Editing IPCC:AR6/WGI/Chapter-3 (section)

===== 3.3.1.1.2 Detection and attribution =====

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

Looking at periods preceding the instrumental record, AR5 assessed with ''high confidence'' that the 20th century annual mean surface temperature warming reversed a 5000-year cooling trend in Northern Hemisphere mid- to high latitudes caused by orbital forcing, and attributed the reversal to anthropogenic forcing with ''high confidence'' (see also ( [[IPCC:Wg1:Chapter:Chapter-2#2.3.1.1|Section 2.3.1.1]] ). Since AR5, the combined response to solar, volcanic and greenhouse gas forcing was detected in all Northern Hemisphere continents ( [[#PAGES%202k-PMIP3%20group--2015|PAGES 2k-PMIP3 group, 2015]] ) over the period 864 to 1840. In contrast, the effect of those forcings was not detectable in the Southern Hemisphere ( [[#Neukom--2018|Neukom et al., 2018]] ). Global and Northern Hemisphere temperature changes from reconstructions over this period have been attributed mostly to volcanic forcing ( [[#Schurer--2014|Schurer et al., 2014]] ; [[#McGregor--2015|McGregor et al., 2015]] ; [[#Otto-Bliesner--2016|Otto-Bliesner et al., 2016]] ; [[#PAGES%202k%20Consortium--2019|PAGES 2k Consortium, 2019]] ; [[#Büntgen--2020|Büntgen et al., 2020]] ), with a smaller role for changes in greenhouse gas forcing, and solar forcing playing a minor role ( [[#Schurer--2014|Schurer et al., 2014]] ; [[#PAGES%202k%20Consortium--2019|PAGES 2k Consortium, 2019]] ).

Focusing now on warming over the historical period, AR5 assessed that it was ''extremely likely'' that human influence was the dominant cause of the observed warming since the mid-20th century, and that it was ''virtually certain'' that warming over the same period could not be explained by internal variability alone. Since AR5 many new attribution studies of changes in global surface temperature have focused on methodological advances (see also ( [[#3.2|Section 3.2]] ). Those advances include better accounting for observational and model uncertainties, and internal variability ( [[#Ribes--2013|Ribes and Terray, 2013]] ; [[#Hannart--2016|Hannart, 2016]] ; [[#Ribes--2017|Ribes et al., 2017]] ; [[#Schurer--2018|Schurer et al., 2018]] ); formulating the attribution problem in a counterfactual framework ( [[#Hannart--2018|Hannart and Naveau, 2018]] ); and reducing the dependence of the attribution on uncertainties in climate sensitivity and forcing ( [[#Otto--2015|Otto et al., 2015]] ; [[#Haustein--2017|Haustein et al., 2017]] , 2019). Studies now account for uncertainties in the statistics of internal variability, either explicitly ( [[#Hannart--2016|Hannart, 2016]] ; [[#Hannart--2018|Hannart and Naveau, 2018]] ; [[#Ribes--2021|Ribes et al., 2021]] ) or implicitly ( [[#Ribes--2013|Ribes and Terray, 2013]] ; [[#Schurer--2018|Schurer et al., 2018]] ; [[#Gillett--2021|Gillett et al., 2021]] ), thus addressing concerns about over-confident attribution conclusions. Accounting for observational uncertainty increases the range of warming attributable to greenhouse gases by only 10 to 30% ( [[#Jones--2017|Jones and Kennedy, 2017]] ; [[#Schurer--2018|Schurer et al., 2018]] ). While some attribution studies estimate attributable changes in globally-complete GSAT ( [[#Schurer--2018|Schurer et al., 2018]] ; [[#Gillett--2021|Gillett et al., 2021]] ; [[#Ribes--2021|Ribes et al., 2021]] ), others attribute changes in observational GMST, but this makes little difference to attribution conclusions ( [[#Schurer--2018|Schurer et al., 2018]] ). Moreover, based on a synthesis of observational and modelling evidence, Cross-Chapter Box 2.3 assesses that the current best estimate of the scaling factor between GMST and GSAT is one, and therefore attribution studies of GMST and GSAT are here treated together in deriving assessed warming ranges. Studies also increasingly validate their multi-model approaches using imperfect model tests ( [[#Schurer--2018|Schurer et al., 2018]] ; [[#Gillett--2021|Gillett et al., 2021]] ; [[#Ribes--2021|Ribes et al., 2021]] ). Alternative techniques, based purely on statistical or econometric approaches, without the need for climate modelling, have also been applied ( [[#Estrada--2013|Estrada et al., 2013]] ; [[#Stern--2014|Stern and Kaufmann, 2014]] ; [[#Dergiades--2016|Dergiades et al., 2016]] ) and match the results of physically-based methods. The larger range of attribution techniques and improvements to those techniques increase confidence in the results compared to AR5.

In contrast, studies published since AR5 indicate that closely constraining the separate contributions of greenhouse gas changes and aerosol changes to observed temperature changes remains challenging. Nonetheless, attribution of warming to greenhouse gas forcing has been found as early as the end of the 19th century ( [[#Schurer--2014|Schurer et al., 2014]] ; [[#Owens--2017|Owens et al., 2017]] ; [[#PAGES%202k%20Consortium--2019|PAGES 2k Consortium, 2019]] ). [[#Hegerl--2019|Hegerl et al. (2019)]] found that volcanism cooled global temperatures by about 0.1°C between 1870 and 1910, then a lack of volcanic activity warmed temperatures by about 0.1°C between 1910 and 1950, with anthropogenic aerosols cooling temperatures throughout the 20th century, especially between 1950 and 1980 when the estimated range of aerosol cooling was about 0.1°C to 0.5°C. [[#Jones--2016|Jones et al. (2016)]] attributed a warming of 0.87 to 1.22°C per century over the period 1906 to 2005 to greenhouse gases, partially offset by a cooling of −0.54°C to −0.22°C per century attributed to aerosols. But they also found that detection of the greenhouse gas or the aerosol signal often fails, because of uncertainties in modelled patterns of change and internal variability. That point is illustrated by Figure 3.7, which shows two- and three-way fingerprinting regression coefficients for 13 CMIP6 models and the corresponding attributable warming ranges, derived using HadCRUT4 ( [[#Gillett--2021|Gillett et al., 2021]] ). Regression coefficients with an uncertainty range that includes zero mean that detection has failed. Models with regression coefficients significantly less than one significantly overpredict the temperature response to the corresponding forcing. Conversely, models with regression coefficients significantly greater than one underpredict the response to these forcings. While estimates of warming attributable to anthropogenic influence derived using individual models are generally consistent, estimates of warming attributable to greenhouse gases and aerosols separately based on individual models are not all consistent, and detection of the aerosol influence fails more often than that of greenhouse gases. Hence, results of recent studies emphasize the need to use multi-model means to better constrain estimates of GSAT changes attributable to greenhouse gas and aerosol forcing ( [[#Schurer--2018|Schurer et al., 2018]] ; [[#Gillett--2021|Gillett et al., 2021]] ; [[#Ribes--2021|Ribes et al., 2021]] ).

<div id="_idContainer020" class="Basic-Text-Frame"></div>

[[File:d3a10075e7a486c6d96c99e82c0e505c IPCC_AR6_WGI_Figure_3_7.png]]

Figure 3.7 | '''Regression coefficients and corresponding attributable warming estimates for individual CMIP6 models.''' Upper panels show regression coefficients based on a two-way regression '''(left)''' and three-way regression '''(right)''' , of observed five-year mean, globally averaged, masked and blended surface temperature (HadCRUT4) onto individual model response patterns, and a multi-model mean, labelled ‘Multi’. Anthropogenic, natural, greenhouse gas, and other anthropogenic (aerosols, ozone, land-use change) regression coefficients are shown. Regression coefficients are the scaling factors by which the model responses must be multiplied to best match observations. Regression coefficients consistent with one indicate a consistent magnitude response in observations and models, and regression coefficients significantly greater than zero indicate a detectable response to the forcing concerned. Lower panels show corresponding observationally-constrained estimates of attributable warming in globally-complete GSAT for the period 2010–2019, relative to 1850–1900, and the horizontal black line shows an estimate of observed warming in GSAT for this period. Figure is adapted from [[#Gillett--2021|Gillett et al. (2021)]] , their Extended Data Figure 3. Further details on data sources and processing are available in the chapter data table (Table 3.SM.1).

Figure 3.8 compares attributable changes in globally complete GSAT for the period 2010–2019 relative to 1850–1900 from three detection and attribution studies, two of which use CMIP6 multi-model means ( [[#Gillett--2021|Gillett et al., 2021]] ; [[#Ribes--2021|Ribes et al., 2021]] ), and an estimate based on assessed effective radiative forcing and transient and equilibrium climate sensitivity (see Section 7.3.5.3). The reference period 1850–1900 is used to assess attributable temperature changes because this is when the earliest gridded surface temperature records start, this is when the CMIP6 historical simulations start, this is the earliest base period used in attribution literature, and this is a reference period used in IPCC SR1.5 and earlier reports. It should, however, be noted that Cross-Chapter Box 1.2 assesses with ''medium confidence'' that there was an anthropogenic warming with a ''likely'' range of 0.0°C–0.2°C between 1750 and 1850–1900. Figure 3.8 also shows the GSAT changes directly simulated in response to these forcings in thirteen CMIP6 models. In spite of their different methodologies and input datasets, the three attribution approaches yield very similar results, with the anthropogenic attributable warming range encompassing observed warming, and the natural attributable warming being close to zero. The warming driven by greenhouse gas increases is offset in part by cooling due to other anthropogenic forcing agents, mostly aerosols, although uncertainties in these contributions are larger than the uncertainty in the net anthropogenic warming, as discussed above. Estimates based on physical understanding of forcing and ECS made by ( [[IPCC:Wg1:Chapter:Chapter-7|Chapter 7]] are close to estimates from attribution studies, despite being the products of a different approach. This agreement enhances confidence in the magnitude and causes of attributable surface temperature warming.

<div id="_idContainer022" class="_idGenObjectStyleOverride-1"></div>

[[File:b5453f34bf66d2a892da5bef7f33d5e4 IPCC_AR6_WGI_Figure_3_8.png]]

'''Figure 3.8 | Assessed contributions to observed warming, and supporting lines of evidence.''' Shaded bands show assessed ''likely'' ranges of temperature change in GSAT, 2010–2019 relative to 1850–1900, attributable to net human influence, well-mixed greenhouse gases, other human forcings (aerosols, ozone, and land-use change), natural forcings, and internal variability, and the 5–95% range of observed warming. Bars show 5–95% ranges based on (left to right) [[#Haustein--2017|Haustein et al. (2017)]] , [[#Gillett--2021|Gillett et al. (2021)]] and [[#Ribes--2021|Ribes et al. (2021)]] , and crosses show the associated best estimates. No 5–95% ranges were provided for the [[#Haustein--2017|Haustein et al. (2017)]] greenhouse gas or other human forcings contributions. The [[#Ribes--2021|Ribes et al. (2021)]] results were updated using a revised natural forcing time series, and the [[#Haustein--2017|Haustein et al. (2017)]] results were updated using HadCRUT5. The ( [[IPCC:Wg1:Chapter:Chapter-7|Chapter 7]] best estimates and ranges were derived using assessed forcing time series and a two-layer energy balance model as described in Section 7.3.5.3. Coloured symbols show the simulated responses to the forcings concerned in each of the models indicated. Further details on data sources and processing are available in the chapter data table (Table 3.SM.1).

The AR5 found ''high confidence'' for a major role for anthropogenic forcing in driving warming over each of the inhabited continents, except for Africa where they found only ''medium confidence'' because of limited data availability ( [[#Bindoff--2013|Bindoff et al., 2013]] ). At the hemispheric scale, [[#Friedman--2020|Friedman et al. (2020)]] and [[#Bonfils--2020|Bonfils et al. (2020)]] detected an anthropogenically forced response of inter-hemispheric contrast in surface temperature change, which has a complex time evolution but shows the Northern Hemisphere cooling relative to the Southern Hemisphere until around 1975 but then warming after that. [[#Bonfils--2020|Bonfils et al. (2020)]] attribute the Northern Hemisphere reversal to a combination of reduced aerosol forcing and greenhouse gas induced warming of Northern Hemisphere land masses. [[#Friedman--2020|Friedman et al. (2020)]] found that CMIP5 models simulate the correct sign of the inter-hemispheric contrast when forced with all forcings but underestimate its magnitude. Figure 3.9 shows global surface temperature change in CMIP6 all-forcing and natural-only simulations globally, averaged over continents, and separately over land and ocean surfaces. All-forcing simulations encompass observed temperature changes for all regions, while natural-only simulations fail to do so in recent decades except in Antarctica, based on the annual means shown. As stated above, warming results from a partial offset of greenhouse gas warming by aerosol cooling. That offset is stronger over land than ocean. Regionally, models show a large range of possible temperature responses to greenhouse gas and aerosol forcing, which complicates single-forcing attribution. A more detailed discussion of regional attribution can be found in Section 10.4. Over global land surfaces, [[#Chan--2015|Chan and Wu (2015)]] used CMIP5 simulations to attribute a warming trend of 0.3 (2.5%–97.5% confidence interval: 0.2–0.36) °C per decade to anthropogenic forcing, with natural forcing only contributing 0.05 (0.02–0.06) °C per decade. Accounting for unsampled sources of uncertainty and the availability of only a single study, their result suggests that it is ''very likely'' that human influence is the main driver of warming over land.

<div id="_idContainer024" class="Basic-Text-Frame"></div>

[[File:85198fe800e9fd13ec9c5f9973af10cb IPCC_AR6_WGI_Figure_3_9.png]]

'''Figure 3.9 | Global, land, ocean and continental annual mean near-surface air temperatures anomalies in CMIP6 models and observations.''' Time series are shown for CMIP6 historical anthropogenic and natural (brown), natural-only (green), greenhouse gas only (grey) and aerosol only (blue) simulations (thick lines show multi-model means and shaded regions show the 5th to 95th percentile ranges) and for HadCRUT5 (black). All models have been subsampled using the HadCRUT5 observational data mask. Temperature anomalies are shown relative to 1950–2010 for Antarctica and relative to 1850–1900 for other continents. CMIP6 historical simulations are extended using the SSP2-4.5 scenario simulations. All available ensemble members were used (see [[#3.2|Section 3.2]] ). Regions are defined by [[#Iturbide--2020|Iturbide et al. (2020)]] . Further details on data sources and processing are available in the chapter data table (Table 3.SM.1).

In summary, since the publication of AR5, new literature has emerged that better accounts for methodological and climate model uncertainties in attribution studies ( [[#Ribes--2017|Ribes et al., 2017]] ; [[#Hannart--2018|Hannart and Naveau, 2018]] ) and that concludes that anthropogenic warming is approximately equal to observed warming over the 1951–2010 period. The IPCC SR1.5 reached the same conclusion for 2017 relative to 1850–1900 based on anthropogenic warming and associated uncertainties calculated using the method of [[#Haustein--2017|Haustein et al. (2017)]] . Moreover, the improved understanding of the causes of the apparent slowdown in warming over the beginning of the 21st century and the difference in simulated and observed warming trends over this period (Cross-Chapter Box 3.1) further improve our confidence in the assessment of the dominant anthropogenic contribution to observed warming. In deriving our assessments, these considerations are balanced against new literature that raises questions about the ability of some models to simulate variability in surface temperatures over a range of time scales ( [[#Laepple--2014|Laepple and Huybers, 2014]] ; [[#Parsons--2017|Parsons et al., 2017]] ; [[#Friedman--2020|Friedman et al., 2020]] ), and the finding that some CMIP6 models exhibit substantially higher multi-decadal internal variability than that seen in CMIP5, which remains to be fully understood ( [[#Parsons--2020|Parsons et al., 2020]] ; [[#Ribes--2021|Ribes et al., 2021]] ). Further, uncertainties in simulated aerosol-cloud interactions are still large (Section 7.3.3.2.2), resulting in very diverse spatial responses of different climate models to aerosol forcing, and inter-model differences in the historical global mean temperature evolution and in diagnosed cooling attributable to aerosols (Figure 3.8). Moreover, like previous generations of coupled model simulations, historical and single forcing CMIP6 simulations follow a common experimental design ( [[#Eyring--2016a|Eyring et al., 2016a]] ; [[#Gillett--2016|Gillett et al., 2016]] ) and are thus all driven by the same common set of forcings, even though these forcings are uncertain. Hence, forcing uncertainty is not directly accounted for in most of the attribution and model evaluation studies assessed here, although this limitation can to some extent be addressed by comparing with previous generation multi-model ensembles or individual model studies using different sets of forcings.

The IPCC SR1.5 best estimate and ''likely'' range of anthropogenic attributable GMST warming was 1.0 ± 0.2°C in 2017 with respect to the period 1850–1900. Here, the best estimate is expressed in terms of GSAT and is calculated as the average of the three estimates shown in Figure 3.9, yielding a value of 1.07°C. Ranges for attributable GSAT warming are derived by finding the smallest ranges with a precision of 0.1°C which span all of the 5–95% ranges from the attribution studies shown in Figure 3.9. These ranges are then assessed as ''likely'' rather than ''very likely'' because the studies may underestimate the importance of the structural limitations of climate models, which probably do not represent all possible sources of internal variability; use too simple climate models, which may underestimate the role of internal variability; or underestimate model uncertainty, especially when using model ensembles of limited size and inter-dependent models, for example through common errors in forcings across models, as discussed above. This leads to a ''likely'' range for anthropogenic attributable warming in 2010–2019 relative to 1850–1900 of 0.8 to 1.3°C in terms of GSAT. This range encompasses the best estimate and ''very likely'' range of observed GSAT warming of 1.06 [0.88 to 1.21] °C over the same period (Cross-Chapter Box 2.3). There is ''medium confidence'' that the best estimate and ''likely'' ranges of attributable warming expressed in terms of GMST are equal to those for GSAT (Cross-Chapter Box 2.3). Repeating the process for other time periods leads to the best estimates and ''likely'' ranges listed in Table 3.1. GSAT change attributable to natural forcings is −0.1 to +0.1°C. The ''likely'' range of GSAT warming attributable to greenhouse gases is assessed in the same way to be 1.0 to 2.0°C while the GSAT change attributable to aerosols, ozone and land-use change is −0.8 to 0.0°C. Progress in attribution techniques allows the important advance of attributing observed surface temperature warming since 1850–1900, instead of since 1951 as was done in AR5.

<div id="_idContainer025"></div>

Table 3.1 | '''Estimates of warming in GSAT attributable to human influence for different periods in °C, all relative to the 185''' '''0''' '''–1900 base period.''' Uncertainty ranges are 5–95% ranges for individual studies and ''likely'' ranges for the assessment. The results shown in the table use the methods described in the three studies indicated, but applied to additional periods and the warming trend. [[#Ribes--2021|Ribes et al. (2021)]] results were updated using a corrected natural forcing time series, and [[#Haustein--2017|Haustein et al. (2017)]] results were updated to use HadCRUT5.

{| class="wikitable"
|-
|
| 1986–2005

| 1995–2014

| 2006–2015

| 2010–2019

| Warming Rate 2010–2019

|-
| [[#Ribes--2021|Ribes et al. (2021)]]

| 0.65 (0.52 to 0.77)

| 0.82 (0.69 to 0.94)

| 0.94 (0.8 to 1.08)

| 1.03 (0.89 to 1.17)

| 0.23 (0.18 to 0.29)

|-
| [[#Gillett--2021|Gillett et al. (2021)]]

| 0.63 (0.32 to 0.94)

| 0.84 (0.63 to 1.06)

| 0.98 (0.74 to 1.22)

| 1.11 (0.92 to 1.30)

| 0.35 (0.30 to 0.41)

|-
| [[#Haustein--2017|Haustein et al. (2017)]]

| 0.73 (0.58 to 0.82)

| 0.88 (0.75 to 0.98)

| 0.98 (0.87 to 1.10)

| 1.06 (0.94 to 1.22)

| 0.23 (0.19 to 0.35)

|-
| Assessment

| 0.68 (0.3 to 1.0)

| 0.85 (0.6 to 1.1)

| 0.97 (0.7 to 1.3)

| 1.07 (0.8 to 1.3)

| 0.2 (0.1 to 0.3)

|}

The IPCC AR5 assessed the ''likely'' range of the contribution of internal variability to GMST warming to be −0.1 to +0.1°C over the period 1951–2010. Since then, several studies have downplayed the contribution of internal modes of variability to global temperature variability, often by arguing for a forced component to those internal modes ( [[#Mann--2014|Mann et al., 2014]] ; [[#Folland--2018|Folland et al., 2018]] ; [[#Haustein--2019|Haustein et al., 2019]] ; [[#Liguori--2020|Liguori et al., 2020]] ). [[#Haustein--2017|Haustein et al. (2017)]] found a 5–95% confidence interval of −0.09°C to +0.12°C for the contribution of internal variability to warming between 1850–1879 and 2017. [[#Ribes--2021|Ribes et al. (2021)]] imply a contribution of internal variability of −0.02°C ± 0.16°C to warming between 2010–2019 and 1850–1900, assuming independence between errors in the observations and in the estimate of the forced response. Based on these studies, but allowing for unsampled sources of error, we assess the ''likely'' range of the contribution of internal variability to GSAT warming between 2010–2019 and 1850–1900 to be −0.2°C to +0.2°C.

The IPCC SR1.5 gave a ''likely'' range for the human-induced warming rate of 0.1°C to 0.3°C per decade in 2017, with a best estimate of 0.2°C per decade ( [[#Allen--2018|Allen et al., 2018]] ). Table 3.1 lists the estimates of attributable anthropogenic warming rate over the period 2010–2019 based on the three studies that underpin the assessment of GSAT warming ( [[#Haustein--2017|Haustein et al., 2017]] ; [[#Gillett--2021|Gillett et al., 2021]] ; [[#Ribes--2021|Ribes et al., 2021]] ). Estimates from [[#Haustein--2017|Haustein et al. (2017)]] , based on observed warming, and [[#Ribes--2021|Ribes et al. (2021)]] , based on CMIP6 simulations constrained by observed warming, are in good agreement. The [[#Gillett--2021|Gillett et al. (2021)]] estimate, also based on CMIP6 models, corresponds to a larger anthropogenic attributable warming rate, because of a smaller warming rate attributed to natural forcing than in [[#Ribes--2021|Ribes et al. (2021)]] . This disagreement does not support a decrease in uncertainty compared to the SR1.5 assessment. So the range for anthropogenic attributable surface temperature warming rate of 0.1°C to 0.3°C per decade is again assessed to be ''likely'' , with a best estimate of 0.2°C per decade.

<div id="3.3.1.2" class="h3-container"></div>

<span id="upper-air-temperature"></span>