Editing IPCC:Wg1:Chapter:Chapter-1-comments (section)

== 1.4 AR6 Foundations and Concepts ==

<div id="h1-5-siblings" class="h1-siblings"></div>

The AR6 WGI builds on previous assessments using well established foundations and concepts. This section highlights some of the cross-cutting methods applied in the climate change literature and topics discussed repeatedly throughout this Report. First, the choices related to ‘baselines’, or ‘reference periods’, are highlighted ( [[#1.4.1|Section 1.4.1]] ), including a specific discussion on the pre-industrial baseline used in AR6 WGI (Cross-Chapter Box 1.2). The relationships between long-term trends, climate variability and the concept of ‘emergence of changes’ ( [[#1.4.2|Section 1.4.2]] ) and the sources of uncertainty in climate simulations ( [[#1.4.3|Section 1.4.3]] ) are discussed next. The topic of low-likelihood outcomes, storylines, abrupt changes and surprises follows ( [[#1.4.4|Section 1.4.4]] ), including a description of AR6 WGI risk framing (Cross-Chapter Box 1.3). The Cross-Working Group Box on Attribution describes attribution methods, including those for extreme events. Various sets of geographical regions used in later chapters are also defined and introduced ( [[#1.4.5|Section 1.4.5]] ).

<div id="1.4.1" class="h2-container"></div>

<span id="baselines-reference-periods-and-anomalies"></span>
=== 1.4.1 Baselines, Reference Periods and Anomalies ===

<div id="h2-19-siblings" class="h2-siblings"></div>

Several baselines or reference periods are used consistently throughout AR6 WGI. Baseline refers to a period against which differences are calculated, whereas reference period is used more generally to indicate a time period of interest, or a period over which some relevant statistics are calculated (Glossary). Variations in observed and simulated climate variables over time are often presented as ‘anomalies’, that is, the differences relative to a baseline, rather than using the absolute values. This is done for several reasons.

First, anomalies are often used when combining data from multiple locations, because the absolute values can vary over small spatial scales which are not densely observed or simulated, whereas anomalies are representative for much larger scales (e.g., for temperature; [[#Hansen--1987|Hansen and Lebedeff, 1987]] ). Since their baseline value is zero by definition, anomalies are also less susceptible to biases arising from changes in the observational network. Second, the seasonality in different climate indicators can be removed using anomalies to more clearly distinguish variability from long-term trends.

Third, different datasets can have different absolute values for the same climate variable that should be removed to allow effective comparisons of variations over time. This is often required when comparing climate simulations with each other, or when comparing simulations with observations, as simulated climate variables are also affected by model bias that can be removed when they are presented as anomalies. It can also be required when comparing observational datasets or reanalyses ( [[#1.5.2|Section 1.5.2]] ) with each other, due to systematic differences in the underlying measurement system (Figure 1.11). Understanding the reasons for any absolute difference is important, but whether the simulated absolute value matters when projecting future change will depend on the variable of interest. For example, there is not a strong relationship between climate sensitivity of a model (which is an indicator of the degree of future warming) and the simulated absolute global surface temperature ( [[#Mauritsen--2012|Mauritsen et al., 2012]] ; [[#Hawkins--2016|Hawkins and Sutton, 2016]] ).

<div id="_idContainer039" class="_idGenObjectStyleOverride-1"></div>

<!-- START IMG -->
<!-- IMG FILE -->
[[File:a16dd7036cdc2bae1bfbdae8995f8310 IPCC_AR6_WGI_Figure_1_11.png]]
<!-- IMG TITLE + CAPTION -->

'''Figure 1.11 |''' '''Choice of baseline matters when comparing observations and model simulations.''' Global mean surface air temperature (GSAT, grey) from a range of CMIP6 historical simulations (1850–2014; 25 models) and SSP1-2.6 (2015–2100) using absolute values '''(top)''' and anomalies relative to two different baselines: 1850–1900 '''(middle)''' and 1995–2014 '''(bottom)''' . An estimate of GSAT from a reanalysis (ERA-5, orange, 1979–2020) and an observation-based estimate of global mean surface air temperature (GMST) (Berkeley Earth, black, 1850–2020) are shown, along with the mean GSAT for 1961–1990 estimated by [[#Jones--1999|Jones et al. (1999)]] , light blue shading (14.0°C ± 0.5°C). Using the more recent baseline (bottom) allows the inclusion of datasets which do not include the periods of older baselines. The middle and bottom panels have scales which are the same size but offset. Further details on data sources and processing are available in the chapter data table (Table 1.SM.1).
<!-- END IMG -->

For some variables, such as precipitation, anomalies are often expressed as percentages in order to more easily compare changes in regions with very different climatological means. However, for situations where there are important thresholds (e.g., phase transitions around 0°C) or for variables which can only take a particular sign or be in a fixed range (e.g., sea ice extent or relative humidity), absolute values are normally used.

The choice of a baseline period has important consequences for evaluating both observations and simulations of the climate, for comparing observations with simulations, and for presenting climate projections. There is usually no perfect choice of baseline as many factors have to be considered and compromises may be required ( [[#Hawkins--2016|Hawkins and Sutton, 2016]] ). It is important to evaluate the sensitivity of an analysis or assessment to the choice of the baseline.

For example, the collocation of observations and reanalyses within the model ensemble spread depends on the choice of the baseline, and uncertainty in future projections of climate is reduced if using a more recent baseline, especially for the near term (Figure 1.11). The length of an appropriate baseline or reference period depends on the variable being considered, the rates of change of the variable and the purpose of the chosen period, but is usually 20 to 50 years long. The World Meteorological Organization (WMO) uses 30-year periods to define ‘climate normals’, which indicate conditions expected to be experienced in a given location.

For AR6WGI, the period 1995–2014 is used as a baseline to calculate the changes in future climate using model projections and also as a ‘modern’ or ‘recent past’ reference period when estimating past observed warming. The equivalent period in AR5 was 1986–2005, and in SR1.5, SROCC and SRCCL it was 2006–2015. The primary reason for the different choice in AR6 is that 2014 is the final year of the historical CMIP6 simulations. These simulations subsequently assume different emissions scenarios and so choosing any later baseline end date would require selecting a particular emissions scenario. For certain assessments, the most recent decade possible (e.g., 2010–2019 or 2011–2020, depending on the availability of observations) is also used as a reference period (Cross-Chapter Box 2.3).

Figure 1.12 shows changes in observed global mean surface temperature (GMST) relative to 1850–1900 and illustrates observed global warming levels for a range of reference periods that are either used in AR6 or were used in previous IPCC reports. This allows changes to be calculated between different periods and compared to previous assessments. For example, AR5 assessed the change in GMST from the 1850–1900 baseline to 1986–2005 reference period as 0.61 [0.55 to 0.67] °C, whereas it is now assessed to be 0.69 [0.52 to 0.82] °C using improved GMST datasets (Cross-Chapter Box 2.3).

<div id="_idContainer041" class="_idGenObjectStyleOverride-1"></div>

<!-- START IMG -->
<!-- IMG FILE -->
[[File:fb052bf0932690600517b1ce338f6255 IPCC_AR6_WGI_Figure_1_12.png]]
<!-- IMG TITLE + CAPTION -->

'''Figure 1.12 |''' '''Global warming over the instrumental period.''' Observed global mean surface temperature (GMST) from four datasets, relative to the average temperature of 1850–1900 in each dataset (see Cross-Chapter Box 2.3 and [[IPCC:Wg1:Chapter:Chapter-2#2.3.1.1|Section 2.3.1.1]] for more details). The shaded grey band indicates the assessed ''likely'' range for the period around 1750 (Cross-Chapter Box 1.2). Different reference periods are indicated by the coloured horizontal lines, and an estimate of total GMST change up to that period is given, enabling a translation of the level of warming between different reference periods. The reference periods are all chosen because they have been used in AR6 or previous IPCC assessment reports. The value for the 1981–2010 reference period, used as a ‘climate normal’ period by the World Meteorological Organization, is the same as the 1986–2005 reference period shown. Further details on data sources and processing are available in the chapter data table (Table 1.SM.1).
<!-- END IMG -->

The commonly used metric for global surface warming tends to be GMST but, as shown in Figure 1.11, climate model simulations tend to use global surface air temperature (GSAT). Although GMST and GSAT are closely related, the two measures are physically distinct. GMST is a combination of land surface air temperature (LSAT) and sea surface temperature (SST), whereas GSAT is surface air temperatures over land, ocean and ice. A key development in AR6 is the assessment that long-term changes in GMST and GSAT differ by at most 10% in either direction, with ''low confidence'' in the sign of any differences (see Cross Chapter Box 2.3 for details).

Three future reference periods are used in AR6 WGI for presenting projections: ''near term'' (2021–2040), ''mid-term'' (2041–2060) and ''long-term'' (2081–2100; Figure 1.11). In AR6, 20-year reference periods are considered long enough to show future changes in many variables when averaging over ensemble members of multiple models, and short enough to enable the time dependence of changes to be shown throughout the 21st century. Projections with alternative recent baselines (such as 1986–2005 or the current WMO climate-normal period of 1981–2010) and a wider range of future reference periods are presented in the Interactive Atlas. Note that ‘long term’ is also sometimes used in a more general sense to refer to durations of centuries to millennia when examining past climate, as well as future climate change beyond the year 2100. Cross-Chapter Box 2.1 discusses the paleo-reference periods used in AR6.

<div id="cross-chapter-box-1.2" class="h2-container box-container"></div>

'''Cross-Chapter Box 1.2 | Changes in Global Temperature Betwee''' '''n 1750 and 1850'''

<div id="h2-20-siblings" class="h2-siblings"></div>

'''Contributing Authors:''' Ed Hawkins (United Kingdom), Paul Edwards (United States of America), Piers Forster (United Kingdom), Darrell S. Kaufman (United States of America), Jochem Marotzke (Germany), Malte Meinshausen (Australia/Germany), Maisa Rojas (Chile), Bjørn H. Samset (Norway), Peter Thorne (Ireland/United Kingdom)

The Paris Agreement aims to limit global temperatures to specific thresholds ‘above pre-industrial levels’. In AR6 WGI, as in previous IPCC reports, observations and projections of changes in global temperature are generally expressed relative to 1850–1900 as an approximate pre-industrial state (SR1.5, [[#IPCC--2018|IPCC, 2018]] ). This is a pragmatic choice based upon data availability considerations, though both anthropogenic and natural changes to the climate occurred before 1850. The remaining carbon budgets, the chance of crossing global temperature thresholds, and projections of extremes and sea level rise at a particular level of global warming can all be sensitive to the chosen definition of the approximate pre-industrial baseline ( [[#Millar--2017b|Millar et al., 2017b]] ; [[#Schurer--2017|Schurer et al., 2017]] ; [[#Pfleiderer--2018|Pfleiderer et al., 2018]] ; [[#Rogelj--2019|Rogelj et al., 2019]] ; [[#Tokarska--2019|Tokarska et al., 2019]] ). This Cross-Chapter Box assesses the evidence on change in radiative forcing and global temperature from the period around 1750 to 1850–1900; variations in the climate before 1750 are discussed in Chapter 2.

Although there is some evidence for human influence on climate before 1750 (e.g., [[#Ruddiman--2001|Ruddiman and Thomson, 2001]] ; [[#Koch--2019|Koch et al., 2019]] ), the magnitude of the effect is still disputed (Section 5.1.2.3; e.g., [[#Joos--2004|Joos et al., 2004]] ; J. [[#Beck--2018|]] [[#Beck--2018|Beck et al., 2018]] ), and most studies analyse the human influence on climate over the industrial period. Historically, the widespread use of coal-powered machinery started the Industrial Revolution in Britain in the late 18th century ( [[#Ashton--1997|Ashton, 1997]] ), but the global effects were small for several decades. In line with this, previous IPCC assessment reports considered changes in radiative forcing relative to 1750, and temperature changes were often reported relative to the ‘late 19th century’. The AR5 and SR1.5 made the specific pragmatic choice to approximate pre-industrial global temperatures by using the average of the 1850 – 1900 period, when permanent surface observing networks emerged that provide sufficiently accurate and continuous measurements on a near-global scale (Sections [[#1.3.1|1.3.1]] and [[IPCC:Wg1:Chapter:Chapter-2#2.3.1.1|2.3.1.1]] ), and because model simulations of the historical period used 1850 as their start date. For the same reasons, to ensure continuity with previous assessments, and because of larger uncertainties and lower confidence in climatic changes before 1850 than after, AR6 makes the same choice to approximate pre-industrial global temperatures by using the the average of the 1850–1900 period.

Here weassess improvements in our understanding of climatic changes in the period 1750–1850. Anthropogenic influences on climate between 1750 and 1900 were primarily increased anthropogenic GHG and aerosol emissions, and changes in land use. Between 1750 and 1850 atmospheric CO <sub>2</sub> levels increased from about 278 ppm to about 285 ppm (equivalent to around 3 years of current rates of increase; Chapter 2, [[IPCC:Wg1:Chapter:Chapter-2#2.2.3|Section 2.2.3]] ), corresponding to about 55 GtCO <sub>2</sub> in the atmosphere. Estimates of emissions from fossil fuel burning (about 4 GtCO <sub>2</sub> , [[#Boden--2017|Boden et al., 2017]] ) cannot explain the pre-1850 increase, so CO <sub>2</sub> emissions from land-use changes are implicated as the dominant source. The atmospheric concentration of other GHGs also increased over the same period, and there was a cooling influence from other anthropogenic radiative forcings (such as aerosols and land-use changes), but with a larger uncertainty than for GHGs (Sections 2.2.6 and 7.3.5.2, and Cross-Chapter Box 1.2, Figure 1; e.g., [[#Carslaw--2017|Carslaw et al., 2017]] ; [[#Owens--2017|Owens et al., 2017]] ; [[#Hamilton--2018|Hamilton et al., 2018]] ). It is ''likely'' that there was a net anthropogenic forcing of 0.0 – 0.3 Wm <sup>–2</sup> in 1850 – 1900 relative to 1750 ( ''medium confidence'' ). The net radiative forcing from changes in solar activity and volcanic activity in 1850 – 1900, compared to the period around 1750, is estimated to be smaller than ±0.1 W m <sup>–2</sup> , but note there were several large volcanic eruptions between 1750 and 1850 (Cross-Chapter Box 1.2, Figure 1).

Several studies since AR5 have estimated changes in global temperatures following industrialisation and before 1850. [[#Hawkins--2017|Hawkins et al. (2017)]] used observations, radiative forcing estimates and model simulations to estimate the warming from 1720–1800 until 1986–2005 and assessed a ''likely'' range of 0.55°C–0.80°C, slightly broader than the equivalent range starting from 1850–1900 (0.6°C–0.7°C). From proxy evidence, [[#PAGES%202k%20Consortium--2019|PAGES 2k Consortium (2019)]] found that GMST for 1850–1900 was 0.02 [–0.22 to +0.16] °C warmer than the 30-year period centred on 1750. [[#Schurer--2017|Schurer et al. (2017)]] used climate model simulations of the last millennium to estimate that the increase in GHG concentrations before 1850 caused an additional ''likely'' range of 0.0°C –0.2°C global warming when considering multiple reference periods. [[#Haustein--2017|Haustein et al. (2017)]] implies an additional warming of around 0.05°C attributable to human activity from 1750 to 1850–1900, and the AR6 emulator (Section 7.3.5.3) estimates the ''likely'' range of this warming to be 0.04°C–0.14°C.

Combining these different sources of evidence, we assess that from the period around 1750 to 1850–1900 there was a change in global temperature of around 0.1 [–0.1 to +0.3] °C ( ''medium confidence'' ), with an anthropogenic component in a ''likely'' range of 0.0°C–0.2°C ( ''medi'' ''um confidence'' ).

<!-- START IMG -->
<!-- IMG FILE -->
[[File:6a7bf6431e46d4e07cd5a74e974a4398 IPCC_AR6_WGI_CCBox_1_2_Figure_1.png]]
<!-- IMG TITLE + CAPTION -->

'''Cross-Chapter Box 1.2, Figure 1''' | '''Changes in radiative forcing from 1750–2019''' . The radiative forcing estimates from the AR6 emulator (Cross-Chapter Box 7.1) are split into GHG, other anthropogenic (mainly aerosols and land use) and natural forcings, with the average over the 1850–1900 baseline shown for each. Further details on data sources and processing are available in the chapter data table (Table 1.SM.1).
<!-- END IMG -->
<div id="1.4.2" class="h2-container"></div>

<span id="variability-and-emergence-of-the-climate-change-signal"></span>
=== 1.4.2 Variability and Emergence of the Climate Change Signal ===

<div id="h2-21-siblings" class="h2-siblings"></div>

Climatic changes since the pre-industrial era are a combination of long-term anthropogenic changes and natural variations on time scales from days to decades. The relative importance of these two factors depends on the climate variable or region of interest. Natural variations consist of both natural radiatively forced trends (e.g., due to volcanic eruptions or solar variations) and ‘internal’ fluctuations of the climate system which occur even in the absence of any radiative forcings. The internal ‘modes of variability’, such as the El Niño–Southern Oscillation (ENSO) and the North Atlantic Oscillation (NAO), are discussed further in Annex IV.

<div id="1.4.2.1" class="h3-container"></div>

<span id="climate-variability-can-influence-trends-over-short-periods"></span>
==== 1.4.2.1 Climate Variability Can Influence Trends Over Short Periods ====

<div id="h3-13-siblings" class="h3-siblings"></div>

Natural variations in both weather and longer time scale phenomena can temporarily mask or enhance any anthropogenic trends (e.g., [[#Deser--2012|Deser et al., 2012]] ; [[#Kay--2015|Kay et al., 2015]] ). These effects are more important on small spatial and temporal scales but can also occur on the global scale (Cross-Chapter Box 3.1).

Since AR5, many studies have examined the role of internal variability through the use of ‘large ensembles’. Each such ensemble consists of many different simulations by a single climate model for the same time period and using the same radiative forcings. These simulations differ only in their phasing of the internal climate variations (also see [[#1.5.4.2|Section 1.5.4.2]] ). A set of illustrative examples using one such large ensemble ( [[#Maher--2019|Maher et al., 2019]] ) demonstrates how variability can influence trends on decadal time scales (Figure 1.13). The long-term anthropogenic trends in this set of climate indicators are clearly apparent when considering the ensemble as a whole (grey shading), and all the individual ensemble members have very similar trends for ocean heat content (OHC), which is a robust estimate of the total energy stored in the climate system (e.g., [[#Palmer--2014|Palmer and McNeall, 2014]] ). However, the individual ensemble members can exhibit very different decadal trends in global surface air temperature (GSAT), UK summer temperatures, and Arctic sea ice variations. More specifically, for a representative 11-year period, both positive and negative trends can be found in all these surface indicators, even though the long-term trend is for increasing temperatures and decreasing sea ice. Periods in which the long-term trend is substantially masked or enhanced for more than 20 years are also visible in these regional examples. This highlights the fact that observations are expected to exhibit short-term trends which are larger or smaller than the long-term trend or that differ from the average projected trend from climate models, especially on continental spatial scales or smaller (Cross-Chapter Box 3.1). The actual observed trajectory can be considered as one realization of many possible alternative worlds that experienced different weather; this is also demonstrated by the construction of ‘observation-based large ensembles’, which are alternate possible realizations of historical observations that retain the statistical properties of observed regional weather (e.g., [[#McKinnon--2018|McKinnon and Deser, 2018]] ).

<div id="_idContainer043" class="_idGenObjectStyleOverride-1"></div>

<!-- START IMG -->
<!-- IMG FILE -->
[[File:fc7c3fbc2db6d0af8a18d4187bea1d57 IPCC_AR6_WGI_Figure_1_13.png]]
<!-- IMG TITLE + CAPTION -->

'''Figure 1.13 |''' '''Simulated changes in various climate indicators under historical and RCP4.5 scenarios using the MPI ESM Grand Ensemble.''' The grey shading shows the 5–95% range from the 100-member ensemble. The coloured lines represent individual example ensemble members, with linear trends for the 2011–2021 period indicated by the dashed lines. Changes in ocean heat content (OHC) over the top 2000 m represents the integrated signal of global warming '''(left)''' . The '''top row''' shows surface air temperature-related indicators (annual GSAT change and UK summer temperatures) and The '''bottom row''' shows Arctic sea ice-related indicators (annual ice volume and September sea ice extent). For smaller regions and for shorter time-period averages the variability increases and simulated short-term trends can temporarily mask or enhance anthropogenic changes in climate. Data from [[#Maher--2019|Maher et al. (2019)]] . Further details on data sources and processing are available in the chapter data table (Table 1.SM.1).
<!-- END IMG -->

<div id="1.4.2.2" class="h3-container"></div>

<span id="the-emergence-of-the-climate-change-signal"></span>
==== 1.4.2.2 The Emergence of the Climate Change Signal ====

<div id="h3-14-siblings" class="h3-siblings"></div>

In the 1930s it was noted that temperatures were increasing at both local and global scales (Figure 1.8; [[#Kincer--1933|Kincer, 1933]] ; [[#Callendar--1938|Callendar, 1938]] ). At the time it was unclear whether the observed changes were part of a longer-term trend or a natural fluctuation; the ‘signal’ had not yet clearly emerged from the ‘noise’ of natural variability. Numerous studies have since focused on the emergence of changes in temperature using instrumental observations (e.g., [[#Madden--1980|Madden and Ramanathan, 1980]] ; [[#Wigley--1981|Wigley and Jones, 1981]] ; [[#Mahlstein--2011|Mahlstein et al., 2011]] , 2012; [[#Lehner--2015|Lehner and Stocker, 2015]] ; [[#Lehner--2017|Lehner et al., 2017]] ) and paleo-temperature data (e.g., [[#Abram--2016|Abram et al., 2016]] ).

Since the IPCC Third’s Assessment Report in 2001, the observed signal of climate change has been unequivocally detected at the global scale ( [[#1.3|Section 1.3]] ), and this signal is increasingly emerging from the noise of natural variability on smaller spatial scales and in a range of climate variables (FAQ 1.2). In this Report emergence of a climate change signal or trend refers to when a change in climate (the ‘signal’) becomes larger than the amplitude of natural or internal variations (defining the ‘noise’). This concept is often expressed as a ‘signal-to-noise’ ratio (S/N) and emergence occurs at a defined threshold of this ratio (e.g., S/N &gt;1 or 2). Emergence can be estimated using observations and/or model simulations and can refer to changes relative to a historical or modern baseline (Section 12.5.2 and Glossary). The concept can also be expressed in terms of time (the ‘time of emergence’; Glossary) or in terms of a global warming level (Section 11.2.5; [[#Kirchmeier-Young--2019|Kirchmeier-Young et al., 2019]] ) and is also used to refer to a time when we can expect to see a response of mitigation activities that reduce emissions of GHGs or enhance their sinks (emergence with respect to mitigation; [[IPCC:Wg1:Chapter:Chapter-4#4.6.3.1|Section 4.6.3.1]] ). Whenever possible, emergence should be discussed in the context of a clearly defined level of S/N or other quantification, such as ‘the signal has emerged at the level of S/N &gt;2’, rather than as a simple binary statement. For an extended discussion, see [https://www.ipcc.ch/report/ar6/wg1/chapter/chapter-10 Chapter 10] (Section 10.4.3).

Related to the concept of emergence is the detection of change (Chapter 3). Detection of change is defined as the process of demonstrating that some aspect of the climate, or a system affected by climate, has changed in some defined statistical sense, often using spatially aggregating methods that try to maximize S/N, such as ‘fingerprints’ (e.g., [[#Hegerl--1996|Hegerl et al., 1996]] ), without providing a reason for that change. An identified change is detected in observations if its likelihood of occurrence by chance due to internal variability alone is determined to be small, for example, &lt;10% (Glossary).

An example of observed emergence in surface air temperatures is shown in Figure 1.14. Both the largest changes in temperature and the largest amplitude of year-to-year variations are observed in the Arctic, with lower latitudes showing less warming and smaller year-to-year variations. For the six example regions shown in Figure 1.14, the emergence of changes in temperature is more apparent in Northern South America, East Asia and Central Africa, than for northern North America or Northern Europe. This pattern was predicted by [[#Hansen--1988|Hansen et al. (1988)]] and noted in subsequent observations by [[#Mahlstein--2011|Mahlstein et al. (2011)]] (Sections 10.3.4.3 and 12.5.2). Overall, tropical regions show earlier emergence of temperature changes than at higher latitudes ( ''hi'' ''gh confidence'' ).

<div id="_idContainer045" class="_idGenObjectStyleOverride-1"></div>

<!-- START IMG -->
<!-- IMG FILE -->
[[File:f99318b822c49734ff81d7990164dfbb IPCC_AR6_WGI_Figure_1_14.png]]
<!-- IMG TITLE + CAPTION -->

'''Figure 1.14 |''' '''The observed emergence of changes in temperature.''' '''(Top left)''' The total change in temperature estimated for 2020 relative to 1850–1900 (following [[#Hawkins--2020|Hawkins et al., 2020]] ), showing the largest warming occurring in the Arctic. '''(Top right)''' The amplitude of estimated year-to-year variations in temperature. '''(Middle''' '''left)''' The ratio of the observed total change in temperature and the amplitude of temperature variability (the ‘signal-to-noise (S/N) ratio’), showing that the warming is most apparent in the tropical regions (also see FAQ 1.2). '''(Middle right)''' The global warming level at which the change in local temperature becomes larger than the local year-to-year variability. The '''bottom''' panels show time series of observed annual mean surface air temperatures over land in various example regions, as indicated by the boxes in the top-left panel. The 1 and 2 standard deviations ( σ ) of estimated year-to-year variations for that region are shown by the pink shaded bands. Observed temperature data from Berkeley Earth ( [[#Rohde--2020|Rohde and Hausfather, 2020]] ). Further details on data sources and processing are available in the chapter data table (Table 1.SM.1).
<!-- END IMG -->

Since AR5, the emergence of projected future changes has also been extensively examined, in variables including surface air temperature ( [[#Hawkins--2012|Hawkins and Sutton, 2012]] ; [[#Kirtman--2013|Kirtman et al., 2013]] ; [[#Tebaldi--2013|Tebaldi and Friedlingstein, 2013]] ), ocean temperatures and salinity ( [[#Banks--2002|Banks and Wood, 2002]] ), mean precipitation ( [[#Giorgi--2009|Giorgi and Bi, 2009]] ; [[#Maraun--2013|Maraun, 2013]] ), drought ( [[#Orlowsky--2013|Orlowsky and Seneviratne, 2013]] ), extremes ( [[#Diffenbaugh--2011|Diffenbaugh and Scherer, 2011]] ; [[#Fischer--2014|Fischer et al., 2014]] ; [[#King--2015|King et al., 2015]] ; [[#Schleussner--2020|Schleussner and Fyson, 2020]] ), and regional sea level change ( [[#Lyu--2014|Lyu et al., 2014]] ). The concept has also been applied to climate change impacts such as effects on crop growing regions ( [[#Rojas--2019|Rojas et al., 2019]] ). In AR6, the emergence of oceanic signals such as regional sea level change and changes in water mass properties is assessed in [[IPCC:Wg1:Chapter:Chapter-9|Chapter 9]] (Section 9.6.1.4); emergence of future regional changes is assessed in [https://www.ipcc.ch/report/ar6/wg1/chapter/chapter-10 Chapter 10] (Section 10.4.3); the emergence of extremes as a function of global warming levels is assessed in [https://www.ipcc.ch/report/ar6/wg1/chapter/chapter-11 Chapter 11] (Section 11.2.5); and the emergence of climatic impact-drivers for AR6 regions and many climate variables is assessed in [https://www.ipcc.ch/report/ar6/wg1/chapter/chapter-12 Chapter 12] (Section 12.5.2).

Although the magnitude of any change is important, regions which have a larger signal of change relative to the background variations will potentially face greater risks than other regions, as they will see unusual or novel climate conditions more quickly ( [[#Frame--2017|Frame et al., 2017]] ). As in Figure 1.14, the signal of temperature change is often smaller in tropical countries, but their lower amplitude of variability means they may experience the effects of climate change earlier than the mid-latitudes. In addition, these tropical countries are often among the most exposed, due to large populations ( [[#Lehner--2015|Lehner and Stocker, 2015]] ), and often more vulnerable ( [[#Harrington--2016|Harrington et al., 2016]] ; [[#Harrington--2018|Harrington and Otto, 2018]] ; [[#Russo--2019|Russo et al., 2019]] ). Higher levels of exposure and vulnerability increase the risk from climate-related impacts (Cross-Chapter Box 1.3). The rate of change is also important for many hazards (e.g., [[#Loarie--2009|Loarie et al., 2009]] ). Providing more information about changes and variations on regional scales, and the associated attribution to particular causes (Cross-Working Group Box: Attribution), is therefore important for adaptation planning.

<div id="1.4.3" class="h2-container"></div>

<span id="sources-of-uncertainty-in-climate-simulations"></span>
=== 1.4.3 Sources of Uncertainty in Climate Simulations ===

<div id="h2-22-siblings" class="h2-siblings"></div>

When evaluating and analysing simulations of the physical climate system, several different sources of uncertainty need to be considered (e.g., [[#Hawkins--2009|Hawkins and Sutton, 2009]] ; [[#Lehner--2020|Lehner et al., 2020]] ). Broadly, these sources are: uncertainties in radiative forcings (both those observed in the past and those projected for the future); uncertainty in the climate response to particular radiative forcings; internal and natural variations of the climate system (which may be somewhat predictable); and interactions among these sources of uncertainty.

Ensembles of climate simulations ( [[#1.5.4.2|Section 1.5.4.2]] ), such as those produced as part of the sixth phase of the Coupled Model Intercomparison Project (CMIP6), can be used to explore these different sources of uncertainty and estimate their magnitude. Relevant experiments with climate models include both historical simulations constrained by past radiative forcings, and projections of future climate which are constrained by specified drivers, such as GHG concentrations, emissions, or radiative forcings. (The term ‘prediction’ is usually reserved for estimates of the future climate state which are also constrained by the observed initial conditions of the climate system, analogous to a weather forecast.)

<div id="1.4.3.1" class="h3-container"></div>

<span id="sources-of-uncertainty"></span>
==== 1.4.3.1 Sources of Uncertainty ====

<div id="h3-15-siblings" class="h3-siblings"></div>

<div id="1.4.3.1.1" class="h4-container"></div>

<span id="radiative-forc-ing-uncertainty"></span>
===== ''1.4.3.1.1 Radiative forcing uncertainty'' =====

<div id="h4-1-siblings" class="h4-siblings"></div>

Future radiative forcing is uncertain due to as-yet-unknown societal choices that will determine future anthropogenic emissions; this is considered ‘scenario uncertainty’. The RCP and SSP scenarios, which form the basis for climate projections assessed in this Report, are designed to span a plausible range of future pathways ( [[#1.6|Section 1.6]] ) and can be used to estimate the magnitude of scenario uncertainty, but the real world may also differ from any one of these example pathways.

Uncertainties also exist regarding past emissions and radiative forcings. These are especially important for simulations of paleoclimate time periods, such as the Pliocene, Last Glacial Maximum or the last millennium, but are also relevant for the CMIP historical simulations of the instrumental period since 1850. In particular, historical radiative forcings due to anthropogenic and natural aerosols are less well constrained by observations than the GHG radiative forcings. There is also uncertainty in the size of large volcanic eruptions (and in the location for some that occurred before around 1850), and the amplitude of changes in solar activity, before satellite observations. The role of historical radiative forcing uncertainty was considered previously ( [[#Knutti--2002|Knutti et al., 2002]] ; [[#Forster--2013|Forster et al., 2013]] ) but, since AR5, specific simulations have been performed to examine this issue, particularly for the effects of uncertainty in anthropogenic aerosol radiative forcing (e.g., [[#Jiménez-de-la-Cuesta--2019|Jiménez-de-la-Cuesta and Mauritsen, 2019]] ; [[#Dittus--2020|Dittus et al., 2020]] ).

<div id="1.4.3.1.2" class="h4-container"></div>

<span id="climate-respo-nse-uncertainty"></span>

===== ''1.4.3.1.2 Climate respo'' ''nse uncertainty'' =====

<div id="h4-2-siblings" class="h4-siblings"></div>

Under any particular scenario ( [[#1.6.1|Section 1.6.1]] ), there is uncertainty in how the climate will respond to the specified emissions or radiative forcing combinations. A range of climate models is often used to estimate the range of uncertainty in our understanding of the key physical processes and to define the ‘model response uncertainty’ (Sections [[#1.5.4|1.5.4]] and [[IPCC:Wg1:Chapter:Chapter-4#4.2.5|4.2.5]] ). However, this range does not necessarily represent the full ‘climate response uncertainty ''’'' in how the climate may respond to a particular radiative forcing or emissions scenario. This is because, for example, the climate models used in CMIP experiments have structural uncertainties not explored in a typical multi-model exercise (e.g., [[#Murphy--2004|Murphy et al., 2004]] ) and are not entirely independent of each other ( [[#1.5.4.8|Section 1.5.4.8]] ; [[#Masson--2011|Masson and Knutti, 2011]] ; [[#Abramowitz--2019|Abramowitz et al., 2019]] ); there are small spatial-scale features which cannot be resolved; and long time-scale processes or tipping points are not fully represented. [[#1.4.4|Section 1.4.4]] discusses how some of these issues can still be considered in a risk assessment context. For some metrics, such as equilibrium climate sensitivity (ECS), the CMIP6 model range is found to be broader than the ''very likely'' range assessed by combining multiple lines of evidence (Sections 4.3.4 and 7.5.6).

<div id="1.4.3.1.3" class="h4-container"></div>

<span id="natural-and-internal-cli-mate-variations"></span>
===== ''1.4.3.1.3 Natural and internal cli'' ''mate variations'' =====

<div id="h4-3-siblings" class="h4-siblings"></div>

Even without any anthropogenic radiative forcing, there would still be uncertainty in projecting future climate because of unpredictable natural factors such as variations in solar activity and volcanic eruptions. For projections of future climate, such as those presented in Chapter 4, the uncertainty in these factors is not normally considered. However, the potential effects on the climate of large volcanic eruptions (Cross-Chapter Box 4.1; [[#Zanchettin--2016|Zanchettin et al., 2016]] ; [[#Bethke--2017|Bethke et al., 2017]] ) and large solar variations ( [[#Feulner--2010|Feulner and Rahmstorf, 2010]] ; [[#Maycock--2015|Maycock et al., 2015]] ) are studied. On longer time scales, orbital effects and plate tectonics also play a role.

Further, even in the absence of any anthropogenic or natural changes in radiative forcing, Earth’s climate fluctuates on time scales from days to decades or longer. These ‘internal’ variations, such as those associated with modes of variability (e.g., ENSO, Pacific Decadal Variability (PDV), or Atlantic Multi-decadal Variability (AMV); Annex IV) are unpredictable on time scales longer than a few years ahead and are a source of uncertainty for understanding how the climate might become in a particular decade, especially regionally. The increased use of ‘large ensembles’ of complex climate model simulations to sample this component of uncertainty is discussed above in [[#1.4.2.1|Section 1.4.2.1]] and further in Chapter 4.

<div id="1.4.3.1.4" class="h4-container"></div>

<span id="interactions-between-variability-and-rad-iative-forcings"></span>
===== ''1.4.3.1.4 Interactions between variability and rad'' ''iative forcings'' =====

<div id="h4-4-siblings" class="h4-siblings"></div>

It is plausible that there are interactions between radiative forcings and climate variations, such as influences on the phasing or amplitude of internal or natural climate variability ( [[#Zanchettin--2017|Zanchettin, 2017]] ). For example, the timing of volcanic eruptions may influence Atlantic Multi-decadal Variability (e.g., [[#Otterå--2010|Otterå et al., 2010]] ; [[#Birkel--2018|Birkel et al., 2018]] ) or ENSO (e.g., [[#Maher--2015|Maher et al., 2015]] ; [[#Khodri--2017|Khodri et al., 2017]] ; [[#Zuo--2018|Zuo et al., 2018]] ), and anthropogenic aerosols may influence decadal modes of variability in the Pacific (e.g., [[#Smith--2016|Smith et al., 2016]] ). In addition, melting of glaciers and ice caps due to anthropogenic influences has been speculated to increase volcanic activity (e.g., a specific example for Iceland is discussed in [[#Swindles--2018|Swindles et al., 2018]] ).

<div id="1.4.3.2" class="h3-container"></div>

<span id="uncertainty-quantification"></span>
==== 1.4.3.2 Uncertainty Quantification ====

<div id="h3-16-siblings" class="h3-siblings"></div>

Not all of these listed sources of uncertainty are of the same type. For example, internal climate variations are an intrinsic uncertainty that can be estimated probabilistically, and could be more precisely quantified, but cannot usually be reduced. However, advances in decadal prediction offer the prospect of narrowing uncertainties in the trajectory of the climate for a few years ahead ( [[IPCC:Wg1:Chapter:Chapter-4#4.2.3|Section 4.2.3]] ; e.g., [[#Meehl--2014|Meehl et al., 2014]] ; [[#Yeager--2017|Yeager and Robson, 2017]] ).

Other sources of uncertainty, such as model response uncertainty, can in principle be reduced, but are not amenable to a frequency-based interpretation of probability, and Bayesian methods to quantify the uncertainty have been considered instead (e.g., [[#Tebaldi--2004|Tebaldi, 2004]] ; [[#Rougier--2007|Rougier, 2007]] ; [[#Sexton--2012|Sexton et al., 2012]] ). The scenario uncertainty component is distinct from other uncertainties, given that future anthropogenic emissions can be considered as the outcome of a set of societal choices ( [[#1.6.1|Section 1.6.1]] ).

For climate model projections it is possible to approximately quantify the relative amplitude of various sources of uncertainty (e.g., [[#Hawkins--2009|Hawkins and Sutton, 2009]] ; [[#Lehner--2020|Lehner et al., 2020]] ). A range of different climate models are used to estimate the model response uncertainty to a particular emissions pathway, and multiple pathways are used to estimate the scenario uncertainty. The unforced component of internal variability can be estimated from individual ensemble members of the same climate model ( [[#1.5.4.8|Section 1.5.4.8]] ; e.g., [[#Deser--2012|Deser et al., 2012]] ; [[#Maher--2019|Maher et al., 2019]] ).

Figure 1.15 illustrates the relative size of these different uncertainty components using a ‘cascade of uncertainty’ ( [[#Wilby--2010|Wilby and Dessai, 2010]] ), with examples shown for global mean temperature, Northern South American annual temperatures and East Asian summer precipitation changes. For global mean temperature, the role of internal variability is small, and the total uncertainty is dominated by emissions scenario and model response uncertainties. Note that there is considerable overlap between individual simulations for different emissions scenarios, even for the mid-term (2041–2060). For example, the slowest-warming simulation for SSP5-8.5 produces less mid-term warming than the fastest-warming simulation for SSP1-1.9. For the long term, emissions scenario uncertainty becomes dominant.

<div id="_idContainer047" class="_idGenObjectStyleOverride-1"></div>

<!-- START IMG -->
<!-- IMG FILE -->
[[File:5e6738df5d3cb730b505b1733656e44f IPCC_AR6_WGI_Figure_1_15.png]]
<!-- IMG TITLE + CAPTION -->

'''Figure 1.15 |''' '''The ‘cascade of uncertainties’ in CMIP6 projections.''' Changes in: GSAT '''(left)''' ; Northern South America temperature '''(middle)''' ; and East Asia summer (June–July–August, JJA) precipitation '''(right)''' . These are shown for two time periods: 2041–2060 '''(top)''' and 2081–2100 '''(bottom)''' . The SSP–radiative forcing combination is indicated at the top of each cascade at the value of the multi-model mean for each scenario. This branches downwards to show the ensemble mean for each model, and further branches into the individual ensemble members, although often only a single member is available. These diagrams highlight the relative importance of different sources of uncertainty in climate projections, which varies for different time periods, regions and climate variables. See ( [[#1.4.5|Section 1.4.5]] for the definition of the regions used. Further details on data sources and processing are available in the chapter data table (Table 1.SM.1).
<!-- END IMG -->

Therelative uncertainty due to internal variability and model uncertainty increases for smaller spatial scales. In the regional example shown in Figure 1.15 for changes in temperature, the same scenario and model combination has produced two simulations which differ by 1°C in their projected 2081–2100 averages due solely to internal climate variability. For regional precipitation changes, emissions scenario uncertainty is often small relative to model response uncertainty. In the example shown in Figure 1.15, the SSPs overlap considerably, but SSP1-1.9 shows the largest precipitation change in the near term, even though global mean temperature warms the least; this is due to differences between regional aerosol emissions projected in this and other scenarios ( [[#Wilcox--2020|Wilcox et al., 2020]] ). These cascades of uncertainty would branch out further if applying the projections to derive estimates of changes in hazard (e.g., [[#Wilby--2010|Wilby and Dessai, 2010]] ; [[#Halsnæs--2018|Halsnæs and Kaspersen, 2018]] ; [[#Hattermann--2018|Hattermann et al., 2018]] ).

<div id="1.4.4" class="h2-container"></div>

<span id="considering-an-uncertain-future"></span>
=== 1.4.4 Considering an Uncertain Future ===

<div id="h2-23-siblings" class="h2-siblings"></div>

Since AR5 there have been developments in how to consider and describe future climate outcomes which are considered possible but ''very unlikely ,'' highly uncertain, or potentially surprising. To examine such futures there is a need to move beyond the usual ''likely'' or ''very likely'' assessed ranges and consider low-likelihood outcomes, especially those that would result in significant impacts if they occurred (e.g., [[#Sutton--2018|Sutton, 2018]] ; [[#Sillmann--2021|Sillmann et al., 2021]] ). This section briefly outlines some of the different approaches used in the AR6 WGI.

<div id="1.4.4.1" class="h3-container"></div>

<span id="low-likelihood-outcomes"></span>
==== 1.4.4.1 Low-Likelihood Outcomes ====

<div id="h3-17-siblings" class="h3-siblings"></div>

In the AR6, certain low-likelihood outcomes are described and assessed because they may be associated with high levels of risk, and the greatest risks may not be associated with the most likely outcome. The aim of assessing these possible futures is to better inform risk assessment and decision-making. Two types are considered: (i) low-likelihood high-warming (LLHW) scenarios, which describe the climate in a world with very high climate sensitivity; and (ii) low-likelihood, high-impact outcomes that have a low likelihood of occurring, but would cause large potential impacts on societies or ecosystems.

An illustrative example of how low-likelihood outcomes can produce significant additional risks is shown in Figure 1.16. The Reasons for Concern (RFCs) produced by the IPCC AR5 WGII define the additional risks due to climate change at different global warming levels. These have been combined with [[IPCC:Wg1:Chapter:Chapter-4|Chapter 4]] assessments of projected global temperature for different emissions scenarios (SSPs; [[#1.6|Section 1.6]] ), and [[IPCC:Wg1:Chapter:Chapter-7|Chapter 7]] assessments about ECS. For example, even following an intermediate emissions scenario could result in high levels of additional risk if ECS is at the upper end of the ''very likely'' range. However, not all possible low-likelihood outcomes relate to ECS, and AR6 considers these issues in more detail than previous IPCC assessment reports (see Table 1.1 and [[#1.4.4.2|Section 1.4.4.2]] for some examples).

<div id="_idContainer049" class="_idGenObjectStyleOverride-1"></div>

<!-- START IMG -->
<!-- IMG FILE -->
[[File:35397456082e6f0e68f69d968a9044f0 IPCC_AR6_WGI_Figure_1_16.png]]
<!-- IMG TITLE + CAPTION -->

'''Figure 1.16 |''' '''Illustrating concepts of low-likelihood outcomes.''' '''Left:''' schematic likelihood distribution consistent with the IPCC AR6 assessments that equilibrium climate sensitivity (ECS) is ''likely'' in the range 2.5°C to 4.0°C, and ''very likely'' between 2.0°C and 5.0°C (Chapter 7). ECS values outside the assessed ''very likely'' range are designated low-likelihood outcomes in this example (light grey). '''Middle''' and '''right-hand columns:''' additional risks due to climate change for 2020–2090 using the Reasons For Concern (RFCs, see [[#IPCC--2014b|IPCC, 2014b]] ), specifically RFC1 describing the risks to unique and threatened systems and RFC3 describing risks from the distribution of impacts ( [[#O’Neill--2017b|O’Neill et al., 2017b]] ; [[#Zommers--2020|Zommers et al., 2020]] ). The projected changes of GSAT used are the 95%, median and 5% assessed ranges from [[IPCC:Wg1:Chapter:Chapter-4|Chapter 4]] for each SSP (top, middle and bottom); these are designated High ECS, Mid-range ECS and Low ECS respectively. The ‘burning-ember’ risk spectrum of graduated colours is usually associated with levels of committed GSAT change; instead, this illustration associates the risk spectrum with the GSAT temperature reached in each year from 2020 to 2090. Note that this illustration does not include the vulnerability aspect of each SSP scenario. Further details on data sources and processing are available in the chapter data table (Table 1.SM.1).
<!-- END IMG -->

<div id="1.4.4.2" class="h3-container"></div>

<span id="storylines"></span>
==== 1.4.4.2 Storylines ====

<div id="h3-18-siblings" class="h3-siblings"></div>

As societies are increasingly experiencing the impacts of climate change-related events, the climate science community is developing climate information tailored for particular regions and sectors. There is a growing focus on explaining and exploring complex physical chains of events or on predicting climate under various future socio-economic developments. Since AR5, ‘storylines’ or ‘narratives’ approaches have been used to better inform risk assessment and decision-making, to assist understanding of regional processes, and represent and communicate climate projection uncertainties more clearly. The aim is to help build a cohesive overall picture of potential climate change pathways that moves beyond the presentation of data and figures (Glossary; Fløttum and Gjerstad, 2017; [[#Moezzi--2017|Moezzi et al., 2017]] ; [[#Dessai--2018|Dessai et al., 2018]] ; T.G. [[#Shepherd--2018|]] [[#Shepherd--2018|Shepherd et al., 2018]] ).

In the broader IPCC context, the term ‘scenario storyline’ refers to a narrative description of one or more scenarios, highlighting their main characteristics, relationships between key driving forces and the dynamics of their evolution (e.g., emissions of short-lived climate forcers assessed in [[IPCC:Wg1:Chapter:Chapter-6|Chapter 6]] are driven by ‘scenario storylines’; see [[#1.6|Section 1.6]] ). The AR6 WGI is mainly concerned with ‘physical climate storylines’. A physical climate storyline is a self-consistent and plausible physical trajectory of the climate system, or a weather or climate event, on time scales from hours to multiple decades (T.G. [[#Shepherd--2018|]] [[#Shepherd--2018|Shepherd et al., 2018]] ). This approach can be used to constrain projected changes or specific events on specified explanatory elements such as projected changes of large-scale indicators (Box 10.2). For example, [[#Hazeleger--2015|Hazeleger et al. (2015)]] suggested using ‘tales of future weather’, blending numerical weather prediction with a climate projection to illustrate the potential behaviour of future high-impact events (also see [[#Hegdahl--2020|Hegdahl et al., 2020]] ). Several studies describe how possible large changes in atmospheric circulation would affect regional precipitation and other climate variables, and discuss the various climate drivers that could cause such a circulation response ( [[#James--2015|James et al., 2015]] ; [[#Zappa--2017|Zappa and Shepherd, 2017]] ; [[#Mindlin--2020|Mindlin et al., 2020]] ). Physical climate storylines can also help frame the causal factors of extreme weather events ( [[#Shepherd--2016|Shepherd, 2016]] ) and then be linked to event attribution (Section 11.2.2 and Cross-Working Group Box: Attribution).

Storyline approaches can be used to communicate and contextualize climate change information in the context of risk for policymakers and practitioners (Box 10.2; e.g., [[#de%20Bruijn--2016|de Bruijn et al., 2016]] ; [[#Dessai--2018|Dessai et al., 2018]] ; [[#Scott--2018|Scott et al., 2018]] ; [[#Jack--2020|Jack et al., 2020]] ). They can also help in assessing risks associated with LLHI events ( [[#Weitzman--2011|Weitzman, 2011]] ; [[#Sutton--2018|Sutton, 2018]] ), because they consider the ‘physically self-consistent unfolding of past events, or of plausible future events or pathways’ ( [[#Shepherd--2018|]] [[#Shepherd--2018|Shepherd et al., 2018]] ), which would be masked in a probabilistic approach. These aspects are important as the greatest risk need not be associated with the highest-likelihood outcome, and in fact will often be associated with low-likelihood outcomes. The storyline approach can also acknowledge that climate-relevant decisions in a risk-oriented framing will rarely be taken on the basis of physical climate change alone; instead, such decisions will normally take into account socio-economic factors as well ( [[#Shepherd--2019|Shepherd, 2019]] ).

In the AR6 WGIAssessment Report, these different storyline approaches are used in several places (see Table 1.1). [[IPCC:Wg1:Chapter:Chapter-4|Chapter 4]] uses a storyline approach to assess the upper tail of the distribution of global warming levels (the storylines of high global warming levels) and their manifestation in global patterns of temperature and precipitation changes. [[IPCC:Wg1:Chapter:Chapter-9|Chapter 9]] uses a storyline approach to examine the potential for, and early warning signals of, a high-end sea level scenario, in the context of deep uncertainty related to our current understanding of the physical processes that contribute to long-term sea level rise. [https://www.ipcc.ch/report/ar6/wg1/chapter/chapter-10 Chapter 10] assesses the use of physical climate storylines and narratives as a way to explore uncertainties in regional climate projections, and to link to the specific risk and decision context relevant to a user, for developing integrated and context-relevant regional climate change information. [https://www.ipcc.ch/report/ar6/wg1/chapter/chapter-11 Chapter 11] uses the term storyline in the framework of extreme event attribution. [https://www.ipcc.ch/report/ar6/wg1/chapter/chapter-12 Chapter 12] assesses the use of a storylines approach with narrative elements for communicating climate (change) information in the context of climate services (Cross-Chapter Box 12.2).

<div id="1.4.4.3" class="h3-container"></div>

<span id="abrupt-change-tipping-points-and-surprises"></span>
==== 1.4.4.3 Abrupt Change, Tipping Points and Surprises ====

<div id="h3-19-siblings" class="h3-siblings"></div>

An ‘abrupt change’ is defined in this report as a change that takes place substantially faster than the rate of change in the recent history of the affected component of a system (Glossary). In some cases, abrupt change occurs because the system state actually becomes unstable, such that the subsequent rate of change is independent of the forcing. We refer to this class of abrupt change as a ‘tipping point’ '','' defined as a critical threshold beyond which a system reorganizes, often abruptly and/or irreversibly (Glossary; [[#Lenton--2008|Lenton et al., 2008]] ). Some of the abrupt climate changes and climate tipping points discussed in this Report could have severe local climate responses, such as extreme temperature, droughts, forest fires, ice-sheet loss and collapse of the thermohaline circulation (Sections 4.7.2, 5.4.9, 8.6 and 9.2.3).

There is evidence of abrupt changes in Earth’s history, and some of these events have been interpreted as tipping points ( [[#Dakos--2008|Dakos et al., 2008]] ). Some of these are associated with significant changes in the global climate, such as deglaciations in the Quaternary (past 2.5 million years) and rapid warming at the Palaeocene–Eocene Thermal Maximum (around 55.5 million years ago; [[#Bowen--2015|Bowen et al., 2015]] ; [[#Hollis--2019|Hollis et al., 2019]] ). Such events changed the planetary climate for tens to hundreds of thousands of years, but at a rate that is actually much slower than projected anthropogenic climate change over this century, even in the absence of tipping points.

Such paleoclimate evidence has even fuelled concerns that anthropogenic GHGs could tip the global climate into a permanent hot state ( [[#Steffen--2018|Steffen et al., 2018]] ). However, there is no evidence of such non-linear responses at the global scale in climate projections for the next century, which indicates a near-linear dependence of global temperature on cumulative GHG emissions (Sections 1.3.5, 5.5 and 7.4.3.1). At the regional scale, abrupt changes and tipping points, such as Amazon rainforest dieback and permafrost collapse, have occurred in projections with Earth System Models ( [[IPCC:Wg1:Chapter:Chapter-4#4.7.3|Section 4.7.3]] ; [[#Drijfhout--2015|Drijfhout et al., 2015]] ; [[#Bathiany--2020|Bathiany et al., 2020]] ). In such simulations, tipping points occur in narrow regions of parameter space (e.g., CO <sub>2</sub> concentration or temperature increase), and for specific climate background states. This makes them difficult to predict using Earth system models (ESMs) relying on parmeterizations of known processes. In some cases, it is possible to detect forthcoming tipping points through time-series analysis that identifies increased sensitivity to perturbations as the tipping point is approached (e.g., ‘critical slowing-down’, [[#Scheffer--2012|Scheffer et al., 2012]] ).

Some suggested climate tipping points prompt transitions from one steady state to another (Figure 1.17). Transitions can be prompted by perturbations such as climate extremes which force the system outside of its current well of attraction in the stability landscape; this is called noise-induced tipping (Figure 1.17a,b; [[#Ashwin--2012|Ashwin et al., 2012]] ). For example, the tropical forest dieback seen in some ESM projections is accelerated by longer and more frequent droughts over tropical land ( [[#Good--2013|Good et al., 2013]] ).

<div id="_idContainer051" class="_idGenObjectStyleOverride-1"></div>

<!-- START IMG -->
<!-- IMG FILE -->
[[File:23aa7d4ef86b70d128c44c971f6234a3 IPCC_AR6_WGI_Figure_1_17.png]]
<!-- IMG TITLE + CAPTION -->

'''Figure 1.17 |''' '''Illustration of two types of tipping points: noise-induced (a, b) and bifurcation (c, d).''' '''(a)''' and '''(c)''' are example time-series (coloured lines) through the tipping point, with solid-black lines indicating stable climate states (e.g., low or high rainfall) and dashed lines representing the boundary between stable states. '''(b)''' and '''(d)''' are stability landscapes, which provide an intuitive understanding of the different types of tipping point. The ‘valleys’ represent different climate states the system can occupy, with ‘hilltops’ separating the stable states. The resilience of a climate state is implied by the depth of the valley. The current state of the system is represented by a ball. Both scenarios assume that the ball starts in the left-hand valley (dashed-black lines) and then through different mechanisms dependent on the type of tipping transitions to the right-hand valley (coloured lines). Noise-induced tipping events (a, b), for instance drought events causing sudden dieback of the Amazon rainforest, develop from fluctuations within the system. The stability landscape in this scenario remains fixed and stationary. A series of perturbations in the same direction, or one large perturbation, are required to force the system over the hilltop and into the alternative stable state. Bifurcation tipping events (c, d), such as a collapse of the thermohaline circulation in the Atlantic Ocean under climate change, occur when a critical level in the forcing is reached. Here the stability landscape is subjected to a change in shape. Under gradual anthropogenic forcing the left-hand valley begins to shallow and eventually vanishes at the tipping point, forcing the system to transition to the right-hand valley.
<!-- END IMG -->

Alternatively, transitions from one state to another can occur if a critical threshold is exceeded; this is called ‘bifurcation tipping’ (Figure 1.17c,d; [[#Ashwin--2012|Ashwin et al., 2012]] ). The new state is defined as ‘irreversible’ on a given time scale if the recovery from this state takes substantially longer than the time scale of interest, which is decades to centuries for the projections presented in this report. A well-known example is the modelled irreversibility of the ocean’s thermohaline circulation in response to North Atlantic changes such as freshwater input from rainfall and ice-sheet melt ( [[#Rahmstorf--2005|Rahmstorf et al., 2005]] ; [[#Alkhayuon--2019|Alkhayuon et al., 2019]] ), which is assessed in detail in [[IPCC:Wg1:Chapter:Chapter-9|Chapter 9]] (Section 9.2.3).

The tipping point concept is most commonly framed for systems in which the forcing changes relatively slowly. However, this is not the case for most scenarios of anthropogenic forcing projected for the 21st century. Systems with inertia lag behind rapidly increasing forcing, which can lead to the failure of early warning signals or even the possibility of temporarily overshooting a bifurcation point without provoking tipping ( [[#Ritchie--2019|Ritchie et al., 2019]] ).

‘Surprises’ are a class of risk that can be defined as low-likelihood but well-understood events: they are events that cannot be predicted with current understanding. The risk from such surprises can be accounted for in risk assessments ( [[#Parker--2015|Parker and Risbey, 2015]] ). Examples relevant to climate science include: a series of major volcanic eruptions or a nuclear war, either of which would cause substantial planetary cooling ( [[#Robock--2007|Robock et al., 2007]] ; [[#Mills--2014|Mills et al., 2014]] ); significant 21st century sea level rise due to marine ice sheet instability (MISI; Box 9.4); the potential for collapse of the stratocumulus cloud decks ( [[#Schneider--2019|Schneider et al., 2019]] ) or other substantial changes in climate feedbacks (Section 7.4); and unexpected biological epidemics among humans or other species, such as the COVID-19 pandemic (Cross-Chapter Box 6.1; [[#Forster--2020|Forster et al., 2020]] ; [[#Le%20Quéré--2020|Le Quéré et al., 2020]] ). The discovery of the hole in the ozone layerwas also a surprise even though some of the relevant atmospheric chemistry was known at the time. The term ‘unknownunknowns’ ( [[#Parker--2015|Parker and Risbey, 2015]] ) is also sometimes used in this context to refer to events that cannot be anticipated with presentknowledge or were of an unanticipated nature before they occurred.

<div id="cross-chapter-box-1.3" class="h2-container box-container"></div>

'''Cross-Chapter Box 1.3 | Risk Fram''' '''ing in IPCC AR6'''

<div id="h2-24-siblings" class="h2-siblings"></div>

'''Contributing Authors:''' Andy Reisinger (New Zealand), Maisa Rojas (Chile), Aïda Diongue-Niang (Senegal), Maarten K. van Aalst (The Netherlands), Mathias Garschagen (Germany), Mark Howden (Australia), Margot Hurlbert (Canada), Katharine Mach (United States of America), Sawsan Khair Elsied Abdel Rahim Mustafa (Sudan), Brian O’Neill (United States of America), Roque Pedace (Argentina), Jana Sillmann (Norway/Germany), Carolina Vera (Argentina), David Viner (United Kingdom)

The IPCC Special Report on Managing the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation (SREX; [[#IPCC--2012|IPCC, 2012]] ) presented a framework for assessing risks from climate change, which linked hazards (due to changes in climate) with exposure and vulnerability ( [[#Cardona--2012|Cardona et al., 2012]] ). This framework was further developed by AR5 WGII ( [[#IPCC--2014b|IPCC, 2014b]] ), while AR5 WGI focussed only on the hazard component of risk. As part of AR6, a cross-Working Group process expanded and refined the concept of risk to allow for a consistent risk framing to be used across the three IPCC Working Groups ( [[#IPCC--2019b|IPCC, 2019b]] ; Box 2 in [[#Abram--2019|Abram et al., 2019]] ; [[#Reisinger--2020|Reisinger et al., 2020]] ).

In this revised definition, risk is defined as:

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. Relevant adverse consequences include those on lives, livelihoods, health and well-being, economic, social and cultural assets and investments, infrastructure, services (including ecosystem services), ecosystems and species.

In the context of climate change impacts, risks result from dynamic interactions between climate-related hazards with the exposure and vulnerability of the affected human or ecological system to the hazards. Hazards, exposure and vulnerability may each be subject to uncertainty in terms of magnitude and likelihood of occurrence, and each may change over time and space due to socio-economic changes and human decision-making (see also risk management, adaptation and mitigation).

In the context of climate change responses, risks result from the potential for such responses not achieving the intended objective(s), or from potential trade-offs with, or negative side-effects on, other societal objectives, such as the Sustainable Development Goals (SDGs) (see also risk trade-off). Risks can arise, for example, from uncertainty in implementation, effectiveness or outcomes of climate policy, climate-related investments, technology development or adoption, and system transitions.

Cross-Chapter Box 1.3

The following concepts are also relevant for the definition of risk (Glossary):

'''Exposure:''' The presence of people; livelihoods; species or ecosystems; environmental functions, services, and resources; infrastructure; or economic, social, or cultural assets in places and settings that could be adversely affected.

'''Vulnerability:''' The propensity or predisposition to be adversely affected. Vulnerability encompasses a variety of concepts and elements including sensitivity or susceptibility to harm and lack of capacity to cope and adapt.

'''Hazard:''' The potential occurrence of a natural or human-induced physical event or trend that may cause loss of life, injury, or other health impacts, as well as damage and loss to property, infrastructure, livelihoods, service provision, ecosystems and environmental resources.

'''Impacts:''' The consequences of realized risks on natural and human systems, where risks result from the interactions of climate-related hazards (including extreme weather/climate events), exposure, and vulnerability. Impacts generally refer to effects on lives, livelihoods, health and well-being, ecosystems and species, economic, social and cultural assets, services (including ecosystem services), and infrastructure. Impacts may be referred to as consequences or outcomes and can be adverse or beneficial.

''''''Risk in AR6 WGI''''''
The revised risk framing clarifies the role and contribution of WGI to risk assessment. ‘Risk’ in IPCC terminology applies only to human or ecological systems, not to physical systems on their own.

'''Climatic impact-drivers (CIDs):''' 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.

InAR6, WGI uses the term ‘climatic impact-drivers’ to describe changes in physical systems rather than ‘hazards’, because the term hazard already assumes an adverse consequence. The terminology of ‘climatic impact-driver’ therefore allows WGI to provide a more value-neutral characterization of climatic changes that may be relevant for understanding potential impacts, without pre-judging whether specific climatic changes necessarily lead to adverse consequences, as some could also result in beneficial outcomes depending on the specific system and associated values. [https://www.ipcc.ch/report/ar6/wg1/chapter/chapter-12 Chapter 12] and the [[IPCC:Wg1:Chapter:Atlas|Atlas]] assess and provide information on climatic impact-drivers for different regions and sectors to support and link to the WGII assessment of the impacts and risks (or opportunities) related to the changes in the climatic impact-drivers. Although CIDs can lead to adverse or beneficial outcomes, focus is given to CIDs connected to hazards, and hence inform risk.

‘Extremes’ are a category of CID, corresponding to unusual events with respect to the range of observed values of the variable. [https://www.ipcc.ch/report/ar6/wg1/chapter/chapter-11 Chapter 11] assesses changes in weather and climate extremes, their attribution and future projections.

As examples of the use of this terminology, the term ‘flood risk’ should not be used if it only describes changes in the frequency and intensity of flood events (a hazard); the risk from flooding to human and ecological systems is caused by the flood hazard, the exposure of the system affected (e.g., topography, human settlements or infrastructure in the area potentially affected by flooding) and the vulnerability of the system (e.g., design and maintenance of infrastructure, existence of early warning systems). As another example, climate-related risk to food security can arise from both potential climate change impacts and responses to climate change and can be exacerbated by other stressors. Drivers for risks related to climate change impacts include climatic impact- drivers (e.g., drought, temperature extremes, humidity) mediated by other climatic impact-drivers (e.g., increased CO <sub>2</sub> fertilization of certain types of crops may help increase yields), the potential for indirect climate-related impacts (e.g., pest outbreaks triggered by ecosystem responses to weather patterns), exposure of people (e.g., how many people depend on a particular crop) and vulnerability or adaptability (how able are affected people to substitute other sources of food, which may be related to financial access and markets).

Information provided by WGI may or may not be relevant to understand risks related to climate change responses. For example, the risk to a company arising from emissions pricing, or the societal risk from reliance on an unproven mitigation technology, is not directly dependent on actual or projected changes in climate but arise largely from human choices. However, WGI climate information may be relevant to understand the potential for maladaptation, such as the potential for specific adaptation responses not achieving the desired outcome or having negative side effects. For example, WGI information about the range of sea level rise can help inform understanding of whether coastal protection, accommodation, or retreat would be the most effective risk management strategy in a particular context.

Cross-Chapter Box 1.3

From a WGI perspective, low-likelihood, high-impact outcomes and the concept of deep uncertainty are also relevant for risk assessment.

'''Low-likelihood, high-impact (LLHI) 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.

The AR6 WGI Report provides more detailed information about these types of events compared to AR5 (Table 1.1, [[#1.4.4|Section 1.4.4]] ).

Recognizing the need for assessing and managing risk in situations of high uncertainty, SROCC advanced the treatment of situations with deep uncertainty ( [[#1.2.3|Section 1.2.3]] ; [[#IPCC--2019b|IPCC, 2019b]] ; Box 5 in [[#Abram--2019|Abram et al., 2019]] ). A situation of deep uncertainty exists when experts or stakeholders do not know or cannot agree on: (i) appropriate conceptual models that describe relationships among key driving forces in a system; (ii) the probability distributions used to represent uncertainty about key variables and parameters; and/or (iii) how to weigh and value desirable alternative outcomes ( [[#Abram--2019|Abram et al., 2019]] ). The concept of deep uncertainty can complement the IPCC calibrated uncertainty language and thereby broaden the communication of risk.

<div id="cross-working-group-box" class="h2-container box-container"></div>

'''Cross-Working Group B''' '''ox | Attribution'''

<div id="h2-25-siblings" class="h2-siblings"></div>

'''Contributing Authors:''' Pandora Hope (Australia), Wolfgang Cramer (France/Germany), Gregory M. Flato (Canada), Katja Frieler (Germany), Nathan P. Gillett (Canada), Christian Huggel (Switzerland), Jan Minx (Germany), Friederike Otto (United Kingdom/Germany), Camille Parmesan (France, United Kingdom/United States of America), Joeri Rogelj (United Kingdom/Belgium), Maisa Rojas (Chile), Sonia I. Seneviratne (Switzerland), Aimée B.A. Slangen (The Netherlands), Daithi Stone (New Zealand), Laurent Terray (France), Maarten K. van Aalst (The Netherlands), Robert Vautard (France), Xuebin Zhang (Canada)

''''''Introduction''''''
Changes in the climate system are becoming increasingly apparent, as are the climate-related impacts on natural and human systems. Attribution is the process of evaluating the contribution of one or more causal factors to such observed changes or events. Typical questions addressed by the IPCC include: ‘To what extent is an observed change in global temperature induced by anthropogenic GHG and aerosol concentration changes, or influenced by natural variability?’ and ‘What is the contribution of climate change to observed changes in crop yields, which are also influenced by changes in agricultural management?’ Changes in the occurrence and intensity of extreme events can also be attributed, addressing questions such as: ‘Have human GHG emissions increased the likelihood or intensity of an observed heatwave?’

This Cross-Working Group Box briefly describes why attribution studies are important. It also describes some new developments in the methods used in those studies and provides recommendations for interpretation.

Attribution studies serve to evaluate and communicate linkages associated with climate change, for example: between the human-induced increase in GHG concentrations and the observed increase in air temperature or extreme weather events (AR6 WGI Chapters 3, 10 and 11); or between observed changes in climate and changing species distributions and food production (AR6 WGII Chapters 2 and others, summarized in WGII Chapter 16; e.g., [[#Verschuur--2021|Verschuur et al., 2021]] ); or between climate change mitigation policies and atmospheric GHG concentrations (AR6 WGI Chapter 5; AR6 WGIII Chapter 14). As such, they support numerous statements made by the IPCC (AR6 WGI [[#1.3|Section 1.3]] and Appendix 1A; [[#IPCC--2013b|IPCC, 2013b]] , 2014b).

Attribution assessments can also serve to monitor mitigation and assess the efficacy of applied climate protection policies (AR6 WGI [[IPCC:Wg1:Chapter:Chapter-4#4.6.3|Section 4.6.3]] ; e.g., [[#Nauels--2019|Nauels et al., 2019]] ; [[#Banerjee--2020|Banerjee et al., 2020]] ), inform and constrain projections (WGI [[IPCC:Wg1:Chapter:Chapter-4#4.2.3|Section 4.2.3]] ; [[#Gillett--2021|Gillett et al., 2021]] ; [[#Ribes--2021|Ribes et al., 2021]] ) or inform the loss and damages estimates and potential climate litigation cases by estimating the costs of climate change ( [[#Huggel--2015|Huggel et al., 2015]] ; [[#Marjanac--2017|Marjanac et al., 2017]] ; [[#Frame--2020|Frame et al., 2020]] ). These findings can thus inform mitigation decisions as well as risk management and adaptation planning (e.g., [[#CDKN--2017|CDKN, 2017]] ).

''''''Steps towards an attribu''' '''tion assessment''''''
The unambiguous framing of what changes are being attributed to what causes is a crucial first step for an assessment ( [[#Easterling--2016|Easterling et al., 2016]] ; [[#Hansen--2016|Hansen et al., 2016]] ; [[#Stone--2021|Stone et al., 2021]] ), followed by the identification of the possible and plausible drivers of change and the development of a hypothesis or theory for the linkage (Cross-Working Group Box: Attribution, Figure 1). The next step is to clearly define the indicators of the observed change or event and note the quality of the observations. There has been significant progress in the compilation of fragmented and distributed observational data, broadening and deepening the data basis for attribution research (WGI [[#1.5|Section 1.5]] ; e.g., [[#Poloczanska--2013|Poloczanska et al., 2013]] ; [[#Ray--2015|Ray et al., 2015]] ; [[#Cohen--2018|Cohen et al., 2018]] ). The quality ofthe observational record of drivers should also be considered (e.g., volcanic eruptions: WGI [[IPCC:Wg1:Chapter:Chapter-2#2.2.2|Section 2.2.2]] ). Impacted systems also change in the absence of climate change; this baseline and its associated modifiers – such as agricultural developments or population growth – need to be considered, alongside the exposure and vulnerability of people depending on these systems.

<div id="_idContainer053" class="Basic-Text-Frame"></div>

<!-- START IMG -->
<!-- IMG FILE -->
[[File:638aa4fa277b50207bb63cce1961b263 IPCC_AR6_WGI_CCBOX_Attribution_Figure_1.png]]
<!-- IMG TITLE + CAPTION -->

'''Cross-Working Group Box: Attribution, Figure 1 |'''
<!-- END IMG -->

'''Schematic of the steps to develop an attribution assessment, and the purposes of such assessments. Methods and systems used to test the attribution hypothesis or theory include: model-based fingerprinting; other model-based methods; evidence-based fingerprinting; process-based approaches; empirical or decomposition methods; and the use of multiple lines of evidence.''' Many of the methods are based on the comparison of the observed state of a system to a hypothetical counterfactual world that does not include the driver of interest to help estimate the causes of the observed response.

There are many attribution approaches, and several methods are detailed below. In physical and biological systems, attribution often builds on the understanding of the mechanisms behind the observed changes and numerical models are used, while in human systems other methods of evidence-building are employed. Confidence in the attribution can be increased if more than one approach is used and the model is evaluated as fit-for-purpose (WGI [[#1.5|Section 1.5]] , WGI Section 3.8, WGI Section 10.3.3.4 ; Hegerl et al. , 2010; Vautard et al. , 2019; Otto et al. , 2020; Philip et al. , 2020) . The final step includes appropriate communication of the attribution assessment and the accompanying confidence in the result (e.g., [[#Lewis--2019|Lewis et al., 2019]] ).

''''''Attribution methods''''''
<span id="attribution-of-changes-in-atmospheric-greenhouse-gas-concentrations-to-anthropogenic-activity"></span>
=== Attribution of changes in atmospheric greenhouse gas concentrations to anthropogenic activity ===

The AR6 WGI [[IPCC:Wg1:Chapter:Chapter-5|Chapter 5]] presents multiple lines of evidence that unequivocally establish the dominant role of human activities in the growth of atmospheric CO <sub>2</sub> , including through analysing changes in atmospheric carbon isotope ratios and the atmospheric O <sub>2</sub> –N <sub>2</sub> ratio (WGI Section 5.2.1.1). Decomposition approaches can be used to attribute emissions underlying those changes to various drivers such as population, energy efficiency, consumption or carbon intensity ( [[#Hoekstra--2003|Hoekstra and van den Bergh, 2003]] ; [[#Raupach--2007|Raupach et al., 2007]] ; [[#Rosa--2012|Rosa and Dietz, 2012]] ). Combined with attribution of their climate outcomes, the attribution of the sources of GHG emissions can inform the attribution of anthropogenic climate change to specific countries or actors ( [[#Matthews--2016|Matthews, 2016]] ; [[#Otto--2017|Otto et al., 2017]] ; [[#Skeie--2017|Skeie et al., 2017]] ; [[#Nauels--2019|Nauels et al., 2019]] ), and in turn inform discussions on fairness and burden sharing (WGIII Chapter 14).

<span id="attribution-of-observed-climate-change-to-anthropogenic-forcing"></span>
=== Attribution of observed climate change to anthropogenic forcing ===

Changes in large-scale climate variables (e.g., global mean temperature) have been reliably attributed to anthropogenic and natural forcings (WGI [[#1.3.4|Section 1.3.4]] ; e.g., [[#Hegerl--2010|Hegerl et al., 2010]] ; [[#Bindoff--2013|Bindoff et al., 2013]] ). The most established method is to identify the ‘fingerprint’ of the expected space-time response to a particular climate forcing agent such as the concentration of anthropogenically induced GHGs or aerosols, or natural variation of solar radiation. This technique disentangles the contribution of individual forcing agents to an observed change (e.g., [[#Gillett--2021|Gillett et al., 2021]] ). New statistical approaches have been applied to better account for internal climate variability and the uncertainties in models and observations (WGI [[IPCC:Wg1:Chapter:Chapter-3#3.2|Section 3.2]] ; e.g., Naveau et al. , 2018; Santer et al. , 2019) . There are many other approaches, for example, global mean sea level change has been attributed to anthropogenic climate forcing by attributing the individual contributions from, for example, glacier melt or thermal expansion, while also examining which aspects of the observed change are inconsistent with internal variability (WGI Sections 3.5.2 and 9.6.1.4).

Specific regional conditions and responses may simplify or complicate attribution on those scales. For example, some human forcings, such as regional land-use change or aerosols, may enhance or reduce regional signals of change (WGI Sections 10.4.2, 11.1.6 and 11.2.2; Lejeune et al. , 2018; Undorf et al. , 2018; Boé et al. , 2020; Thiery et al. , 2020) . In general, regional climate variations are larger than the global mean climate, adding additional uncertainty to attribution (e.g., in regional sea level change, WGI Section 9.6.1). These statistical limitations may be reduced by ‘process-based attribution’, focusing on the physical processes known to influence the response to external forcing and internal variability (WGI Section 10.4.2).

<span id="attribution-of-weather-and-climate-events-to-anthropogenic-forcing"></span>
=== Attribution of weather and climate events to anthropogenic forcing ===

New methods have emerged since AR5 to attribute the change in likelihood or characteristics of weather or climate events or classes of events to underlying drivers (WGI Sections 10.4.1 and 11.2.2; [[#NA%20SEM--2016|NA SEM, 2016]] ; Stott et al. , 2016; Jézéquel et al. , 2018; Wehner et al. , 2018; Wang et al. , 2021) . Typically, historical changes, simulated under observed forcings, are compared to a counterfactual climate simulated in the absence of anthropogenic forcing. Another approach examines facets of the weather and thermodynamic status of an event through process-based attribution (WGI [https://www.ipcc.ch/report/ar6/wg1/chapter/chapter-11 Chapter 11] and Section 10.4.1; Hauser et al. , 2016; Shepherd et al. , 2018; Grose et al. , 2019) . Events where attributable human influence have been found include hot and cold temperature extremes (including some with widespread impacts), heavy precipitation, and certain types of droughts and tropical cyclones (AR6 WGI Section 11.9; e.g., [[#Vogel--2019|Vogel et al., 2019]] ; [[#Herring--2021|Herring et al., 2021]] ). Event attribution techniques have sometimes been extended to ‘end-to-end’ assessments from climate forcing to the impacts of events on natural or human systems ( [[#Otto--2017|Otto, 2017]] ).

<span id="attribution-of-observed-changes-in-natural-or-human-systems-to-climate-related-drivers"></span>
=== Attribution of observed changes in natural or human systems to climate-related drivers ===

The attribution of observed changes to climate-related drivers across a diverse set of sectors, regions and systems is part of each chapter in the WGII contribution to AR6 and is synthesized in WGII Chapter 16 (Section 16.2). The number of attribution studies on climate change impacts has grown substantially since AR5, generally leading to higher confidence levels in attributing the causes of specific impacts. New studies include the attribution of changes in socio-economic indicators such as economic damages due to river floods (e.g., [[#Schaller--2016|Schaller et al., 2016]] ; [[#Sauer--2021|Sauer et al., 2021]] ), the occurrence of heat-related human mortality (e.g., [[#Vicedo-Cabrera--2018|Vicedo-Cabrera et al., 2018]] ; [[#Sera--2020|Sera et al., 2020]] ) or economic inequality (e.g., [[#Diffenbaugh--2019|Diffenbaugh and Burke, 2019]] ).

Impact attribution covers a diverse set of qualitative and quantitative approaches, building on experimental approaches, observations from remote sensing, long-term in situ observations, and monitoring efforts, teamed with local knowledge, process understanding and empirical or dynamical modelling (WGII Section 16.2; [[#Stone--2013|Stone et al., 2013]] ; [[#Cramer--2014|Cramer et al., 2014]] ). The attribution of a change in a natural or human system (e.g., wild species, natural ecosystems, crop yields, economic development, infrastructure or human health) to changes in climate-related systems (i.e., climate, ocean acidification, permafrost thawing or sea level rise) requires accounting for other potential drivers of change, such as technological and economic changes in agriculture affecting crop production ( [[#Hochman--2017|Hochman et al., 2017]] ; [[#Butler--2018|Butler et al., 2018]] ), changes in human population patterns and vulnerability affecting flood- or wildfire-induced damages ( [[#Huggel--2015|Huggel et al., 2015]] ; [[#Sauer--2021|Sauer et al., 2021]] ), or habitat loss driving declines in wild species ( [[#IPBES--2019|IPBES, 2019]] ). These drivers are accounted for by estimating a baseline condition that would exist in the absence of climate change. The baseline might be stationary and be approximated by observations from the past, or it may change over time and be simulated by statistical or process-based impact models (WGII Section 16.2; Cramer et al. , 2014) . Assessment of multiple independent lines of evidence, taken together, can provide rigorous attribution when more quantitative approaches are not available ( [[#Parmesan--2013|Parmesan et al., 2013]] ). These include paleodata, physiological and ecological experiments, natural ‘experiments’ from very long-term datasets indicating consistent responses to the same climate trend/event, and ‘fingerprints’ in species’ responses that are uniquely expected from climate change (e.g. poleward range boundaries expanding and equatorial range boundaries contracting in a coherent pattern worldwide; [[#Parmesan--2003|Parmesan and Yohe, 2003]] ) . Meta-analyses of species/ecosystem responses, when conducted with wide geographic coverage, also provide a globally coherent signal of climate change at an appropriate scale for attribution to anthropogenic climate change ( [[#Parmesan--2003|Parmesan and Yohe, 2003]] ; [[#Parmesan--2013|Parmesan et al., 2013]] ).

Impact attribution does notalways involve attribution to anthropogenic climate forcing. However, a growing number of studies include this aspect (e.g., [[#Frame--2020|Frame et al. (2020)]] for the attribution of damages induced by Hurricane Harvey; or [[#Diffenbaugh--2019|Diffenbaugh and Burke (2019)]] for the attribution of economic inequality between countries; or [[#Schaller--2016|Schaller et al. (2016)]] for flood damages).

<div id="1.4.5" class="h2-container"></div>

<span id="climate-regions-used-in-ar6"></span>
=== 1.4.5 Climate Regions Used in AR6 ===

<div id="h2-26-siblings" class="h2-siblings"></div>

<div id="1.4.5.1" class="h3-container"></div>

<span id="defining-climate-regions"></span>
==== 1.4.5.1 Defining Climate Regions ====

<div id="h3-20-siblings" class="h3-siblings"></div>

The AR5 assessed regional-scale detection and attribution and assessed key regional climate phenomena and their relevance for future regional climate projections. This report shows that past and future climate changes and extreme weather events can be substantial on local and regional scales (Chapters 8–12 and Atlas), where they may differ considerably from global trends, not only in intensity but even in the direction of change (e.g., [[#Fischer--2013|Fischer et al., 2013]] ).

Although the evolution of global climate trends emerges as the net result of regional phenomena, average or aggregate estimates often do not reflect the intensity, variability and complexity of regional climate changes ( [[#Stammer--2018|Stammer et al., 2018]] ; [[#Shepherd--2019|Shepherd, 2019]] ). A fundamental aspect of the study of regional climate changes is the definition of characteristic climate zones, clusters or regions, across which the emergent climate change signal can be properly analysed and projected (see Atlas). Suitable sizes and shapes of such zones strongly depend not only on the climate variable and process of interest, but also on relevant multi-scale feedbacks.

There are several approaches to the classification of climate regions. When climate observation data was sparse and limited, the aggregation of climate variables was implicitly achieved through the consideration of biomes, giving rise to the traditional vegetation-based classification of [[#Köppen--1936|Köppen (1936)]] . In the last decades, the substantial increases in climate observations, climate modelling, and data processing capabilities have allowed new approaches to climate classification, for example through interpolation of aggregated global data from thousands of stations ( [[#Peel--2007|Peel et al., 2007]] ; [[#Belda--2014|Belda et al., 2014]] ; [[#Beck--2018|]] [[#Beck--2018|Beck et al., 2018]] ) or through data-driven approaches applied to delineate ecoregions that behave in a coherent manner in response to climate variability ( [[#Papagiannopoulou--2018|Papagiannopoulou et al., 2018]] ). Experience shows that each method has strengths and weaknesses through trade-offs between detail and convenience. For instance, a very detailed classification, with numerous complexly shaped regions derived from a large set of variables, may be most useful for the evaluation of climate models ( [[#Rubel--2010|Rubel and Kottek, 2010]] ; [[#Belda--2015|Belda et al., 2015]] ; [[#Beck--2018|]] [[#Beck--2018|Beck et al., 2018]] ) and climate projections ( [[#Feng--2014|Feng et al., 2014]] ; [[#Belda--2016|Belda et al., 2016]] ). In contrast, geometrically simple regions are often best suited for regional climate modelling and downscaling (e.g., the Coordinated Regional Climate Downscaling Experiment (CORDEX) domains; [[#1.5.3|Section 1.5.3]] ; [[#Giorgi--2015|Giorgi and Gutowski, 2015]] ).

<div id="1.4.5.2" class="h3-container"></div>

<span id="types-of-regions-used-in-ar6"></span>
==== 1.4.5.2 Types of Regions Used in AR6 ====

<div id="h3-21-siblings" class="h3-siblings"></div>

IPCC’s recognition of the importance of regional climates can be traced back to its First Assessment Report (FAR; [[#IPCC--1990a|IPCC, 1990a]] ), where climate projections for 2030 were presented for five sub-continental regions (see [[#1.3.6|Section 1.3.6]] for an assessment of those projections). In subsequent reports, there has been a growing emphasis on the analysis of regional climate, including two special reports: one on regional impacts ( [[#IPCC--1998|IPCC, 1998]] ) and another on extreme events (SREX, [[#IPCC--2012|IPCC, 2012]] ). A general feature of previous IPCC reports is that the number and coverage of climate regions vary according to the subject and across Working Groups. Such varied definitions have the advantage of optimizing the results for a particular application (e.g., national boundaries are crucial for decision-making, but they rarely delimit distinctive climate regions), whereas variable definitions of regions may have the disadvantage of hindering multidisciplinary assessments and comparisons between studies or Working Groups.

In this Report, regional climate change is primarily addressed through the introduction of four classes of regions (unless otherwise explicitly mentioned and justified). The first two are the unified WGI Reference Sets of (i) Land Regions and (ii) Ocean Regions, which are used throughout the Report. These are supplemented by additional sets of (iii) Typological Regions – used in Chapters 5, 8–12 and [[IPCC:Wg1:Chapter:Atlas|Atlas]] – and (iv) Continental Regions, which are mainly used for linking Chapters 11, 12 and [[IPCC:Wg1:Chapter:Atlas|Atlas]] with Working Group II (Figure 1.18). All four classes of regions are defined and described in detail in the Atlas. Here we summarize their basic features.

<div id="_idContainer055" class="_idGenObjectStyleOverride-1"></div>

<!-- START IMG -->
<!-- IMG FILE -->
[[File:1c73b6615276e5d22b28b1b6b48ce8fc IPCC_AR6_WGI_Figure_1_18.png]]
<!-- IMG TITLE + CAPTION -->

'''Figure 1.18 |''' '''Main region types used in this report.''' '''(a)''' AR6 WGI Reference Set of Land and Ocean Regions ( [[#Iturbide--2020|Iturbide et al., 2020]] ), consisting of 46 land regions and 15 ocean regions, including 3 hybrid regions (CAR, MED, SEA) that are both land and ocean regions. Abbreviations are explained to the right of the map. Notice that RAR, SPO, NPO and EPO extend beyond the 180º meridian, therefore appearing at both sides of the map (indicated by dashed lines). A comparison with the previous reference regions of AR5 WGI ( [[#IPCC--2013a|IPCC, 2013a]] ) is presented in the Atlas. '''(b)''' Example of typological regions: monsoon domains (see Chapter 8). Abbreviations are explained to the right of the map. The black contour lines represent the global monsoon zones, while the coloured regions denote the regional monsoon domains. The two stippled regions (EqAmer and SAfri) do receive seasonal rainfall, but their classification as monsoon regions is still under discussion. '''(c)''' Continental Regions used mainly in [https://www.ipcc.ch/report/ar6/wg1/chapter/chapter-12 Chapter 12] and the Atlas. Stippled zones define areas that are assessed in both regions (e.g., the Caribbean is assessed as Small Islands and also as part of Central America). Small Islands are ocean regions containing small islands with consistent climate signals and/or climatological coherence.
<!-- END IMG -->

The Reference Sets of Land and Ocean Regions are polygonal, sub-continental domains, defined through a combination of environmental, climatic and non-climatic (e.g., pragmatic, technical, historical) factors, in accordance with the literature and climatological reasoning based on observed and projected future climate. Merging the diverse functions and purposes of the regions assessed in the literature into a common reference set implies a certain degree of compromise between simplicity, practicality and climate consistency. For instance, Spain is fully included in the Mediterranean (MED) Reference Region, but is one of the most climatically diverse countries in the world. Likewise, a careful comparison of panels a and b of Figure 1.18 reveals that the simplified southern boundary of the Sahara (SAH) Reference Region slightly overlaps the northern boundary of the West African Monsoon Typological Region. As such, the resulting Reference Regions are not intended to precisely represent climates, but rather to provide simple domains suitable for regional synthesis of observed and modelled climate and climate change information ( [[#Iturbide--2020|Iturbide et al., 2020]] ). In particular, CMIP6 model results averaged over Reference Regions are presented in the Atlas.

The starting point for defining the AR6 Reference Sets of Land Regions was the collection of 26 regions introduced in SREX ( [[#IPCC--2012|IPCC, 2012]] ). The SREX collection was then revised, reshaped, complemented and optimized to reflect the recent scientific literature and observed climate-change trends, giving rise to the novel AR6 Reference Set of 46 Land Regions. Additionally, AR6 introduces a new Reference Set of 15 Ocean Regions (including 3 Hybrid Regions that are treated as both: land and ocean), which complete the coverage of the whole Earth ( [[#Iturbide--2020|Iturbide et al., 2020]] ).

Particular aspects of regional climate change are described by specialized domains called Typological Regions (Figure 1.18b). These regions cover a wide range of spatial scales and are defined by specificfeatures, called typologies. Examples of typologies include: tropical forests, deserts, mountains, monsoon regions and megacities, among others. Typological Regions are powerful tools to summarize complex aspects of climate defined by a combination of multiple variables. For this reason, they are used in many chapters of AR6 WGI and WGII (e.g., Chapters 8–12 and Atlas).

Finally, consistency with WGII is also pursued in Chapters 11, 12 and the [[IPCC:Wg1:Chapter:Atlas|Atlas]] through the use of a set of Continental Regions (Figure 1.18c), based on the nine continental domains defined in AR5 WGII Part B ( [[#Hewitson--2014|Hewitson et al., 2014]] ). These are classical geopolitical divisions of Africa, Asia, Australasia, Europe, North America, Central and South America, plus Small Islands, Polar Regions, and the Ocean. In AR6 WGI, five hybrid zones (Caribbean–Small Islands, East Europe–Asia, European Arctic, North American Arctic, and Northern Central America) are also identified, which are assessed in more than one Continental Region. Additional consistency with WGIII is pursued in [[IPCC:Wg1:Chapter:Chapter-6|Chapter 6]] through the use of sub-continental domains which essentially form a subset of the Continental Set of Regions (Figure 1.18c and Section 6.1).

<div id="1.5" class="h1-container"></div>

<span id="major-developments-and-their-implications"></span>