Eastern South America, Southern South America,
Northern South America, South American Monsoon and TS.4.3.2.5 Europe
North-Eastern South America (medium confidence) for a 2°C
GWL and above. Increases of agricultural and ecological Additional regional changes in Europe, besides those features
drought are projected in South America Monsoon and described in Section TS.4.3.1, include observed increases
Southern South America, and fire weather is projected to in pluvial flooding in Northern Europe and hydrological
increase over several regions (Northern South America, the and agricultural/ecological droughts in the Mediterranean
South American Monsoon, North-Eastern South America (high confidence), which have been attributed to human
and South-Western South America) (high confidence). {8.3, influence with high and medium confidence, respectively.
8.4, 11.3, 11.4, 11.9, Table 11.13, Table 11.14, Table 11.15, Increased mean precipitation amounts at high latitudes
12.4.4.2, Atlas.7.1, Atlas.7.2} in boreal winter and reduced summer precipitation in
southern Europe are projected starting from a 2°C GWL
In addition to the main changes summarized above and in Section (high confidence). Aridity, agricultural and hydrological
TS.4.3.1, further details are given below. droughts and fire weather conditions will increase in
the Mediterranean region starting from 2°C GWL (high
Heat and cold: Observed and projected increases in mean confidence). Pluvial flooding will increase everywhere
temperature and a shift toward heat extreme characteristics are with high confidence except for medium confidence in the
broadly similar to the generic pattern described in Section TS.4.3.1. Mediterranean; in Western and Central Europe this also
{11.3.2, 11.3.5, Table 11.13, 12.4.4.1, Atlas.7.1.2, Atlas.7.2.2, applies to river flooding starting from a 2°C GWL (high
Atlas.7.2.4} confidence). Most periglacial processes in Northern Europe
are projected to disappear by the end of the 21st century,
Wet and dry: Mean precipitation is projected to change in a dipole even for a low warming scenario (medium confidence). {8.3,
pattern with increases in North-Western and South-Eastern South 11.3, 11.9, 12.4.5, 12.5.2, Atlas.8.2, Atlas.8.4}
America and decreases in North-Eastern and South-Western South
America (high confidence) and with further decreases in Northern

140
 Technical Summary

In addition to the main changes summarized above and in Section America and Northern Central America (from medium to
TS.4.3.1, further details are given below. high confidence). Severe wind storms, tropical cyclones
and dust storms in North America are shifting toward more
Heat and cold: Observed and projected increases in mean extreme characteristics (medium confidence), and both
temperature and a shift toward heat extreme characteristics are observations and projections point to strong changes in the
broadly similar to the generic pattern described in Section TS.4.3.1. seasonal and geographic range of snow and ice conditions
{11.3, 11.9, 12.4.5.1, 12.5.2, Atlas.8.2, Atlas.8.4} in the coming decades (very high confidence). General
findings for relative sea level, coastal flooding and erosion
Wet and dry: There is medium confidence that annual mean will not apply for areas with substantial land uplift around
precipitation has increased in Northern Europe, West and Central the Hudson Bay and Southern Alaska. {8.4, 11.4, 11.5, 11.7,
Europe, and Eastern Europe since the early 20th century and 11.9, 12.4, Atlas.9.4}
high confidence for increases in extreme precipitation. In the
European Mediterranean, the magnitude and sign of observed land In addition to the main changes summarized above and in Section
precipitation trends depend on time period and exact study region TS.4.3.1, further details are given below.
(medium confidence). There is medium confidence that river floods TS
will decrease in Northern, Eastern and southern Europe for high Heat and cold: Observed and projected increases in mean
warming levels. {8.3.1.3, 11.3, 11.9, 12.4.5.2, Atlas.8.2, Atlas.8.4} temperature and a shift toward heat extreme characteristics are
broadly similar to the generic pattern described in Section TS.4.3.1.
Wind: Mean wind speed over land has decreased (medium {11.3, 11.9, 12.4.6.1, Atlas.9.2, Atlas.9.4}
confidence), but the role of human-induced climate change has not
been established. There is high confidence that mean wind speeds Wet and dry: Annual precipitation increased over parts of Eastern
will decrease in Mediterranean areas and medium confidence for and Central North America during 1960–2015 (high confidence) and
such decreases in Northern Europe for GWLs exceeding 2°C. The has decreased in parts of south-western United States and north-
frequency of Medicanes (tropical-like cyclones in the Mediterranean) western Mexico (medium confidence). River floods are projected to
is projected to decrease (medium confidence). {11.9, 12.4.5.3} increase for all North American regions other than Northern Central
America (medium confidence). {8.4.2.4, 11.4, 11.5, 11.9, 12.4.6.2,
Snow and ice: In the Alps, snow cover will decrease below elevations Atlas.9.2, Atlas.9.4}
of 1500–2000 m throughout the 21st century (high confidence).
A reduction of glacier ice volume is projected in the European Alps Agricultural and ecological drought increases have been observed in
and Scandinavia with high confidence and with medium confidence Western North America (medium confidence), and aridity is projected
for the timing and mass change rates. {9.5.2, 12.4.5.4} to increase in the south-western United States and Northern Central
America, with lower summer soil moisture across much of the
Coastal and oceanic: Over the last three decades, relative sea level continental interior (medium confidence). {8.4.1, 11.6.2, 12.4.6.2}
has increased at a lower rate than GMSL in the sub-polar North
Atlantic coasts of Europe. The present-day 1-in-100-year ETWL is Wind: Projections indicate a greater number of the most intense
between 0.5–1.5 m in the Mediterranean basin and 2.5–5.0 m in TCs, with slower translation speeds and higher rainfall potential for
the western Atlantic European coasts, around the United Kingdom Mexico’s Pacific Coast, the Gulf Coast and the United States East
and along the North Sea coast, and lower at 1.5–2.5 m along the Coast (medium confidence). Mean wind speed and wind power
Baltic Sea coast. Satellite-derived shoreline change estimates over potential are projected to decrease in Western North America (high
1984–2015 indicate shoreline retreat rates of around 0.5 m yr –1 confidence), with differences between global and regional models
along the sandy coasts of Central Europe and the Mediterranean lending low confidence elsewhere. {11.4, 11.7, 12.4.6.3}
and more or less stable shorelines in Northern Europe. Over the
period 1982–2016, the coastlines of Europe experienced on average Snow and ice: It is likely that some high-latitude regions will
more than 2.0 MHW per year, with the eastern Mediterranean and experience an increase in winter snow water equivalent due to the
Scandinavia experiencing 2.5–3 MHWs per year. {12.4.5.5} snowfall increase prevailing over the warming trend. At sustained
GWLs between 3°C and 5°C, nearly all glacial mass in Western
TS.4.3.2.6 North America Canada and Western North America will disappear (medium
confidence). {9.5.1, 9.5.3, 12.4.6.4, Atlas.9.4}
Additional regional changes in North America, besides those
features described in Section TS.4.3.1, include changes Coastal and oceanic: Around North America, relative sea level has
in North American wet and dry CIDs, which are largely increased over the last three decades at a rate lower than GMSL
organized by the north-east (more wet) to south-west (more in the subpolar North Atlantic and in the East Pacific, while it has
dry) pattern of mean precipitation change, although heavy increased at a rate higher than GMSL in the subtropical North Atlantic.
precipitation increases are widespread (high confidence). Observations indicate that episodic coastal flooding is increasing
Increasing evaporative demand will expand agricultural along many coastlines in North America. Shoreline retreat rates of
and ecological drought and fire weather (particularly in around 1 m yr –1 have been observed during 1984–2015 along the
summertime) in Central North America, Western North sandy coasts of North-Western North America and Northern Central

141
 Technical Summary

America, while portions of the United States Gulf Coast have seen Permafrost warming, loss of seasonal snow cover, and
a retreat rate approaching 2.5 m yr –1. Sandy shorelines along Eastern glacier melt will be widespread (high confidence). There is
North America and Western North America have remained more or high confidence that both the Greenland and Antarctic ice
less stable during 1984–2014, but a shoreline progradation rate of sheets have lost mass since 1992 and will continue to lose
around 0.5 m yr –1 has been observed in North-Eastern North America. mass throughout this century under all emissions scenarios.
{12.4.6.5} Relative sea level and coastal flooding are projected to
increase in areas other than regions with substantial land
TS.4.3.2.7 Small Islands uplift (medium confidence). {2.3, 3.4, 4.3, 4.5, 7.4, 8.2, 8.4,
Box 8.2, 9.5, 12.4.9, Atlas.11.1, Atlas.11.2}
Additional regional changes in Small Islands, besides those
features described in Section TS.4.3.1, include a likely In addition to the main changes summarized above and in Section
decrease in rainfall during boreal summer in the Caribbean TS.4.3.1, further details are given below.
and in some parts of the Pacific islands poleward of 20°
latitude in both the Northern and Southern Hemispheres. Heat and cold: Changes in Antarctica showed larger spatial
TS These drying trends will likely continue in coming decades. variability, with very likely warming in the Antarctic Peninsula since
Fewer but more intense tropical cyclones are projected the 1950s and no overall trend in East Antarctica. Less warming
starting from a 2°C GWL (medium confidence). {9.6, 11.3, and weaker polar amplification are projected as very likely over
11.4, 11.7, 11.9, 12.4.7, Atlas.10.2, Atlas.10.4, Cross-Chapter the Antarctic than in the Arctic, with a weak polar amplification
Box Atlas.2} projected as very likely by the end of the 21st century. {4.3.1, 4.5.1,
7.4.4, 12.4.9.1, Atlas.11.1, Atlas.11.2}
In addition to the main changes summarized above and in Section
TS.4.3.1, further details are given below. Wet and dry: Recent decades have seen a general decrease in
Arctic aridity (high confidence), with increased moisture transport
Heat and cold: It is very likely that most Small Islands have warmed leading to higher precipitation, humidity and streamflow and a
over the period of instrumental records, and continued temperature corresponding decrease in dry days. Antarctic precipitation showed
increases in the 21st century will further increase heat stress in these a positive trend during the 20th century. The water cycle is projected
regions. {11.3.2, 11.9, 12.4.7.1, Atlas.10.2, Atlas.10.4, Cross-Chapter to intensify in both polar regions, leading to higher precipitation
Box Atlas.2} totals (and a shift to more heavy precipitation) and higher fraction
of precipitation falling as rain. In the Arctic, this will result in higher
Wet and dry: Observed and projected rainfall trends vary spatially river flood potential and earlier meltwater flooding, altering seasonal
across the Small Islands. Higher evapotranspiration under a warming characteristics of flooding (high confidence). A lengthening of the fire
climate can partially offset future increases or amplify future season (medium confidence) and encroachment of fire regimes into
reductions in rainfall, resulting in increased aridity as well as more tundra regions (high confidence) are projected. {8.2.3, 8.4.1, Box 8.2,
severe agricultural and ecological drought in the Caribbean (medium 9.4.1, 9.4.2, 12.4.9.2, Atlas.11.1, Atlas.11.2}
confidence). {11.4.2, 11.9, 12.4.7.2, Atlas.10.2, Atlas.10.4, Cross-
Chapter Box Atlas.2} Wind: There is medium confidence in mean wind decrease over
the Russian Arctic and Arctic North-East North America, but low
Wind: Global changes indicate that Small Islands will face fewer confidence of changes in other Arctic regions and Antarctica.
but more intense TCs, with spatial inconsistency in projections given {12.4.9.3}
poleward shifts in TC tracks (medium confidence). {11.7.1.2, 11.7.1.5,
12.4.7.3} Snow and ice: Reductions in spring snow cover extent have
occurred across the Northern Hemisphere since at least 1978 (very
Coastal and oceanic: Continued relative sea level rise is very likely high confidence). Permafrost warming and thawing have been
in the ocean around Small Islands and, along with storm surges widespread in the Arctic since the 1980s (high confidence), causing
and waves, will exacerbate coastal inundation with the potential to strong heterogeneity in surface conditions. There is high confidence
increase saltwater intrusion into aquifers in small islands. Shoreline in future glacier- and ice-sheet loss, permafrost warming, decreasing
retreat is projected along sandy coasts of most small islands (high permafrost extent and decreasing seasonal duration and extent of
confidence). {9.6.3.3, 12.4.7.4, Cross-Chapter Box Atlas.2} snow cover in the Arctic. Decline in seasonal sea ice coverage along
the majority of the Arctic coastline in recent decades is projected to
TS.4.3.2.8 Polar continue, contributing to an increase in coastal hazards (including
open water storm surge, coastal erosion and flooding). {2.3.2, 3.4.2,
It is virtually certain that surface warming in the Arctic 3.4.3, 9.4.1, 9.4.2, 9.5, 12.4.6, 12.4.9, Atlas.11.2}
will continue to be more pronounced than the global
average warming over the 21st century. An intensification Coastal and oceanic: Higher sea levels contribute to high
of the polar water cycle will increase mean precipitation, confidence for projected increases of Arctic coastal flooding and
with precipitation intensity becoming stronger and more higher coastal erosion (aided by sea ice loss) (medium confidence),
likely to be rainfall rather than snowfall (high confidence).

142
 Technical Summary

with lower confidence for those regions with substantial land uplift TS.4.3.2.10 Other Typological Domains
(Arctic North-East North America and Greenland). {12.4.9.5}
Some types of regions found in different continents face
TS.4.3.2.9 Ocean common climate challenges regardless of their location.
These include biodiversity hot spots that will very likely
The Indian Ocean, western equatorial Pacific Ocean and see even more extreme heat and droughts, mountain areas
western boundary currents have warmed faster than the where a projected raising in the freezing level height will
global average (very high confidence), with the largest alter snow and ice conditions (high confidence), and tropical
changes in the frequency of marine heatwaves (MHWs) forests that are increasingly prone to fire weather (medium
projected in the western tropical Pacific and the Arctic confidence). {8.4, Box 8.2, 9.5, 12.3, 12.4}
Ocean (medium confidence). The Pacific and Southern Ocean
are projected to freshen and the Atlantic to become more Biodiversity hotspots located around the world will each face unique
saline (medium confidence). Anthropogenic warming is very challenges in CID changes. Heat, drought and length of dry season,
likely to further decrease ocean oxygen concentrations, and wildfire weather, sea surface temperature and deoxygenation are
this deoxygenation is expected to persist for thousands relevant drivers to terrestrial and freshwater ecosystems and have TS
of years (medium confidence). Arctic sea ice losses are marked increasing trends. {12.3, 12.4.10.1}
projected to continue, leading to a practically ice-free Arctic
in September by the end of the 21st century under high Desert and semi-arid areas are strongly affected by CIDs such as
CO2 emissions scenarios (high confidence). {2.3, 5.3, 9.2, 9.3, extreme heat, drought and dust storms, with large-scale aridity
Box 9.2, 12.3.6, 12.4.8} trends contributing to expanding drylands in some regions (high
confidence). {12.3, 12.4.10.3}
In addition to the main changes summarized above and in Section
TS.4.3.1, further details are given below. Average warming in mountain areas varies with elevation, but
the pattern is not globally uniform (medium confidence). Extreme
Ocean surface temperature: The Southern Ocean, the eastern precipitation is projected to increase in major mountainous regions
equatorial Pacific, and the North Atlantic Ocean have warmed more (medium to high confidence depending on location), with potential
slowly than the global average or slightly cooled. Global warming cascading consequences of floods, landslides and lake outbursts
of 2°C above 1850–1900 levels would result in the exceedance of in all scenarios (medium confidence). {8.4.1.5, Box 8.2, 9.5.1.3,
numerous hazard thresholds for pathogens, seagrasses, mangroves, 9.5.3.3, 9.5.2.3, Cross-Chapter Box 10.4, 11.5.5, 12.3, 12.4.1–12.4.6,
kelp forests, rocky shores, coral reefs and other marine ecosystems 12.4.10.4}
(medium confidence). {9.2.13, 12.4.8}
Most tropical forests are challenged by a mix of emerging warming
Marine heatwaves: Moderate increases in MHW frequency are trends that are particularly large in comparison to historical variability
projected for mid-latitudes, and only small increases are projected (medium confidence). Water cycle changes bring prolonged drought,
for the Southern Ocean (medium confidence). Under the SSP5-8.5 longer dry seasons and increased fire weather to many tropical
scenario, permanent MHWs (more than 360 days per year) are forests (medium confidence). {10.5, 12.3, 12.4}
projected to occur in the 21st century in parts of the tropical ocean,
the Arctic Ocean, and around 45°S; however, the occurrence of
such permanent MHWs can be largely avoided under the SSP1-2.6
scenario. {Box 9.2, 12.4.8}

Ocean acidity: With the rising CO2 concentration, the ocean surface
pH has declined globally over the past four decades (virtually certain).
{2.3.3.5, 5.3.3.2, 12.4.8}

Ocean salinity: At the basin scale, it is very likely that the Pacific and
the Southern Ocean have freshened while the Atlantic has become
more saline. {2.3.3.2, 9.2.2.2, 12.4.8}

Dissolved oxygen: In recent decades, low oxygen zones in ocean

ecosystems have expanded. {2.3.4.2, 5.3.3.2, 12.4.8}

Sea ice: Arctic perennial sea ice is being replaced by thin, seasonal
ice, with earlier spring melt and delayed fall freeze up. There is no
clear trend in the Antarctic sea ice area over the past few decades
and low confidence in its future change. {2.3.2.1.1, 9.3.1.1, 12.4.8,
12.4.9}

143
 Technical Summary

Box TS.14 | Urban Areas

With global warming, urban areas and cities will be affected by more frequent occurrences of extreme climate events,
such as heatwaves, with more hot days and warm nights as well as sea level rise and increases in tropical cyclone
storm surge and rainfall intensity that will increase the probability of coastal city flooding (high confidence). {Box 10.3,
11.3, 11.5, 12.3, 12.4}

Urban areas have special interactions with the climate system, for instance in terms of heat islands and altering the water cycle, and
thereby will be more affected by extreme climate events such as extreme heat (high confidence). With global warming, increasing
relative sea level compounded by increasing tropical cyclone storm surge and rainfall intensity will increase the probability of coastal
city flooding (high confidence). Arctic coastal settlements are particularly exposed to climate change due to sea ice retreat (high
confidence). Improvements in urban climate modelling and climate monitoring networks have contributed to understanding the
mutual interaction between regional and urban climate (high confidence). {Box 10.3, 11.3, 11.5, 12.3, 12.4}
TS
Despite having a negligible effect on global surface temperature (high confidence), urbanization has exacerbated the effects of global
warming through its contribution to the observed warming trend in and near cities, particularly in annual mean minimum temperature
(very high confidence) and increases in mean and extreme precipitation over and downwind of the city, especially in the afternoon
and early evening (medium confidence). {2.3, Box 10.3, 11.3, 11.4, 12.3, 12.4}

Combining climate change projections with urban growth scenarios, future urbanization will amplify (very high confidence) the
projected local air temperature increase, particularly by strong influence on minimum temperatures, which is approximately comparable
in magnitude to global warming (high confidence). Compared to present day, large implications are expected from the combination
of future urban development and more frequent occurrence of extreme climate events, such as heatwaves, with more hot days and
warm nights adding to heat stress in cities (very high confidence). {Box 10.2, 11.3, 12.4}

Both sea levels and air temperatures are projected to rise in most coastal settlements (high confidence). There is high confidence in
an increase in pluvial flood potential in urban areas where extreme precipitation is projected to increase, especially at high global
warming levels. {11.4, 11.5, 12.4}

144