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

== 3.7 Human Influence on Modes of Climate Variability ==

<div id="h1-8-siblings" class="h1-siblings"></div>

This section assesses model evaluation and attribution of changes in the modes of climate variability listed in Cross-Chapter Box 2.2, Table 2. The structure of the modes is described in Annex IV, observed changes in the modes and associated teleconnections are assessed in [[IPCC:Wg1:Chapter:Chapter-2#2.4|Section 2.4]] , and the role of the modes in shaping regional climate is assessed in Section 10.1.3.2.

<div id="3.7.1" class="h2-container"></div>

<span id="north-atlantic-oscillation-and-northern-annular-mode"></span>
=== 3.7.1 North Atlantic Oscillation and Northern Annular Mode ===

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

The Northern Annular Mode (NAM; also known as the Arctic Oscillation) is an oscillation of atmospheric mass between the Arctic and northern mid-latitudes, analogous to the Southern Annular Mode (SAM; [[#3.7.2|Section 3.7.2]] ). It is the leading mode of variability of sea-level pressure in the northern extratropics but also has a clear fingerprint through the troposphere up to the lower stratosphere, with maximum expression in boreal winter ( [[#Kidston--2015|Kidston et al., 2015]] ). The North Atlantic Oscillation (NAO) can be interpreted as the regional expression of the NAM and captures most of the related variance in the troposphere over a broad North Atlantic/Europe domain. Indices measuring the state of the NAO correlate highly with those of the NAM, and teleconnection patterns for both modes are rather similar ( [[#Feldstein--2006|Feldstein and Franzke, 2006]] ). A detailed description of the NAM and the NAO as well as their associated teleconnection over land is given in Annex IV.2.1.

AR5 found that while models simulated correctly most of the spatial properties of the NAM, substantial inter-model differences remained in the details of the associated teleconnection patterns over land ( [[#Flato--2013|Flato et al., 2013]] ). The AR5 reported that most models did not reproduce the observed positive trend of the NAO/NAM indices during the second half of the 20th century. It was unclear to what extent this failure reflected model shortcomings and/or if the observed trend could be simply related to pronounced internal climate variability. The AR5 accordingly did not make an attribution assessment for the NAO/NAM.

New studies since AR5 continue to find that CMIP5 models reproduce the spatial structure and magnitude of the NAM reasonably well ( [[#Lee--2013|Lee and Black, 2013]] ; [[#Zuo--2013|Zuo et al., 2013]] ; [[#Davini--2014|Davini and Cagnazzo, 2014]] ; [[#Ying--2014|Ying et al., 2014]] ; [[#Ning--2016|Ning and Bradley, 2016]] ; [[#Deser--2017b|Deser et al., 2017b]] ; [[#Gong--2017|Gong et al., 2017]] ) although the North Pacific SLP anomalies remain generally too strong ( [[#Zuo--2013|Zuo et al., 2013]] ; [[#Gong--2017|Gong et al., 2017]] ) and the subtropical North Atlantic lobe of SLP anomalies conversely too weak ( [[#Ning--2016|Ning and Bradley, 2016]] ) in many models. Such overall biases noted in both CMIP3 and CMIP5 ( [[#Davini--2014|Davini and Cagnazzo, 2014]] ) persist in CMIP6 historical simulations, even though the multi-model multi-member ensemble mean spatial correlation between modelled and observed NAM is slightly higher (Figure 3.33a,d,g). Regarding the NAO, the majority of CMIP5 models very successfully simulate its spatial structure ( [[#Lee--2019|Lee et al., 2019]] ) and its associations with extratropical jet, storm track and blocking variations over a broad North-Atlantic/Europe domain ( [[#Davini--2014|Davini and Cagnazzo, 2014]] ) and over land through teleconnections ( [[#Volpi--2020|Volpi et al., 2020]] ). The good performance of the models is confirmed in CMIP6 with a marginal improvement of the averaged observation-model spatial correlation (Figure 3.33b,e,h) and better skill based on other evaluation metrics ( [[#Fasullo--2020|Fasullo et al., 2020]] ). The slight underestimation of the SLP anomalies related to the NAO centres of actions over the Azores and Greenland–Iceland–Norwegian Seas remain unchanged compared to CMIP5.

<div id="_idContainer076" class="•-2-columns"></div>

[[File:b133ef1b4fb9fc8d2307ebd86e869aa9 IPCC_AR6_WGI_Figure_3_33.png]]

Figure 3.33 | '''Model evaluation of NAM, NAO and SAM in boreal winter.''' Regression of Mean Sea Level Pressure (MSLP) anomalies (in hPa) onto the normalized principal component (PC) of the leading mode of variability obtained from empirical orthogonal decomposition of the boreal winter (December–February) MSLP poleward of 20°N for the observed Northern Annular Mode '''(NAM, a)''' , over 20°N–80°N, 90°W–40°E for the North Atlantic Oscillation as shown by the black sector '''(NAO, b)''' , and poleward of 20°S for the Southern Annular Mode '''(SAM, c)''' for the JRA-55 reanalysis. Cross marks indicate regions where the anomalies are not significant at the 10% level based on a t-test. The period used to calculate the NAO/NAM is 1958–2014 but 1979–2014 for the SAM. '''(d–f)''' Same but for the multi-model ensemble (MME) mean from CMIP6 historical simulations. Models are weighted in compositing to account for differences in their respective ensemble size. Diagonal lines show regions where less than 80% of the runs agree in sign. '''(g–i)''' Taylor diagrams summarizing the representation of the modes in models and observations following [[#Lee--2019|Lee et al. (2019)]] for CMIP5 (light blue) and CMIP6 (red) historical simulations. The reference pattern is taken from JRA-55 (a–c). The ratio of standard deviation to that of the reference observations (radial distance), spatial correlation (radial angle) and resulting root-mean-squared errors (solid isolines) are given for individual ensemble members (crosses) and for other observational products (ERA5 and NOAA 20CR version 3, black dots). Coloured dots stand for weighted multi-model mean statistics for CMIP5 (blue) and CMIP6 (light red) as well as for AMIP simulations from CMIP6 (orange). '''(j–l)''' Histograms of the trends built from all individual ensemble members and all the models (brown bars). Vertical lines in black show all the observational estimates. The orange, light red, and light blue lines indicate the weighted multi-model mean of CMIP6 AMIP, CMIP6 and CMIP5 historical simulations, respectively. Further details on data sources and processing are available in the chapter data table (Table 3.SM.1).

CMIP5 models with a model top within the stratosphere seriously underestimate the amplitude of the variability of the wintertime NAM expression in the stratosphere, in contrast to CMIP5 models which extend well above the stratopause ( [[#Lee--2015|Lee and Black, 2015]] ). However, even in the latter models, the stratospheric NAM events, and their downward influence on the troposphere, are insufficiently persistent ( [[#Charlton-Perez--2013|Charlton-Perez et al., 2013]] ; [[#Lee--2015|Lee and Black, 2015]] ). Increased vertical resolution does not show any significant added value in reproducing the structure and magnitude of the tropospheric NAM ( [[#Lee--2013|Lee and Black, 2013]] ) nor in the NAO predictability as assessed in a seasonal prediction context with a multi-model approach ( [[#Butler--2016|Butler et al., 2016]] ). On the other hand, there is mounting evidence that a correct representation of the Quasi Biennal Oscillation, extratropical stratospheric dynamics (the polar vortex and sudden stratospheric warmings), and related troposphere-stratosphere coupling, as well as their interplay with ENSO, are important for NAO/NAM timing ( [[#Scaife--2016|Scaife et al., 2016]] ; [[#Karpechko--2017|Karpechko et al., 2017]] ; [[#Domeisen--2019|Domeisen, 2019]] ; [[#Domeisen--2019|Domeisen et al., 2019]] ), in spite of underestimated troposphere–stratosphere coupling found in models compared to observations ( [[#O’Reilly--2019b|O’Reilly et al., 2019b]] ).

The observed trend of the NAM and NAO indices is positive in winter when calculated from the 1960s ( [[IPCC:Wg1:Chapter:Chapter-2#2.4.1.1|Section 2.4.1.1]] ) but it includes large multi-decadal variability, which means that the nature of the trend should be interpreted with caution ( [[#Gillett--2013|Gillett et al., 2013]] ). The multi-model multi-member ensemble mean of the trend estimated from historical simulations over that period is very close to zero for both CMIP5 and CMIP6 (Figures 3.33j,k and 3.34a). Even if one cannot rule out that 1958–2014 was an exceptional period of variability, the observational estimates of the wintertime NAO trend lie outside the 5th–95th percentile range of the distribution of trends in the CMIP6 historical simulations, and the observed NAM trends over the same period lie above the 90th percentile. There is a tendency for the CMIP5 models to systematically underestimate the level of multi-decadal versus interannual variability of the winter NAO and jet stream compared to observations (X. [[#Wang--2017|]] [[#Wang--2017|]] [[#Wang--2017|Wang et al., 2017]] ; [[#Bracegirdle--2018|Bracegirdle et al., 2018]] ; [[#Simpson--2018|Simpson et al., 2018]] ). Results from CMIP6 (Figure 3.33j,k) and over the 1958–2019 period (Figure 3.34a) confirm this conclusion and seriously question the ability of the models to simulate long-term fluctuations of the NAO/NAM, independently of its forced or internal origins.

<div id="_idContainer078" class="•-2-columns"></div>

[[File:0b97dbd737fa427f5e11d81685ab718a IPCC_AR6_WGI_Figure_3_34.png]]

Figure 3.34 | '''Attribution of observed seasonal trends in the annular modes to forcings.''' Simulated and observed trends in NAM indices over 1958–2019 '''(a)''' and in SAM indices over 1979–2019 '''(b)''' and over 2000–2019 '''(c)''' for boreal winter (December–February average; DJF) and summer (June–August average; JJA). The indices are based on the difference of the normalized zonally averaged monthly mean sea level pressure between 35°N and 65°N for the NAM and between 40°S and 65°S for the SAM as defined in [[#Jianping--2003|Jianping and Wang (2003)]] and [[#Gong--1999|Gong and Wang (1999)]] , respectively; the unit is decade <sup>–</sup> <sup>1</sup> . Ensemble mean, interquartile ranges and 5th and 95th percentiles are represented by empty boxes and whiskers for pre-industrial control simulations and historical simulations. The number of ensemble members and models used for computing the distribution is given in the upper-left legend. Grey lines show observed trends from the ERA5 and JRA-55 reanalyses. Multi-model multi-member ensemble means of the forced component of the trends as well as their 5–95% confidence intervals assessed from t-statistics, are represented by filled boxes, based on CMIP6 individual forcing simulations from DAMIP ensembles; greenhouse gases in brown, aerosols in light blue, stratospheric ozone in purple and natural forcing in green. Models with at least three ensemble members are used for the filled boxes, with black dots representing the ensemble means of individual models. Further details on data sources and processing are available in the chapter data table (Table 3.SM.1).

Dedicated SST-forced stand-alone atmospheric model experiments (AMIP) suggest that ocean forcing appears to play a role in decadal variability of the NAO and associated fluctuations in the strength of the jet ( [[#Woollings--2015|Woollings et al., 2015]] ). In particular, Atlantic and Indian Ocean SST anomalies ( [[#Fletcher--2015|Fletcher and Cassou, 2015]] ; [[#Baker--2019|Baker et al., 2019]] ; [[#Douville--2019|Douville et al., 2019]] ; [[#Dhame--2020|Dhame et al., 2020]] ) may have contributed to the long-term positive trend of the winter NAO/NAM over the 20th century, but there is only ''low confidence'' in such a causal relationship because of the limitation of the imposed SST approach in AMIP and the uncertainties in observed SST trends among datasets used as forcing of the atmospheric model. The representation of the NAM and NAO spatial structure is slightly improved in AMIP ensembles (Figure 3.33g,h), which also produce slightly larger trends than the historical simulations for the NAO, but not for the NAM.

When calculated over the most recent two decades, the wintertime NAM/NAO trend is weakly negative since the mid-1990s ( [[#Hanna--2015|Hanna et al., 2015]] ). Recent studies based on observations ( [[#Gastineau--2015|Gastineau and Frankignoul, 2015]] ) and dedicated modelling experiments ( [[#Davini--2015|Davini et al., 2015]] ; [[#Peings--2016|Peings and Magnusdottir, 2016]] ) suggest that the recent dominance of negative NAM/NAO could be partly related to the latest shift of the Atlantic Multi-decadal Variability (AMV) to a warm phase (Sections 2.4.4 and 3.7.7). Some recent modelling studies also find that the Arctic sea ice decline might be partly responsible for more recurrent negative NAM/NAO ( [[#Peings--2013|Peings and Magnusdottir, 2013]] ; B.M. [[#Kim--2014|]] [[#Kim--2014|Kim et al., 2014]] ; [[#Nakamura--2015|Nakamura et al., 2015]] ), while other studies do not robustly identify such responses in models (see also Cross-Chapter Box 10.1).

In contrast to winter, the observed trend of the NAO index over 1958–2014 is overall negative in summer and is associated with more recurrent blocking conditions over Greenland, in particular since the mid-1990s, thus contributing to the acceleration of melting of the Arctic sea ice ( [[#3.4.1.1|Section 3.4.1.1]] ) and Greenland Ice Sheet ( [[#3.4.3.2|Section 3.4.3.2]] ; [[#Fettweis--2013|Fettweis et al., 2013]] ; [[#Hanna--2015|Hanna et al., 2015]] ; [[#Ding--2017|Ding et al., 2017]] ). The origin of the negative trend of the summer NAO has not been clearly identified, and is hypothesized to be the result of combined influences ( [[#Lim--2019|Lim et al., 2019]] ), though trends in summertime NAO should also be interpreted with caution because of the presence of strong multi-decadal variability. The recent observed negative NAO prevalence and related blocking over Greenland is not present in any of the CMIP5 models ( [[#Hanna--2018|Hanna et al., 2018]] ).

Regarding the influence of external forcings since pre-industrial times, AR5 noted that CMIP5 models tend to show an increase in the NAM in response to greenhouse gas increases ( [[#Bindoff--2013|Bindoff et al., 2013]] ). Based on the CMIP5 historical ensemble, [[#Gillett--2013|Gillett and Fyfe (2013)]] however showed that such a trend is not significant in all seasons. A multi-model assessment of eight CMIP5 models found a NAM increase in response to greenhouse gases, but no robust influence of aerosol changes ( [[#Gillett--2013|Gillett et al., 2013]] ). As for ozone depletion, there is no robust detectable influence on long-term trends of the NAO/NAM ( [[#Karpechko--2018|Karpechko et al., 2018]] ) in contrast to the SAM ( [[#3.7.2|Section 3.7.2]] ), but there are indications that extreme Arctic ozone depletion events and their surface expression are linked to an anomalously strong NAM episodes ( [[#Calvo--2015|Calvo et al., 2015]] ; [[#Ivy--2017|Ivy et al., 2017]] ). However, the direction of causality here is not clear.

Conclusions on external forcing influences on the NAM are supported by CMIP6 results based on single forcing ensembles (Figure 3.34a). Positive trends are found in historical simulations over 1958–2019 in boreal winter and are mainly driven by greenhouse gas increases. No significant trends are simulated in response to anthropogenic aerosols, stratospheric ozone or natural forcing. Albeit weak and not statistically significant, the sign of the multi-model mean forced response due to natural forcing is consistent with the observed reduction of solar activity since the 1980s ( [[IPCC:Wg1:Chapter:Chapter-2#2.2.1|Section 2.2.1]] ) whose influence would have favoured the negative phase of wintertime NAM/NAO based on the fingerprint of the nearly periodical 11-year solar cycle extracted from models ( [[#Scaife--2013|Scaife et al., 2013]] ; [[#Andrews--2015|Andrews et al., 2015]] ; [[#Thiéblemont--2015|Thiéblemont et al., 2015]] ) or observations ( [[#Gray--2016|Gray et al., 2016]] ; [[#Lüdecke--2020|Lüdecke et al., 2020]] ). But such an NAO response to solar forcing remains highly uncertain and controversial, being contradicted by longer proxy records over the last millennium ( [[#Sjolte--2018|Sjolte et al., 2018]] ) and modelling evidence ( [[#Gillett--2013|Gillett and Fyfe, 2013]] ; [[#Chiodo--2019|Chiodo et al., 2019]] ). For all seasons and for all individual forcings, uncertainties remain in the estimation of the forced response in the NAM trend as evidenced by considerable model spread (Figure 3.34a) and because the simulated forced component has small amplitude compared to internal variability.

Despite new efforts since AR5 to reconstruct the NAO beyond the instrumental record, it is still very challenging to assess the role of external forcings in the apparent multi-decadal to centennial variability present throughout the last millennium. Large uncertainties remain in the reconstructed NAO index that are sensitive to the types of proxies and statistical methods ( [[#Trouet--2012|Trouet et al., 2012]] ; [[#Ortega--2015|Ortega et al., 2015]] ; [[#Anchukaitis--2019|Anchukaitis et al., 2019]] ; [[#Cook--2019|Cook et al., 2019]] ; [[#Hernández--2020|Hernández et al., 2020]] ; [[#Michel--2020|Michel et al., 2020]] ) and reconstructed NAO variations are often not reproduced using pseudo-proxy approaches in models ( [[#Lehner--2012|Lehner et al., 2012]] ; [[#Landrum--2013|Landrum et al., 2013]] ). At low frequency, it remains challenging to evaluate if the observed or reconstructed signal corresponds to an actual change in the NAO intraseasonal to interannual intrinsic properties or rather to a change in the mean background atmospheric circulation changes projecting on a specific phase of the mode. Consequently, conflicting results emerge in the attribution of reconstructed long-term variations in the NAO to solar forcing, whose influence thus remains controversial ( [[#Gómez-Navarro--2013|Gómez-Navarro and Zorita, 2013]] ; [[#Moffa-Sánchez--2014|Moffa-Sánchez et al., 2014]] ; [[#Ortega--2015|Ortega et al., 2015]] ; [[#Ait%20Brahim--2018|Ait Brahim et al., 2018]] ; [[#Sjolte--2018|Sjolte et al., 2018]] ; [[#Xu--2018|Xu et al., 2018]] ). Influences from major volcanic eruptions appear to be more robust ( [[#Ortega--2015|Ortega et al., 2015]] ; [[#Swingedouw--2017|Swingedouw et al., 2017]] ) even if some modelling experiments question the amplitude of the response, which mostly projects on the positive phase of the NAM/NAO ( [[#Bittner--2016|Bittner et al., 2016]] ). The forced response is dependent on the strength, seasonal timing and location of the eruption but may also depend on the mean climate background state ( [[#Zanchettin--2013|Zanchettin et al., 2013]] ) and/or the phases of the main modes of decadal variability such as the AMV ( [[#3.7.7|Section 3.7.7]] ; [[#Ménégoz--2018|Ménégoz et al., 2018]] ).

Finally, there is some evidence of an apparent signal-to-noise problem referred to as ‘paradox’ in seasonal and decadal hindcasts of the NAO over the period 1979–2018 ( [[#Scaife--2018|Scaife and Smith, 2018]] ), which suggests that the NAO response to external forcing, SST or sea ice anomalies could be too weak in models. The weakness of the signal has been related to troposphere-stratosphere coupling which is too intermittent ( [[#O’Reilly--2019b|O’Reilly et al., 2019b]] ) and to chronic model biases in the persistence of NAO/NAM daily regimes, which is critically underestimated in coupled models ( [[#Strommen--2019|Strommen and Palmer, 2019]] ; [[#Zhang--2019|Zhang and Kirtman, 2019]] ), and which does not exhibit significant improvement when model resolution is increased ( [[#Fabiano--2020|Fabiano et al., 2020]] ). Note, however, that the apparent signal-to-noise problem may be dependent on the period analysed over the 20th century, which questions its interpretation as a general characteristic of coupled models ( [[#Weisheimer--2020|Weisheimer et al., 2020]] ).

In summary, CMIP5 and CMIP6 models are skilful in simulating the spatial features and the variance of the NAM/NAO and associated teleconnections ( ''high confidence'' ). There is ''limited evidence'' for a significant role for anthropogenic forcings in driving the observed multi-decadal variations of the NAM/NAO from the mid 20th century. Confidence in attribution is ''low'' : (i) because there is a large spread in the modelled forced responses which is overwhelmed anyway by internal variability; (ii) because of the apparent signal-to-noise problem; and (iii) because of the chronic inability of models to produce a range of trends which encompasses the observed estimates over the last 60 years.

<div id="3.7.2" class="h2-container"></div>

<span id="southern-annular-mode"></span>
=== 3.7.2 Southern Annular Mode ===

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

The Southern Annular Mode (SAM) consists of a meridional redistribution of atmospheric mass around Antarctica (Figure 3.33c,f), associated with a meridional shift of the jet and surface westerlies over the Southern Ocean. SAM indices are variously defined as the difference in zonal-mean sea level pressure or geopotential height between middle and high latitudes or via a principal-component analysis (Annex IV.2.2). Observational aspects of the SAM are assessed in [[IPCC:Wg1:Chapter:Chapter-2#2.4.1.2|Section 2.4.1.2]] .

AR5 assessed that CMIP5 models have medium performance in reproducing the SAM with biases in pattern ( [[#Flato--2013|Flato et al., 2013]] ). It also concluded that the trend of the SAM toward its positive phase in austral summer since the mid-20th century is ''likely'' to be due in part to stratospheric ozone depletion, and there was ''medium confidence'' that greenhouse gases have also played a role ( [[#Bindoff--2013|Bindoff et al., 2013]] ). Based on proxy reconstructions, AR5 found with ''medium confidence'' that the positive SAM trend since 1950 was anomalous compared to the last 400 years ( [[#Masson-Delmotte--2013|Masson-Delmotte et al., 2013]] ).

Additional research has shown that CMIP5 models reproduce the spatial structure of the SAM well, but tend to overestimate its variability in austral summer at interannual time scales, although this variability is within the observational uncertainty (Figure 3.33c,f,i; [[#Zheng--2013|Zheng et al., 2013]] ; [[#Schenzinger--2015|Schenzinger and Osprey, 2015]] ). This is related to the models’ tendency to simulate slightly more persistent SAM anomalies in summer compared to reanalyses ( [[#Schenzinger--2015|Schenzinger and Osprey, 2015]] ; [[#Bracegirdle--2020|Bracegirdle et al., 2020]] ). This may be due in part to too weak a negative feedback from tropospheric planetary waves ( [[#Simpson--2013|Simpson et al., 2013]] ). CMIP6 models show improved performance in reproducing the spatial structure and interannual variance of the SAM in summer based on [[#Lee--2019|Lee et al. (2019)]] diagnostics (Figure 3.33i), with a better match of its trend with reanalyses over 1979–2014 (Figure 3.33l), more realistic persistence and improved positioning of the westerly jet, which in CMIP5 models on average is located too far equatorward ( [[#Bracegirdle--2020|Bracegirdle et al., 2020]] ; [[#Grose--2020|Grose et al., 2020]] ). In CMIP5, it is also found that models which extend above the stratopause tend to simulate stronger summertime trends in the late 20th century than their counterparts with tops within the stratosphere ( [[#Rea--2018|Rea et al., 2018]] ; [[#Son--2018|Son et al., 2018]] ), though other differences between these sets of models, such as additional physical processes operating in the stratosphere or interactive ozone chemistry, may have also affected these results ( [[#Gillett--2003a|Gillett et al., 2003a]] ; [[#Sigmond--2008|Sigmond et al., 2008]] ; [[#Rea--2018|Rea et al., 2018]] ). At the surface, [[#Ogawa--2015|Ogawa et al. (2015)]] demonstrate with an atmospheric model the importance of sharp mid-latitude SST gradients for stratospheric ozone depletion to affect the SAM in summer. These studies imply that the well resolved stratosphere combined with finer ocean horizontal resolution has contributed to the stronger simulated trends in CMIP6 than in CMIP5.

CMIP6 historical simulations capture the observed positive trend of the summertime SAM when calculated from the 1970s to the 2010s (Figure 3.34b). J.L. [[#Thomas--2015|]] [[#Thomas--2015|Thomas et al. (2015)]] found that the chance of the observed 1980–2004 trend occurring only due to internal variability is less than 10% in many of the CMIP5 models, and results from CMIP6 models suggest that the chance of the 1979–2019 trend being due to internal variability could be even lower (Figure 3.34b). Although paleo-reconstructions of the SAM index are uncertain and vary in terms of long-term trends ( [[IPCC:Wg1:Chapter:Chapter-2#2.4.1.2|Section 2.4.1.2]] ), new reconstructions show that the 60-year summertime SAM trend since the mid-20th century is outside the 5th–95th percentile range of the trends in the pre-industrial variability, which matches the trend range of CMIP5 pre-industrial control simulations well ( [[#Dätwyler--2018|Dätwyler et al., 2018]] ).

In general agreement with AR5, new research continues to indicate that both stratospheric ozone depletion and increasing greenhouse gases have contributed to the trend of the SAM during austral summer toward its positive phase in recent decades ( [[#Solomon--2016|Solomon and Polvani, 2016]] ), with the ozone depletion influence dominating ( [[#Gerber--2014|Gerber and Son, 2014]] ; [[#Son--2018|Son et al., 2018]] ). In CMIP6 historical simulations there are significant positive SAM trends over the 1979–2019 period in austral summer, although the contribution from ozone forcing evaluated with the four available models is not significant (Figure 3.34b). Three of these models share the same standard prescribed ozone forcing and produce significantly positive SAM trends over an extended period (1957–2019). The fourth model, MRI-ESM2-0, has the option of interactive ozone chemistry. Its ozone-only experiment is forced by prescribed ozone derived from its own historical simulations and produces a negative SAM trend associated with weak ozone depletion ( [[#Morgenstern--2020|Morgenstern et al., 2020]] ). [[#Morgenstern--2014|Morgenstern et al. (2014)]] and [[#Morgenstern--2021|Morgenstern (2021)]] find an indirect influence of greenhouse gases on the SAM via induced ozone changes in coupled chemistry-climate simulations, which differ from the prescribed ozone simulations shown in Figure 3.34b. Since about 1997, the effective abundance of ozone-depleting halogen has been decreasing in the stratosphere ( [[#WMO--2018|WMO, 2018]] ), leading to a stabilization or even a reversal of stratospheric ozone depletion (Sections 2.2.5.2 and 6.3.2.2). The ozone stabilization and slight recovery since about 2000 may have caused a pause in the summertime SAM trend (Figure 3.34c; [[#Saggioro--2019|Saggioro and Shepherd, 2019]] ; [[#Banerjee--2020|Banerjee et al., 2020]] ), although some influence from internal variability cannot be ruled out. While some studies find an anthropogenic aerosol influence on the summertime SAM ( [[#Gillett--2013|Gillett et al., 2013]] ; [[#Rotstayn--2013|Rotstayn, 2013]] ), recent studies with larger multi-model ensembles find that this effect is not robust ( [[#Steptoe--2016|Steptoe et al., 2016]] ; [[#Choi--2019|Choi et al., 2019]] ), consistent with CMIP6 single forcing ensembles (Figure 3.34). In the CMIP5 simulations, volcanic stratospheric aerosol has a significant weakening effect on the SAM in autumn and winter (Cross-Chapter Box 4.1; [[#Gillett--2013|Gillett and Fyfe, 2013]] ), but there is no evidence that this effect leads to a significant multi-decadal trend since the late 20th century. Beyond external forcing, [[#Fogt--2017|Fogt et al. (2017)]] show a significant association of tropical SST variability with the summertime SAM trend since the mid-20th century in agreement with [[#Lim--2016|Lim et al. (2016)]] , who, however, demonstrate that such a teleconnection between the summertime SAM and El Niño–Southern Oscillation (Annex IV.2.3), found in observations, is missing in many CMIP5 models.

On longer time scales, last millennium experiments from CMIP5 models fail to capture multicentennial variability evident in the reconstructions for the pre-industrial era ( [[#Abram--2014|Abram et al., 2014]] ; [[#Dätwyler--2018|Dätwyler et al., 2018]] ), which is also the case in those from available CMIP6 models (Figure 3.35). However, there is large uncertainty among reconstructions ( [[IPCC:Wg1:Chapter:Chapter-2#2.4.1.2|Section 2.4.1.2]] ). It is therefore unclear whether this disagreement reflects this observational uncertainty, whether forcings such as variations in the imposed insolation may be too weak, whether models are insufficiently sensitive to such variations, or whether internal variability including that associated with tropical Pacific variability is under-represented ( [[#Abram--2014|Abram et al., 2014]] ). The explanation could be a combination of all these factors. However, despite the aforementioned limitations of the reconstructions, [[IPCC:Wg1:Chapter:Chapter-2#2.4.1.2|Section 2.4.1.2]] assesses that the recent positive trend in the SAM is ''likely'' unprecedented in at least the past millennium ( ''medium confidence'' ). CMIP5 and CMIP6 last-millennium simulations only capture the present anomalous state during the final decades of the simulations which are dominated by human influence; this state is also outside the range of simulated variability characteristic of pre-industrial times.

<div id="_idContainer080" class="_idGenObjectStyleOverride-1"></div>

[[File:2a303798adcb7e1651d072fc363abf1f IPCC_AR6_WGI_Figure_3_35.png]]

Figure 3.35 | '''Southern Annular Mode (SAM) indices in the last millennium. (a)''' Annual-mean SAM reconstructions by [[#Abram--2014|Abram et al. (2014)]] and [[#Dätwyler--2018|Dätwyler et al. (2018)]] . '''(b)''' The annual-mean SAM index defined by [[#Gong--1999|Gong and Wang (1999)]] in CMIP5 and CMIP6 last millennium simulations extended by historical simulations. All indices are normalized with respect to 1961–1990 means and standard deviations. Thin lines and thick lines show seven-year and 70-year moving averages, respectively. Further details on data sources and processing are available in the chapter data table (Table 3.SM.1).

In summary, it is ''very likely'' that anthropogenic forcings have contributed to the observed trend of the summer SAM toward its positive phase since the 1970s. This assessment is supported by further model studies that confirm the human influence on the summertime SAM with improved models since AR5. While ozone depletion contributed to the trend from the 1970s to the 1990s ( ''medium confidence'' ), its influence has been small since 2000, leading to a weaker summertime SAM trend over 2000–2019 ( ''medium confidence'' ). Climate models reproduce the spatial structure of the summertime SAM observed since the late 1970s well ( ''high confidence'' ). CMIP6 models reproduce the spatiotemporal features and recent multi-decadal trend of the summertime SAM better than CMIP5 models ( ''medium confidence'' ). However, there is a large spread in the intensity of the SAM response to ozone and greenhouse gas changes in both CMIP5 and CMIP6 models ( ''high confidence'' ), which limits the confidence in the assessment of the ozone contribution to the observed trends. CMIP5 and CMIP6 models do not capture multicentennial variability of the SAM found in proxy reconstructions ( ''low confidence'' ). This confidence level reflects that it is unclear whether this is due to a model or an observational shortcoming.

<div id="3.7.3" class="h2-container"></div>

<span id="el-niñosouthern-oscillation-enso"></span>
=== 3.7.3 El Niño–Southern Oscillation (ENSO) ===

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

The El Niño–Southern Oscillation (ENSO), which is generated via seasonally modulated interactions between the tropical Pacific ocean and atmosphere, influences severe weather, rainfall, river flow and agricultural production over large parts of the world ( [[#McPhaden--2006|McPhaden et al., 2006]] ). In fact, the remote climate influence of ENSO is so large that knowledge of its current phase and forecasts of its future phase largely underpin many seasonal rainfall and temperature forecasts worldwide (Annex IV.2.3).

AR5 noted that there have been clear improvements in the simulation of ENSO through previous generations of CMIP models ( [[#Flato--2013|Flato et al., 2013]] ), such that many CMIP5 models displayed behaviour that was qualitatively similar to that of the observed ENSO ( [[#Guilyardi--2012|Guilyardi et al., 2012]] ). However, systematic errors were identified in the models’ representation of the tropical Pacific mean state and aspects of their interannual variability that affect quantitative comparisons. The AR5 assessment of ENSO concluded that the considerable observed inter-decadal modulations in ENSO amplitude and spatial pattern were largely consistent with unforced model simulations. Thus, there was ''low confidence'' in the role of a human-induced influence in these ( [[#Bindoff--2013|Bindoff et al., 2013]] ).

Observed ENSO amplitude, which is measured by the standard deviation of SST anomalies in a central equatorial Pacific region often referred to as the Nino 3.4 region, along with the lifecycle of events, are both reasonably well reproduced by most CMIP5 and CMIP6 models (Figure 3.36; [[#Bellenger--2014|Bellenger et al., 2014]] ; [[#Planton--2021|Planton et al., 2021]] ). The average CMIP5 model ENSO amplitude is slightly lower than that observed, while the average CMIP6 model ENSO amplitude is slightly higher than observed (Figure 3.36). The ENSO amplitude of the individual models, however, is highly variable across CMIP5 and CMIP6 models with many displaying either more or less variability than observed ( [[#Stevenson--2012|Stevenson, 2012]] ; [[#Grose--2020|Grose et al., 2020]] ; [[#Planton--2021|Planton et al., 2021]] ).

<div id="_idContainer082" class="•-2-columns"></div>

[[File:58c66b95fe275cdc736db332e2733be7 IPCC_AR6_WGI_Figure_3_36.png]]

Figure 3.36 | '''Life cycle of (left) El Niño and (right) La Niña events in observations (black) and historical simulations from CMIP5 (blue; extended with RCP4.5) and CMIP6 (red).''' An event is detected when the December ENSO index value in year zero exceeds 0.75 times its standard deviation for 1951–2010. '''(a, b)''' Composites of the ENSO index (°C). The horizontal axis represents month relative to the reference December (the grey vertical bar), with numbers in parentheses indicating relative years. Shading and lines represent 5th–95th percentiles and multi-model ensemble means, respectively. '''(c, d)''' Mean durations (months) of El Niño and La Niña events defined as number of months in individual events for which the ENSO index exceeds 0.5 times its December standard deviation. Each dot represents an ensemble member from the model indicated on the vertical axis. The boxes and whiskers represent multi-model ensemble means, interquartile ranges and 5th and 95th percentiles of CMIP5 and CMIP6. The CMIP5 and CMIP6 multi-model ensemble means and observational values are indicated at the top right of each panel. The multi-model ensemble means and percentile values are evaluated after weighting individual members with the inverse of the ensemble size of the same model, so that individual models are equally weighted irrespective of their ensemble sizes. The ENSO index is defined as the SST anomaly averaged over the Niño 3.4 region (5°S–5°N, 170°W–120°W). All results are based on five-month running mean SST anomalies with triangular-weights after linear detrending. Further details on data sources and processing are available in the chapter data table (Table 3.SM.1).

ENSO events are often synchronized to the seasonal cycle in the observations, as the associated SST anomalies tend to peak in boreal winter (November to January) and be at their weakest in the boreal spring (March to April) ( [[#Harrison--1998|Harrison and Larkin, 1998]] ; [[#Larkin--2002|Larkin and Harrison, 2002]] ). The majority of CMIP5 and CMIP6 models broadly reproduce the seasonality of ENSO SST variability in the central equatorial Pacific ( [[#Taschetto--2014|Taschetto et al., 2014]] ; [[#Abellán--2017|Abellán et al., 2017]] ; [[#Grose--2020|Grose et al., 2020]] ; [[#Planton--2021|Planton et al., 2021]] ) (Figure 3.37). However, CMIP5 models, while displaying an improvement on CMIP3 models, appear to under-represent the magnitude of the seasonal variance modulation ( [[#Bellenger--2014|Bellenger et al., 2014]] ). This under-representation of seasonal variance modulation continues in CMIP6 models, which display no statistically significant difference in this behaviour when compared to CMIP5 models ( [[#Planton--2021|Planton et al., 2021]] ) (Figure 3.37).

<div id="_idContainer084" class="•-2-columns"></div>

[[File:b6358ab26bb5a3f75228d78f8ba31f08 IPCC_AR6_WGI_Figure_3_37.png]]

Figure 3.37 | '''ENSO seasonality in observations (black) and historical simulations from CMIP5 (blue; extended with RCP4.5) and CMIP6 (red) for''' '''1951–2010''' '''. (a)''' Climatological standard deviation of the monthly ENSO index (SST anomaly averaged over the Niño 3.4 region; °C). Shading and lines represent 5th–95th percentiles and multi-model ensemble means, respectively. '''(b)''' Seasonality metric, which is defined for each model and each ensemble member as the ratio of the ENSO index climatological standard deviation in November–January (NDJ) to that in March–May (MAM). Each dot represents an ensemble member from the model indicated on the vertical axis. The boxes and whiskers represent the multi-model ensemble means, interquartile ranges and 5th and 95th percentiles of CMIP5 and CMIP6 individually. The CMIP5 and CMIP6 multi-model ensemble means and observational values are indicated at the top right of the panel. The multi-model ensemble means and percentile values are evaluated after weighting individual members with the inverse of the ensemble size of the same model, so that individual models are equally weighted irrespective of their ensemble sizes. All results are based on five-month running mean SST anomalies with triangular-weights after linear detrending. Further details on data sources and processing are available in the chapter data table (Table 3.SM.1).

Observations show strong multi-decadal modulation of ENSO variance throughout the 20th century, with the most recent period displaying larger variability while the mid-century displayed relatively low ENSO variability (Figure 2.36; [[#Li--2013|Li et al., 2013]] ; [[#McGregor--2013|McGregor et al., 2013]] ; [[#Hope--2017|Hope et al., 2017]] ). As assessed in [[IPCC:Wg1:Chapter:Chapter-2#2.4.2|Section 2.4.2]] , ENSO amplitude since 1950 is higher than over the pre-industrial period from 1850 as far back as 1400 ( ''medium confidence'' ), but there is ''low confidence'' that it is higher than the variability over periods prior to 1400. This reported variance increase suggests that external forcing plays a role in the ENSO variance changes ( [[#Hope--2017|Hope et al., 2017]] ). However, large ensembles of single model or multiple model simulations do not find strong trends in ENSO variability over the historical period, suggesting that external forcing has not yet modulated ENSO variability with a magnitude that exceeds the range of internal variability ( [[#Hope--2017|Hope et al., 2017]] ; [[#Maher--2018b|Maher et al., 2018b]] ; [[#Stevenson--2019|Stevenson et al., 2019]] ). This is consistent with the [[IPCC:Wg1:Chapter:Chapter-2|Chapter 2]] assessment that there is no clear evidence for a recent sustained shift in ENSO beyond the range of variability on decadal to millennial time scales ( [[IPCC:Wg1:Chapter:Chapter-2#2.4.2|Section 2.4.2]] ). CMIP5 and CMIP6 models show a decrease in ENSO variance in the mid-Holocene ( [[#Brown--2020|Brown et al., 2020]] ), though not to the extent seen in paleo-proxy records ( [[#Emile-Geay--2016|Emile-Geay et al., 2016]] ). This suggests that both modelled and observed ENSO respond to changes in external forcing, but not necessarily in the same manner.

Most CMIP5 and CMIP6 models are found to represent the general structure of observed SST anomalies during ENSO events well ( [[#Kim--2012|Kim and Yu, 2012]] ; [[#Taschetto--2014|Taschetto et al., 2014]] ; [[#Brown--2020|Brown et al., 2020]] ; [[#Grose--2020|Grose et al., 2020]] ). However, the majority of CMIP5 models display SST anomalies that: i) extend too far to the west ( [[#Taschetto--2014|Taschetto et al., 2014]] ; [[#Capotondi--2015|Capotondi et al., 2015]] ); and ii) have meridional widths that are too narrow ( [[#Zhang--2012|Zhang and Jin, 2012]] ) compared to the observations. CMIP6 models display a statistically significant improvement in the longitudinal representation of ENSO SST anomalies relative to CMIP5 models ( [[#Planton--2021|Planton et al., 2021]] ), however, systematic biases in the zonal extent and meridional width remain in CMIP6 models ( [[#Fasullo--2020|Fasullo et al., 2020]] ; [[#Planton--2021|Planton et al., 2021]] ). The ENSO phase asymmetry, where observed strong El Niño events are larger and have a shorter duration than strong La Niña events ( [[#Ohba--2009|Ohba and Ueda, 2009]] ; [[#Frauen--2010|Frauen and Dommenget, 2010]] ), is also under-represented in both CMIP5 and CMIP6 models ( [[#Zhang--2014|Zhang and Sun, 2014]] ; [[#Planton--2021|Planton et al., 2021]] ). In this instance, both CMIP5 and CMIP6 models typically display El Niño events that have a longer duration than those observed, La Niña events that have a similar duration to those observed, and there is very little asymmetry in the duration of El Niño and La Niña phases (Figure 3.36). [[#Roberts--2018|Roberts et al. (2018)]] find an improvement in amplitude asymmetry in a HighResMIP model, but the under-representation remains.

The continuum of El Niño events are typically stratified into two types (often termed ‘flavours’), Central Pacific and East Pacific, where the name denotes the location of the events’ largest SST anomalies (Annex IV.2.3; [[#Capotondi--2015|Capotondi et al., 2015]] ). As discussed in [[IPCC:Wg1:Chapter:Chapter-2#2.4.2|Section 2.4.2]] , the different types of events tend to produce distinct teleconnections and climatic impacts (e.g., [[#Taschetto--2020|Taschetto et al., 2020]] ). The characteristics of El Niño events of these two flavours in CMIP5 were generally comparable to the observations ( [[#Taschetto--2014|Taschetto et al., 2014]] ). CMIP6 models, however, display a statistically significant improvement in the representation of this ENSO event-to-event SST anomaly diversity when compared with CMIP5 models ( [[#Planton--2021|Planton et al., 2021]] ). In addition to this ENSO event diversity, the short observational record also displays an increase in the number of the Central Pacific-type events in recent decades ( [[#Ashok--2007|Ashok et al., 2007]] ; [[#McPhaden--2011|McPhaden et al., 2011]] ), which has also been identified as unusual in the context of the last 500–800 years based on recent paleo-climatic reconstructions ( [[IPCC:Wg1:Chapter:Chapter-2#2.4.2|Section 2.4.2]] ; Y. [[#Liu--2017|]] [[#Liu--2017|Liu et al., 2017]] ; [[#Freund--2019|Freund et al., 2019]] ). However, the short observational record combined with observational ( [[#L’Heureux--2013|L’Heureux et al., 2013]] ) and paleo-climatic reconstruction uncertainties preclude firm conclusions being made about the long-term changes in the occurrence of different El Niño event types. Initial analysis with a selected number of CMIP3 models suggested that there may be a forced component to this recent prominence of Central Pacific-type events ( [[#Yeh--2009|Yeh et al., 2009]] ), but analysis since then suggests that this behaviour is (i) consistent with that expected from internal variability ( [[#Newman--2011|Newman et al., 2011]] ); and (ii) not apparent across the full CMIP5 ensemble of historical simulations ( [[#Taschetto--2014|Taschetto et al., 2014]] ). Analysis of single-model large ensembles suggests that changes to ENSO event type in response to historical radiative forcing are not significant (e.g., [[#Stevenson--2019|Stevenson et al., 2019]] ). These same results, however, also suggest that multiple forcings can have significant influences on ENSO type and that the net response will depend on the accurate representation of the balance of these forcings ( [[#Stevenson--2019|Stevenson et al., 2019]] ).

The climatic effects of ENSO outside the tropical Pacific largely arise through atmospheric teleconnections that are induced by ENSO-driven changes in deep tropical atmospheric convection and heating ( [[#Yeh--2018|Yeh et al., 2018]] ). The teleconnections to higher latitudes are forced by waves that propagate into the extratropics ( [[#Hoskins--1981|Hoskins and Karoly, 1981]] ) and respectively excite the Pacific-North American pattern ( [[#Horel--1981|Horel and Wallace, 1981]] ) and Pacific-South American pattern ( [[#Karoly--1989|Karoly, 1989]] ; [[#Irving--2016|Irving and Simmonds, 2016]] ) in the Northern and Southern Hemispheres. Given the influence of these teleconnections on climate and extremes around the globe, it is important to understand how well they are reproduced in CMIP models. What has also become clear is that spatial correlations of ENSO’s teleconnections calculated over relatively short periods (&lt;100 years) may not be the most effective way to assess these relationships ( [[#Langenbrunner--2013|Langenbrunner and Neelin, 2013]] ; [[#Perry--2020|Perry et al., 2020]] ). This is because the spatial patterns are significantly affected by internal atmospheric variability on relatively short time scales (&lt;100 years; [[#Batehup--2015|Batehup et al., 2015]] ; [[#Perry--2020|Perry et al., 2020]] ). However, looking at simplified metrics like the agreement in the sign of the teleconnections ( [[#Langenbrunner--2013|Langenbrunner and Neelin, 2013]] ), regional average teleconnection strength over land ( [[#Perry--2020|Perry et al., 2020]] ), or a combination of both ( [[#Power--2018|Power and Delage, 2018]] ) provides a more robust depiction of the teleconnection representation. Examining sign agreement for the teleconnection patterns, ensembles of CMIP5 AMIP simulations display broad spatial regions with high sign agreement with the observations, suggesting that the model ensemble is producing useful information regarding the teleconnected precipitation signal ( [[#Langenbrunner--2013|Langenbrunner and Neelin, 2013]] ). Looking at regional averages of CMIP5 historical simulations, [[#Power--2018|Power and Delage (2018)]] show that the average coupled model teleconnection pattern reproduces the sign of the observed teleconnections in the majority of the 25 regions analysed. The sign agreement between the observed teleconnection and the multi-model mean teleconnection remains strong in CMIP6 (18 out of 20 displayed regions; Figure 3.38), and the observed DJF (December–January–February) teleconnection strength falls within the modelled range in all of the displayed regions for temperature and precipitation. Note, however, that while there is broad agreement in ENSO teleconnections between CMIP6 models and observations during DJF (e.g., [[#Fasullo--2020|Fasullo et al., 2020]] ), there are regions and seasons where the modelled teleconnection strength is outside the observed range ( [[#Chen--2020|Chen et al., 2020]] ).

<div id="_idContainer086" class="•-2-columns"></div>

[[File:101ddaa26d08600548b1cfc9dade867d IPCC_AR6_WGI_Figure_3_38.png]]

Figure 3.38 | '''Model evaluation of ENSO teleconnection for near surface air temperature and precipitation in boreal winter (December–January–February).''' Teleconnections are identified by linear regression with the Niño 3.4 SST index based on Extended Reconstructed Sea Surface Temperature (ERSST) version 5 over the period 1958–2014. Maps show observed patterns for temperature from the Berkeley Earth dataset over land and from ERSST version 5 over ocean (°C, '''top''' ) and for precipitation from GPCC over land (shading, mm day <sup>–</sup> <sup>1</sup> ) and GPCP worldwide (contours, period: 1979–2014). Distributions of regression coefficients (grey histograms) are provided for a subset of AR6 reference regions defined in Atlas.1.3 for temperature '''(top)''' and precipitation '''(bottom)''' . All fields are linearly detrended prior to computation. Multi-model multi-member ensemble means are indicated by thick vertical black lines. Blue vertical lines show three observational estimates of temperature, based on Berkeley Earth, GISTEMP and CRUTS datasets, and two observational estimates of precipitation, based on GPCC and CRUTS datasets. Further details on data sources and processing are available in the chapter data table (Table 3.SM.1).

Most CMIP5 and CMIP6 models exhibit ENSO behaviour during the historical period that, to first order, is qualitatively similar to that of the observed ENSO. Many studies are now delving deeper into the models to understand if they are accurately producing the dynamics driving ENSO and its initiation ( [[#Jin--2006|Jin et al., 2006]] ; [[#Bellenger--2014|Bellenger et al., 2014]] ; [[#Vijayeta--2018|Vijayeta and Dommenget, 2018]] ; [[#Bayr--2019|Bayr et al., 2019]] ; [[#Planton--2021|Planton et al., 2021]] ). For both CMIP3 and CMIP5, diagnostics of ENSO event growth appear to show that the models, while producing ENSO variability that is qualitatively similar to that observed, do not represent the balance of the underlying dynamics well. The atmospheric Bjerknes feedback is too weak in the majority of models, while the surface heat flux feedback is also too weak in the majority of models. The former restricts event growth, while the latter restricts event damping, which when combined allow most models to produce variability in a range that is consistent with the observations ( [[#Bellenger--2014|Bellenger et al., 2014]] ; S.T. [[#Kim--2014|]] [[#Kim--2014|Kim et al., 2014]] ; [[#Vijayeta--2018|Vijayeta and Dommenget, 2018]] ; [[#Bayr--2019|Bayr et al., 2019]] ). Analysis of ENSO representation in a subset of CMIP6 models by [[#Planton--2021|Planton et al. (2021)]] suggests that these issues remain.

To conclude, ENSO representation in CMIP5 models displayed a significant improvement from the representation of ENSO variability in CMIP3 models, which displayed much more intermodel spread in standard deviation, and stronger biennial periodicity ( [[#Guilyardi--2012|Guilyardi et al., 2012]] ; [[#Flato--2013|Flato et al., 2013]] ). In general, there has been no large step change in the representation of ENSO between CMIP5 and CMIP6, however, CMIP6 models appear to better represent some key ENSO characteristics (e.g., [[#Brown--2020|Brown et al., 2020]] ; [[#Planton--2021|Planton et al., 2021]] ). The instrumental record and paleo-proxy evidence through the Holocene all suggest that ENSO can display considerable modulations in amplitude, pattern and period (see also ( [[IPCC:Wg1:Chapter:Chapter-2#2.4.2|Section 2.4.2]] ). For the period since 1850, there is no clear evidence for a sustained shift in ENSO index beyond the range of internal variability. However, paleo-proxy evidence indicates with ''medium confidence'' that ENSO variability since 1950 is greater than at any time between 1400 and 1850 ( [[IPCC:Wg1:Chapter:Chapter-2#2.4.2|Section 2.4.2]] ). Coupled models display large changes of ENSO behaviour in the absence of external forcing changes, and little-to-no variance sensitivity to historical anthropogenic forcing. Thus, there is ''low confidence'' that anthropogenic forcing has led to the changes of ENSO variability inferred from paleo-proxy evidence.

[[IPCC:Wg1:Chapter:Chapter-2|Chapter 2]] reports ''low confidence'' that the apparent change from East Pacific- to Central Pacific-type El Niño events that occurred in the last 20–30 years was representative of a long term change. While some climate models do suggest external forcing may affect the El Niño event type, most climate models suggest that what has been observed is well within the range of natural variability. Thus, there is ''low confidence'' that anthropogenic forcing has had an influence on the observed changes in El Niño event type.

<div id="3.7.4" class="h2-container"></div>

<span id="indian-ocean-basin-and-dipole-modes"></span>
=== 3.7.4 Indian Ocean Basin and Dipole Modes ===

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

The Indian Ocean Basin (IOB) and Dipole (IOD) modes are the two leading modes of interannual SST variability over the tropical Indian Ocean, featuring basin-wide warming/cooling and an east–west dipole of SST anomalies, respectively (Annex IV.2.4). The IOD mode is anchored to boreal summer to autumn by the air–sea feedback, and often develops in concert with ENSO. Driven by matured ENSO events, the IOB mode peaks in boreal spring and often persists into the subsequent summer. Similar patterns of Indian Ocean SST variability also dominate its decadal and longer time scale variability ( [[#Han--2014b|Han et al., 2014b]] ).

AR5 concluded that models show high and medium performance in reproducing the IOB and IOD modes, respectively ( ''medium confidence'' ), with difficulty in reproducing the persistence of the IOB and the pattern and magnitude of the IOD ( [[#Flato--2013|Flato et al., 2013]] ). There was ''low confidence'' that changes in the IOD were detectable or attributable to human influence ( [[#Bindoff--2013|Bindoff et al., 2013]] ).

Since AR5, CMIP5 model representation of these modes has been analysed in detail, finding that most of the models qualitatively reproduce the spatial and seasonal features of the IOB and IOD modes ( [[#Chu--2014|Chu et al., 2014]] ; [[#Liu--2014|Liu et al., 2014]] ; W. [[#Tao--2016|]] [[#Tao--2016|Tao et al., 2016]] ). Improvements in simulating the IOB mode since CMIP3 have been identified in reduced multi-model mean biases and inter-model spread (W. [[#Tao--2016|]] [[#Tao--2016|Tao et al., 2016]] ). CMIP5 models overall capture the transition from the IOD to IOB modes during an ENSO event (W. [[#Tao--2016|]] [[#Tao--2016|Tao et al., 2016]] ). The IOB mode is forced in part through a cross-equatorial wind–evaporation–SST feedback triggered by ENSO-forced anomalous ocean Rossby waves that propagate to the shallow climatological thermocline dome in the tropical south-western Indian Ocean ( [[#Du--2009|Du et al., 2009]] ). Consistently, models with a deeper climatological thermocline dome produce a weaker and less persistent IOB mode (G. [[#Li--2015|Li et al., 2015]] a; [[#Zheng--2016|Zheng et al., 2016]] ). The deep thermocline bias remains in the ensemble mean of CMIP5 models due to a common surface easterly wind bias over the equatorial Indian Ocean ( [[#Lee--2013|Lee et al., 2013]] ) associated with weaker South Asian summer monsoon circulation (G. [[#Li--2015|Li et al., 2015]] b). However, the influence of this systematic bias may be compensated by other biases, resulting in a realistic IOB magnitude (W. [[#Tao--2016|]] [[#Tao--2016|Tao et al., 2016]] ). [[#Halder--2021|Halder et al. (2021)]] found that CMIP6 models reproduce the IOB mode reasonably well, but did not evaluate the progress since CMIP5.

By contrast, the IOD magnitude is overestimated by CMIP5 models on average, though with noticeable improvements from CMIP3 models ( [[#Liu--2014|Liu et al., 2014]] ). The overestimation of the IOD magnitude remains in most of 34 CMIP6 models examined in [[#McKenna--2020|McKenna et al. (2020)]] with worsening on average in July and August. A too steep climatological thermocline slope along the equator due to the surface easterly wind bias in boreal summer and autumn contributes to this IOD magnitude bias through an excessively strong Bjerknes feedback in CMIP5 ( [[#Liu--2014|Liu et al., 2014]] ; G. [[#Li--2015|Li et al., 2015]] b; [[#Hirons--2018|Hirons and Turner, 2018]] ). The surface easterly bias and associated east–west SST gradient bias are not improved in CMIP6 ( [[#Long--2020|Long et al., 2020]] ; [[#3.5.1.2.3|Section 3.5.1.2.3]] ), suggesting that the thermocline bias also remains. [[#McKenna--2020|McKenna et al. (2020)]] additionally find degradation in the positive-negative asymmetry of the IOD but an improvement in IOD frequency in a subset of CMIP6 models compared to CMIP5. In terms of teleconnections, the equatorial surface easterly wind bias also affects the IOD-associated moisture transport anomalies toward tropical eastern Africa ( [[#Hirons--2018|Hirons and Turner, 2018]] ) where the IOD is associated with strong precipitation anomalies in boreal autumn (Annex IV.2.4). CMIP5 and CMIP6 models capture the IOD teleconnection to Southern and Central Australian precipitation although it is weaker on average than observed, with no clear improvements from CMIP5 to CMIP6 ( [[#Grose--2020|Grose et al., 2020]] ). Strong IOD events could also influence the Northern Hemisphere extratropical circulation in winter and in particular the NAM ( [[#3.7.1|Section 3.7.1]] ), based on interference between forced Rossby waves emerging from the Indian Ocean and climatological stationary waves ( [[#Fletcher--2015|Fletcher and Cassou, 2015]] ). The record positive phase of the NAO/NAM in winter 2019–2020 assessed over the instrumental era has been accordingly linked to the record IOD event of autumn 2019 ( [[#Hardiman--2020|Hardiman et al., 2020]] ), which has been associated with the devastating record fire season in Australia ( [[#Wang--2020|Wang and Cai, 2020]] ).

The observed Indian Ocean basin-average SST increase on multi-decadal and centennial time scales is well represented by CMIP5 historical simulations, and has been attributed to the effects of greenhouse gases offset in part by the effects of anthropogenic aerosols mainly through aerosol-cloud interactions ( [[#Dong--2014|Dong and Zhou, 2014]] ; [[#Dong--2014b|Dong et al., 2014b]] ). The observed SST trend is larger in the western than eastern tropical Indian Ocean, which leads to an apparent upward trend of the IOD index, but this trend is statistically insignificant ( [[IPCC:Wg1:Chapter:Chapter-2#2.4.3|Section 2.4.3]] ). CMIP5 models capture this warming pattern, which may be associated with Walker circulation weakening over the Indian Ocean due to greenhouse gas forcing ( [[#Dong--2014|Dong and Zhou, 2014]] ). However, strong internal decadal IOD-like variability and observational uncertainty preclude attribution ( [[#Cai--2013|Cai et al., 2013]] ; [[#Han--2014b|Han et al., 2014b]] ; [[#Gopika--2020|Gopika et al., 2020]] ). Such a positive IOD-like change in equatorial zonal SST gradient suggests an increase in the frequency of extreme positive events ( [[#Cai--2014|Cai et al., 2014]] ) and skewness ( [[#Cowan--2015|Cowan et al., 2015]] ) of the IOD mode. While there is some evidence of an increase in frequency of positive IOD events during the second half of the 20th century, the current level of IOD variability is not unprecedented in a proxy reconstruction for the last millennium ( [[IPCC:Wg1:Chapter:Chapter-2#2.4.3|Section 2.4.3]] ; [[#Abram--2020|Abram et al., 2020]] ). Besides, the IOD magnitude in the late 20th century is not significantly different between CMIP5 simulations forced by historical and natural-only forcings, though this conclusion is based on only five selected ensemble members that realistically reproduce statistical features of the IOD ( [[#Blau--2020|Blau and Ha, 2020]] ). While selected CMIP5 models show weakening ( [[#Thielke--2019|Thielke and Mölg, 2019]] ) and seasonality changes ( [[#Blau--2020|Blau and Ha, 2020]] ) in IOD-induced rainfall anomalies in tropical eastern Africa, no comparison with observational records has been made. Likewise, while a strengthening tendency of the ENSO-IOB mode correlation and resultant intensification of the IOB mode are found in historical or future simulations in selected CMIP5 models ( [[#Hu--2014|Hu et al., 2014]] ; [[#Tao--2015|Tao et al., 2015]] ), such a change has not been detected in observational records.

After linear detrending, Pacific Decadal Variability (PDV; Annex IV.2.6; [[#3.7.6|Section 3.7.6]] ) has been suggested as a driver of decadal to multi-decadal variations in the IOB mode ( [[#Dong--2016|Dong et al., 2016]] ). However, correlation between the PDV and a decadal IOB index, defined from linearly detrended SST, changed from positive to negative during the 1980s ( [[#Han--2014a|Han et al., 2014a]] ). The increase in anthropogenic forcing and recovery from the eruptions of El Chichón in 1982 and Pinatubo in 1991 may have overwhelmed the PDV influence, and explain this change ( [[#Dong--2017|Dong and McPhaden, 2017]] ; L. [[#Zhang--2018a|]] [[#Zhang--2018|Zhang et al., 2018]] a ). However, the low statistical degrees of freedom hamper clear detection of human influence in this correlation change.

To summarize, there is ''medium confidence'' that changes in the interannual IOD variability in the late 20th century inferred from observations and proxy records are within the range of internal variability. There is no evidence of anthropogenic influence on the interannual IOB. On decadal- to multi-decadal time scales, there is ''low confidence'' that human influence has caused a reversal of the correlation between PDV and decadal variations in the IOB mode. The ''low confidence'' in this assessment is due to the short observational record, limited number of models used for the attribution, lack of model evaluation of the decadal IOB mode, and uncertainty in the contribution from volcanic aerosols. Nevertheless, CMIP5 models have medium overall performance in reproducing both the interannual IOB and IOD modes, with an apparently good performance in reproducing the IOB magnitude arising from compensation of biases in the formation process, and overly high IOD magnitude due to the mean state bias ( ''high confidence'' ). There is no clear improvement in the simulation of the IOD from CMIP5 to CMIP6 models, though there is only ''medium'' ''confidence'' in this assessment, since only a subset of CMIP6 models have been examined. There is no evidence for performance changes in simulating the IOB from CMIP5 to CMIP6 models.

<div id="3.7.5" class="h2-container"></div>

<span id="atlantic-meridional-and-zonal-modes"></span>
=== 3.7.5 Atlantic Meridional and Zonal Modes ===

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

The Atlantic Zonal Mode (AZM), often referred to as the Atlantic Equatorial Mode or Atlantic Niño, and the Atlantic Meridional Mode (AMM) are the two leading basin-wide patterns of interannual to decadal variability in the tropical Atlantic. Akin to ENSO in the Pacific, the term Atlantic Niño is broadly used to refer to years when the SSTs in the tropical eastern Atlantic basin along the cold tongue are significantly warmer than the climatological average. The AMM is characterized by anomalous cross-equatorial gradients in SST. Both modes are associated with altered strength of the Inter-tropical Convergence Zone (ITCZ) and/or latitudinal shifts in the ITCZ, which locally affect African and American monsoon systems and remotely affect tropical Pacific and Indian Ocean variability through inter-basins teleconnections. A detailed description of both AZM and AMM, as well as their associated teleconnection over land, is given in Annex IV.2.5

AR5 mentioned the considerable difficulty in simulating both Atlantic Niño and AMM despite some improvements in CMIP5 for some models ( [[#Flato--2013|Flato et al., 2013]] ). Severe biases in mean state and variance for both SST and atmospheric dynamics including rainfall (e.g., a double ITCZ) as well as teleconnections were reported. The AR5 highlighted the complexity of the tropical Atlantic biases, which were explained by multiple factors both in the ocean and atmosphere.

Since AR5, further analysis of the major persistent biases in models has been reported ( [[#Xu--2014|Xu et al., 2014]] ; [[#Jouanno--2017|Jouanno et al., 2017]] ; Y. [[#Yang--2017|]] [[#Yang--2017|Yang et al., 2017]] ; [[#Dippe--2018|Dippe et al., 2018]] ; [[#Lübbecke--2018|Lübbecke et al., 2018]] ; [[#Voldoire--2019a|Voldoire et al., 2019a]] ). Errors in equatorial and basin wide trade winds, cloud cover and ocean vertical mixing and dynamics both locally and in remote subtropical upwelling regions, key thermodynamic ocean–atmosphere feedbacks, and tropical land–atmosphere interaction have been shown to be detrimental to the representation of both the Atlantic Niño and AMM leading to poor teleconnectivity over land ( [[#Rodríguez-Fonseca--2015|Rodríguez-Fonseca et al., 2015]] ; [[#Wainwright--2019|Wainwright et al., 2019]] ) and between tropical basins ( [[#Ott--2015|Ott et al., 2015]] ).

Despite some improvements ( [[#Richter--2014|Richter et al., 2014]] ; [[#Nnamchi--2015|Nnamchi et al., 2015]] ), biases in the mean state are so large that the mean east–west temperature gradient at the equator along the thermocline remains opposite to observed in two thirds of the CMIP5 models ( [[#3.5.1.2.2|Section 3.5.1.2.2]] ), which clearly affects the simulation of the Atlantic Niño and associated dynamics ( [[#Muñoz--2012|Muñoz et al., 2012]] ; [[#Ding--2015|Ding et al., 2015]] ; [[#Deppenmeier--2016|Deppenmeier et al., 2016]] ). The interhemispheric SST gradient is also systematically underestimated in models, with a too cold mean state in the northern part of the tropical Atlantic ocean and too warm conditions in the South Atlantic basin. The seasonality is poorly reproduced and the wind–SST coupling is weaker than observed so that altogether, and despite AMM-like variability in 20th century climate simulations, AMM is not the dominant Atlantic mode in all CMIP5 models ( [[#Liu--2013|Liu et al., 2013]] ; [[#Amaya--2017|Amaya et al., 2017]] ). These biases in mean state translate into biases in modelling the mean ITCZ ( [[#Flato--2013|Flato et al., 2013]] ). Similar biases were found in experiments using CMIP5 models but with different climate background states, such as Last Glacial Maximum, mid-Holocene and future scenario simulations ( [[#Brierley--2018|Brierley and Wainer, 2018]] ). Analyses of CMIP6 show encouraging results in the representation of Atlantic Niño and AMM modes of variability in terms of amplitude and seasonality. Some models now display reduced biases in the spatial structure of the modes and related explained variance but persistent errors still remain on average in the timing of the modes and in the coupled nature of the modes, that is, the strength of the link between ocean (SST, mixed layer depth) and atmospheric (wind) anomalies ( [[#Richter--2020|Richter and Tokinaga, 2020]] ), as well as in the Atlantic Ocean equatorial east–west temperature gradient ( [[#3.5.1.2.2|Section 3.5.1.2.2]] , Figure 3.24).

There are some recent indications that increasing model resolution both vertically and horizontally, in the ocean and atmospheric component ( [[#Richter--2015|Richter, 2015]] ; [[#Small--2015|Small et al., 2015]] ; [[#Harlaß--2018|Harlaß et al., 2018]] ), could partly alleviate some tropical Atlantic biases in mean state ( [[#3.5.1.2.2|Section 3.5.1.2.2]] ), seasonality, interannual- to decadal-variability and associated teleconnectivity over land, such as with the West African monsoon ( [[#Steinig--2018|Steinig et al., 2018]] ). Results from CMIP6 tend to confirm that increasing resolution is not the unique way to address the biases in the tropical Atlantic ( [[#Richter--2020|Richter and Tokinaga, 2020]] ). For instance, the inclusion of a stochastic physics scheme has a nearly equivalent effect in the improvement of the mean number and the strength distribution of tropical Atlantic cyclones ( [[#Vidale--2021|Vidale et al., 2021]] ).

( [[IPCC:Wg1:Chapter:Chapter-2#2.4.4|Section 2.4.4]] assess that there is ''low confidence'' in any sustained changes to the AZM and AMM variability in instrumental observations. Moreover, any attribution of possible human influence on the Atlantic modes and associated teleconnections is limited by the poor fidelity of CMIP5 and CMIP6 models in reproducing the mean tropical Atlantic climate, its seasonality and variability, despite hints of some improvement in CMIP6, as well as other sources of uncertainties related to limited process understanding in the observations ( [[#Foltz--2019|Foltz et al., 2019]] ), the response of the tropical Atlantic climate to anthropogenic aerosol forcing ( [[#Booth--2012|Booth et al., 2012]] ; [[#Zhang--2013a|Zhang et al., 2013a]] ) and the presence of strong multi-decadal fluctuations related to AMV ( [[#3.7.7|Section 3.7.7]] ) and cross-tropical basin interactions ( [[#Martín-Rey--2018|Martín-Rey et al., 2018]] ; [[#Cai--2019|Cai et al., 2019]] ). The fact that most models poorly represent the climatology and variability of the tropical Atlantic combined with the short observational record makes it difficult to place the recent observed changes in the context of internal multi-annual variability versus anthropogenic forcing.

In summary, based on CMIP5 and CMIP6 results, there is no robust evidence that the observed changes in either the Atlantic Niño or AMM modes and associated teleconnections over the second half of the 20th century are beyond the range of internal variability or have been influenced by natural or anthropogenic forcing. Considering the physical processes responsible for model biases in these modes, increasing resolution in both ocean and atmosphere components may be an opportunity for progress in the simulation of the tropical Atlantic changes as evidenced by some individual model studies ( [[#Roberts--2018|Roberts et al., 2018]] ), but this needs confirmation from a multi-model perspective.

<div id="3.7.6" class="h2-container"></div>

<span id="pacific-decadal-variability"></span>
=== 3.7.6 Pacific Decadal Variability ===

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

Pacific Decadal Variability (PDV) is the generic term for the modes of variability in the Pacific Ocean that vary on decadal to inter-decadal time scales. PDV and its related teleconnections encompass the Pacific Decadal Oscillation (PDO; [[#Mantua--1997|Mantua et al., 1997]] ; [[#Zhang--1997|Zhang et al., 1997]] ; [[#Mantua--2002|Mantua and Hare, 2002]] ), and an anomalous SST pattern in the North Pacific, as well as a broader structure associated with Pacific-wide SSTs termed the Inter-decadal Pacific Oscillation (IPO; [[#Power--1999|Power et al., 1999]] ; [[#Folland--2002|Folland et al., 2002]] ; [[#Henley--2015|Henley et al., 2015]] ). Since the PDO and IPO indices are highly correlated, this section assesses them together as the PDV (Annex IV.2.6).

AR5 mentioned an overall ''limited'' level of evidence for both CMIP3 and CMIP5 evaluation of the Pacific modes at inter-decadal time scales, leading to ''low confidence'' statements about the models’ performance in reproducing PDV ( [[#Flato--2013|Flato et al., 2013]] ) and similarly ''low confidence'' in the attribution of observed PDV changes to human influence ( [[#Bindoff--2013|Bindoff et al., 2013]] ).

The implication of PDV in the observed slowdown of the GMST warming rate in the early 2000s (Cross-Chapter Box 3.1) has triggered considerable research on decadal climate variability and predictability since AR5 ( [[#Meehl--2013|Meehl et al., 2013]] , 2016b; [[#England--2014|England et al., 2014]] ; [[#Dai--2015|Dai et al., 2015]] ; [[#Steinman--2015|Steinman et al., 2015]] ; [[#Kosaka--2016|Kosaka and Xie, 2016]] ; [[#Cassou--2018|Cassou et al., 2018]] ). Many studies find that the broad spatial characteristics of PDV are reasonably well represented in unforced climate models ( [[#Newman--2016|Newman et al., 2016]] ; [[#Henley--2017|Henley, 2017]] ) and in historical simulations in CMIP5 and CMIP6 (Figure 3.39), although there is sensitivity to the methodology used to remove the externally-forced component of the SST ( [[#Bonfils--2011|Bonfils and Santer, 2011]] ; [[#Xu--2018|Xu and Hu, 2018]] ). Compared with CMIP3 models, CMIP5 models exhibit overall slightly better performance in reproducing PDV and associated teleconnections ( [[#Polade--2013|Polade et al., 2013]] ; [[#Joshi--2017|Joshi and Kucharski, 2017]] ), and also smaller inter-model spread ( [[#Lyu--2016|Lyu et al., 2016]] ). CMIP6 models on average show slightly improved reproduction of the PDV spatial structure than CMIP5 models (Figure 3.39a–c; [[#Fasullo--2020|Fasullo et al., 2020]] ). SST anomalies in the subtropical South Pacific lobe are, however, too weak relative to the equatorial and North Pacific lobes in CMIP5 pre-industrial control and historical simulations ( [[#Henley--2017|Henley et al., 2017]] ), a bias that remains in CMIP6 (Figure 3.39b).

Biases in the PDV temporal properties and amplitude are present in CMIP5 ( [[#Cheung--2017|Cheung et al., 2017]] ; [[#Henley--2017|Henley, 2017]] ). While model evaluation is severely hampered by short observational records and incomplete observational coverage before satellite measurements started, the duration of PDV phases appears to be shorter in coupled models than in observations, and correspondingly the ratio of decadal to interannual variance is underestimated (Figure 3.39e,f; [[#Henley--2017|Henley et al., 2017]] ). This apparent bias may be associated with overly biennial behaviour of Pacific trade wind variability and related ENSO activity, leaving too weak variability on decadal time scales ( [[#Kociuba--2015|Kociuba and Power, 2015]] ). ENSO influence on the extratropical North Pacific Ocean at decadal time scales is also very diverse among both CMIP3 and CMIP5 models, being controlled by multiple factors ( [[#Nidheesh--2017|Nidheesh et al., 2017]] ). In terms of amplitude, the variance of the PDV index after decadal filtering is significantly weaker in the concatenated CMIP5 ensemble than the three observational estimates used in Figure 3.39e ( ''p'' &lt;0.1 with an F-test). Consequently, the observed PDV fluctuations over the historical period often lie in the tails of the model distributions (Figure 3.39e,f). Even if one cannot rule out that the observed PDV over the instrumental era represents an exceptional period of variability, it is plausible that the tendency of the CMIP5 models to systematically underestimate the low frequency variance is due to an incomplete representation of decadal-scale mechanisms in these models. This situation is slightly improved in CMIP6 historical simulations but remains a concern ( [[#Fasullo--2020|Fasullo et al., 2020]] ). The results of [[#McGregor--2018|McGregor et al. (2018)]] suggest that the under-representation of the variability stems from Atlantic mean SST biases ( [[#3.5.1.2.2|Section 3.5.1.2.2]] ) through inter-basin coupling.

While PDV is primarily understood as an internal mode of variability ( [[#Si--2017|Si and Hu, 2017]] ), there are some indications that anthropogenically induced SST changes project onto PDV and have contributed to its past evolution ( [[#Bonfils--2011|Bonfils and Santer, 2011]] ; [[#Dong--2014a|Dong et al., 2014a]] ; [[#Boo--2015|Boo et al., 2015]] ; [[#Xu--2018|Xu and Hu, 2018]] ). However, the level of evidence is ''limited'' because of the difficulty in correctly separating internal versus externally forced components of the observed SST variations, and because it is unclear whether the dynamics of the PDV are operative in this forced SST change pattern. Over the last two to three decades which encompass the period of slower GMST increase (Cross-Chapter Box 3.1), [[#Smith--2016|Smith et al. (2016)]] found that anthropogenic aerosols have driven part of the PDV change toward its negative phase. A similar result is shown in [[#Takahashi--2016|Takahashi and Watanabe (2016)]] who found intensification of the Pacific Walker circulation in response to aerosol forcing ( [[#3.3.3.1.2|Section 3.3.3.1.2]] ). Indeed, CMIP6 models simulate a negative PDV trend since the 1980s (Figure 3.39f), which is much weaker than internal variability. However, a response to anthropogenic aerosols is not robustly identified in a large ensemble of a model ( [[#Oudar--2018|Oudar et al., 2018]] ), across CMIP5 models ( [[#Hua--2018|Hua et al., 2018]] ), or in idealized model simulations ( [[#Kuntz--2016|Kuntz and Schrag, 2016]] ). Alternatively, inter-basin teleconnections associated with the warming of the North Atlantic Ocean related to the mid-1990s phase shift of the AMV ( [[#McGregor--2014|McGregor et al., 2014]] ; [[#Chikamoto--2016|Chikamoto et al., 2016]] ; [[#Kucharski--2016|Kucharski et al., 2016]] ; X. [[#Li--2016|]] [[#Li--2016|Li et al., 2016]] a; [[#Ruprich-Robert--2017|Ruprich-Robert et al., 2017]] ), and also warming in the Indian Ocean ( [[#Luo--2012|Luo et al., 2012]] ; [[#Mochizuki--2016|Mochizuki et al., 2016]] ), could have favoured a PDV transition to its negative phase in the 2000s. Considering the possible influence of external forcing on Indian Ocean decadal variability ( [[#3.7.4|Section 3.7.4]] ) and AMV ( [[#3.7.7|Section 3.7.7]] ), any such human influence on PDV would be indirect through changes in these ocean basins, and then imported to the Pacific via inter-basin coupling. However, this human influence on AMV, and how consistently such inter-basin processes affect PDV phase shifts, are uncertain. Other modelling studies find that anthropogenic aerosols can influence the PDV ( [[#Verma--2019|Verma et al., 2019]] ; [[#Amiri-Farahani--2020|Amiri-Farahani et al., 2020]] ; [[#Dow--2020|Dow et al., 2020]] ). It is however unclear whether and how much those forcings contributed to the observed variations of PDV. In CMIP6 models, the temporal correlation of the multi-model ensemble mean PDV index with its observational counterpart is insignificant and negligible (Figure 3.39f), suggesting that any externally-driven component in historical PDV variations was weak. Lastly, the multi-model ensemble mean computed from CMIP6 historical simulations shows slightly stronger variation than the CMIP5 counterpart, suggesting a greater simulated influence from external forcings in CMIP6. Still, the fraction of the forced signal to the total PDV is very low (Figure 3.39f), in contrast to AMV ( [[#3.7.7|Section 3.7.7]] ). Consistently, [[#Liguori--2020|Liguori et al. (2020)]] estimate that the variance fraction of the externally-driven to total PDV is up to only 15% in a multi-model large ensemble of historical simulations. These findings support an assessment that PDV is mostly driven by internal variability since the pre-industrial era. The sensitivity of ensemble-mean PDV trends to the ensemble size ( [[#Oudar--2018|Oudar et al., 2018]] ), and the dominance of the ensemble spread over the ensemble mean in the 60-year trend of the equatorial Pacific zonal SST gradient in large ensemble simulations ( [[#Watanabe--2021|Watanabe et al., 2021]] ), also support this statement.

<div id="_idContainer088" class="•-2-columns"></div>

[[File:f624a371b10fb678d8209889f1f1964c IPCC_AR6_WGI_Figure_3_39.png]]

Figure 3.39 | '''Model evaluation of the Pacific Decadal Variability (PDV). (a, b)''' Sea surface temperature (SST) anomalies (°C) regressed onto the Tripole Index (TPI; [[#Henley--2015|Henley et al., 2015]] ) for 1900–2014 in '''(a)''' ERSST version 5 and '''(b)''' CMIP6 multi-model ensemble (MME) mean composite obtained by weighting ensemble members by the inverse of the model ensemble size. A 10-year low-pass filter was applied beforehand. Cross marks in (a) represent regions where the anomalies are not significant at the 10% level based on a t-test. Diagonal lines in (b) indicate regions where less than 80% of the runs agree in sign. '''(c)''' A Taylor diagram summarizing the representation of the PDV pattern in CMIP5 (each ensemble member is shown as a cross in light blue, and the weighted multi-model mean as a dot in dark blue), CMIP6 (each ensemble member is shown as a cross in red, and the weighted multi-model mean as a dot in orange) and observations over 40°S–60°N and 110°E–70°W. The reference pattern is taken from ERSST version 5 and black dots indicate other observational products: Hadley Centre Sea Ice and Sea Surface Temperature data set version 1 (HadISST version 1) and Centennial in situ Observation-Based Estimates of Sea Surface Temperature version 2 (COBE-SST2). '''(d)''' Autocorrelation of unfiltered annual TPI at lag one year and 10-year low-pass filtered TPI at lag 10 years for observations over 1900–2014 (horizontal lines), 115-year chunks of pre-industrial control simulations (open boxes) and individual historical simulations over 1900–2014 (filled boxes) from CMIP5 (blue) and CMIP6 (red). '''(e)''' As in (d), but showing standard deviation of the unfiltered and filtered TPI (°C). Boxes and whiskers show weighted multi-model means, interquartile ranges and 5th and 95th percentiles. '''(f)''' Time series of the 10-year low-pass filtered TPI (°C) in ERSST version 5, HadISST version 1 and COBE-SST2 observational estimates (black) and CMIP5 and CMIP6 historical simulations. The thick red and light blue lines are the weighted multi-model mean for the historical simulations in CMIP5 and CMIP6, respectively, and the envelopes represent the 5th–95th percentile ranges across ensemble members. The 5–95% confidence interval for the CMIP6 multi-model mean is given in thin dashed lines. Further details on data sources and processing are available in the chapter data table (Table 3.SM.1).

In CMIP5 last millennium simulations, there is no consistency in temporal variations of PDV across the ensemble ( [[#Fleming--2016|Fleming and Anchukaitis, 2016]] ). This supports the notion that PDV is internal in nature. However, this issue remains difficult to assess because paleoclimate reconstructions of PDV have too poor a level of agreement for a rigorous model evaluation in past millennia ( [[#Henley--2017|Henley, 2017]] ).

To conclude, there is ''high confidence'' that internal variability has been the main driver of the PDV since pre-industrial times, despite some modelling evidence for potential external influence. This assessment is supported by studies based on large ensemble simulations that found the dominance of internally-driven PDV, and the CMIP6-based assessment (Figure 3.39). As such, PDV is an important driver of decadal internal climate variability which limits detection of human influence on various aspects of decadal climate change on global to regional scales ( ''high confidence'' ). Model evaluation of PDV is hampered by short observational records, spatial incompleteness of observations before the satellite observation era, and poor agreement among paleoclimate reconstructions. Despite the limitations of these model-observation comparisons, CMIP5 models, on average, simulate broadly realistic spatial structures of the PDV, but with a clear bias in the South Pacific ( ''medium confidence'' ). CMIP5 models also ''very'' ''likely'' underestimate PDV magnitude. CMIP6 models tend to show better overall performance in spatial structure and magnitude of PDV, but there is ''low confidence'' in this assessment due to the lack of literature.

<div id="3.7.7" class="h2-container"></div>

<span id="atlantic-multi-decadal-variability"></span>
=== 3.7.7 Atlantic Multi-decadal Variability ===

<div id="h2-27-siblings" class="h2-siblings"></div>

Atlantic Multi-decadal Variability (AMV) refers to a climate mode representing basin-wide multi-decadal fluctuations in surface temperatures in the North Atlantic (Figure 3.40a,f), with teleconnections particularly pronounced over the adjacent continents and the Arctic. The AMV phenomenon is usually assessed through SST anomalies averaged over the entire North Atlantic basin, hereafter the AMV index, but it is associated with many physical processes including three-dimensional ocean circulation, such as AMOC fluctuations ( [[#3.5.4.1|Section 3.5.4.1]] ), gyre adjustments, and salt and heat transport in the entire North Atlantic and subarctic Atlantic basins. The AMV, together with the PDV, has been shown to have modulated GSAT on multi-decadal time scales since pre-industrial times (Cross-Chapter Box 3.1; T. [[#Wu--2019|Wu et al., 2019]] a; [[#Li--2020|Li et al., 2020]] ). A detailed description of the AMV as well as its associated teleconnection over land is given in Annex IV.2.7.

<div id="_idContainer090" class="•-2-columns"></div>

[[File:638c52fbeddf9c608bea6a7a66b83f19 IPCC_AR6_WGI_Figure_3_40.png]]

Figure 3.40 | '''Model evaluation of Atlantic Multi-decadal Variability (AMV). (a, b)''' Sea surface temperature (SST) anomalies (°C) regressed onto the AMV index defined as the 10-year low-pass filtered North Atlantic (0°–60°N, 80°W–0°E) area-weighted SST* anomalies over 1900–2014 in '''(a)''' ERSST version 5 and '''(b)''' the CMIP6 multi-model ensemble (MME) mean composite obtained by weighting ensemble members by the inverse of each model’s ensemble size. The asterisk denotes that the global mean SST anomaly has been removed at each time step of the computation. Cross marks in (a) represent regions where the anomalies are not significant at the 10% level based on a t-test. Diagonal lines in (b) show regions where less than 80% of the runs agree in sign. '''(c)''' A Taylor diagram summarizing the representation of the AMV pattern in CMIP5 (each ensemble member is shown as a cross in light blue, and the weighted multi-model mean is shown as a dot in dark blue), CMIP6 (each ensemble member is shown as a cross in red, and the weighted multi-model mean is shown as a dot in orange) and observations over [0°–60°N, 80°W–0°E]. The reference pattern is taken from ERSST version 5 and black dots indicate other observational products (HadISST version 1 and COBE-SST2). '''(d)''' Autocorrelation of unfiltered annual AMV index at lag one year and 10-year low-pass filtered AMV index at lag 10 years for observations over 1900–2014 (horizontal lines), 115-year chunks of pre-industrial control simulations (open boxes) and individual historical simulations over 1900–2014 (filled boxes) from CMIP5 (blue) and CMIP6 (red). '''(e)''' As in (d), but showing standard deviation of the unfiltered and filtered AMV indices (°C). Boxes and whiskers show the weighted multi-model means, interquartile ranges and 5th and 95th percentiles. '''(f)''' Time series of the AMV index (°C) in ERSST version 5, HadISST version 1 and COBE-SST2 observational estimates (black) and CMIP5 and CMIP6 historical simulations. The thick red and light blue lines are the weighted multi-model mean for the historical simulations in CMIP5 and CMIP6, respectively, and the envelopes represent the 5th–95th percentile ranges obtained from all ensemble members. The 5–95% confidence interval for the CMIP6 multi-model mean is shown by the thin dashed line. Further details on data sources and processing are available in the chapter data table (Table 3.SM.1).

AR5 assessed, based on climate models, that the AMV was primarily internally-driven alongside some contribution from external forcings (mainly anthropogenic aerosols) over the late 20th century ( [[#Bindoff--2013|Bindoff et al., 2013]] ; [[#Flato--2013|Flato et al., 2013]] ). But AR5 also concluded that models show medium performance in reproducing the observed AMV, with difficulties in simulating the time scale, the spatial structure and the coherency between all the physical processes involved ( [[#Flato--2013|Flato et al., 2013]] ).

Climate models analysed since AR5 continue to simulate AMV-like variability as part of their internal variability. This statement is mostly based on CMIP5 pre-industrial control and historical simulations ( [[#Wouters--2012|Wouters et al., 2012]] ; [[#Schmith--2014|Schmith et al., 2014]] ; [[#Menary--2015|Menary et al., 2015]] ; [[#Ruprich-Robert--2015|Ruprich-Robert and Cassou, 2015]] ; [[#Brown--2016b|Brown et al., 2016b]] ; [[#Chen--2016|Chen et al., 2016]] ; [[#Kim--2018a|Kim et al., 2018a]] ) and is also true for the CMIP6 models ( [[#Menary--2018|Menary et al., 2018]] ; [[#Voldoire--2019b|Voldoire et al., 2019b]] ). Models also continue to support links to a wide array of remote climate influences through atmospheric teleconnections ( [[#Martin--2014|Martin et al., 2014]] ; [[#Ruprich-Robert--2017|Ruprich-Robert et al., 2017]] , 2018; [[#Monerie--2019|Monerie et al., 2019]] ; [[#Qasmi--2020|Qasmi et al., 2020]] ; [[#Ruggieri--2021|Ruggieri et al., 2021]] ). Even if debate remains ( [[#Clement--2015|Clement et al., 2015]] ; [[#Cane--2017|Cane et al., 2017]] ; [[#Mann--2020|Mann et al., 2020]] ), there is now stronger evidence for a crucial role of oceanic dynamics in internal AMV that is primarily linked to the AMOC and its interplay with the NAO ( [[#Zhang--2013a|Zhang et al., 2013a]] ; [[#Müller--2015|Müller et al., 2015]] ; [[#O’Reilly--2016b|O’Reilly et al., 2016b]] , 2019a; [[#Delworth--2017|Delworth et al., 2017]] ; [[#Zhang--2017|Zhang, 2017]] ; [[#Sun--2019|Sun et al., 2019]] ; [[#Kim--2020|Kim et al., 2020]] ). However, considerable diversity in the spatio-temporal properties of the simulated AMV is found in both pre-industrial control and historical CMIP5 experiments ( [[#Zhang--2013|Zhang and Wang, 2013]] ; [[#Wills--2019|Wills et al., 2019]] ). Such model diversity is presumably associated with the wide range of coupled processes associated with AMV ( [[#Baker--2017|Baker et al., 2017]] ; [[#Woollings--2018a|Woollings et al., 2018a]] ) including large-scale atmospheric teleconnections and regional feedbacks relating to tropical clouds, Arctic sea ice in the subarctic basins and Saharan dust, whose relative importance and interactions across time scales are specific to each model ( [[#Martin--2014|Martin et al., 2014]] ; [[#Brown--2016b|Brown et al., 2016b]] ).

Additional studies since AR5 corroborate that CMIP5-era models tend to underestimate many aspects of observed AMV and its SST fingerprint. On average, the duration of modelled AMV episodes is too short, the magnitude of AMV is too weak and its basin-wide SST spatial structure is limited by the poor representation of the link between the tropical North Atlantic and the subpolar North Atlantic/Nordic seas ( [[#Martin--2014|Martin et al., 2014]] ; [[#Qasmi--2017|Qasmi et al., 2017]] ). Such mismatches between observed and simulated AMV (Figure 3.40c–e) have been associated with intrinsic model biases in both mean state ( [[#Menary--2015|Menary et al., 2015]] ; [[#Drews--2016|Drews and Greatbatch, 2016]] ) and variability in the ocean and overlying atmosphere. For instance, compared to available observational data CMIP5 models underestimate the ratio of decadal to interannual variability of the main drivers of AMV, namely the AMOC, NAO and related North Atlantic jet variations ( [[#3.7.1|Section 3.7.1]] ; [[#Bracegirdle--2018|Bracegirdle et al., 2018]] ; [[#Kim--2018b|Kim et al., 2018b]] ; [[#Simpson--2018|Simpson et al., 2018]] ; [[#Yan--2018|Yan et al., 2018]] ), which has strong implications for the simulated temporal statistics of AMV, AMV-induced teleconnections ( [[#Ault--2012|Ault et al., 2012]] ; [[#Menary--2015|Menary et al., 2015]] ) and AMV predictability.

The increase of AMV variance in CMIP6 models (stronger magnitude and longer duration) seems to be explained by the enhanced variability in the subpolar North Atlantic SST (Figure 3.40b,c), which is particularly pronounced in some models, associated with greater variability in the AMOC ( [[#3.5.4.1|Section 3.5.4.1]] ; [[#Voldoire--2019a|Voldoire et al., 2019a]] ; [[#Boucher--2020|Boucher et al., 2020]] ) and greater GMST multi-decadal variability ( [[#3.3.1|Section 3.3.1]] and Figure 3.40c–f; [[#Voldoire--2019b|Voldoire et al., 2019b]] ; [[#Parsons--2020|Parsons et al., 2020]] ). The decadal variance in SST in the subpolar North Atlantic seems now to be slightly overestimated in CMIP6 compared to observational estimates, while the AMV-related tropical SST anomalies remain weaker in line with CMIP5 (Figure 3.40b). The mechanisms producing the tropical-extratropical relationship at decadal time scales remain poorly understood despite stronger evidence since AR5 for the importance of the subpolar gyre SST anomalies in generating tropical changes through atmospheric teleconnection ( [[#Caron--2015|Caron et al., 2015]] ; [[#Ruprich-Robert--2017|Ruprich-Robert et al., 2017]] ; [[#Kim--2020|Kim et al., 2020]] ). Significant discrepancies remain in the simulated AMV spatial pattern when historical simulations are compared to multivariate observations ( [[#Yan--2018|Yan et al., 2018]] ; [[#Robson--2020|Robson et al., 2020]] ).

There is additional evidence since AR5 that external forcing has been playing an important role in shaping the timing and intensity of the observed AMV since pre-industrial times ( [[#Bellomo--2018|Bellomo et al., 2018]] ; [[#Andrews--2020|Andrews et al., 2020]] ). The time synchronisation between observed and multi-model mean AMV SST indices is significant in both CMIP5 and CMIP6 historical simulations, while the explained variance of the forced response in CMIP6 appears stronger (Figure 3.40d–f). The competition between greenhouse gas warming and anthropogenic sulphate aerosol cooling has been proposed to be particularly important over the latter half of the 20th century ( [[#Booth--2012|Booth et al., 2012]] ; [[#Steinman--2015|Steinman et al., 2015]] ; [[#Murphy--2017|Murphy et al., 2017]] ; [[#Undorf--2018a|Undorf et al., 2018a]] ; [[#Haustein--2019|Haustein et al., 2019]] ). The latest observed AMV shift from the cold to the warm phase in the mid-1990s at the surface ocean is well captured in the CMIP6 forced component and may be associated with the lagged response to increased AMOC due to strong anthropogenic aerosol forcing over 1955–1985 ( [[#Menary--2020|Menary et al., 2020]] ) in combination with the rapid response through surface flux processes to declining aerosol forcing and increasing greenhouse gas influence since then. However, natural forcings may have also played a significant role. For instance, volcanic forcing has been shown to contribute in part to the cold phases of the AMV-related SST anomalies observed in the 20th century ( [[#Terray--2012|Terray, 2012]] ; [[#Bellucci--2017|Bellucci et al., 2017]] ; [[#Swingedouw--2017|Swingedouw et al., 2017]] ; [[#Birkel--2018|Birkel et al., 2018]] ). Over the last millennium, natural forcings including major volcanic eruptions and fluctuations in solar activity are thought to have driven a larger fraction of the multi-decadal variations in the AMV than in the industrial era, with some interplay with internal processes ( [[#Otterå--2010|Otterå et al., 2010]] ; [[#Knudsen--2014|Knudsen et al., 2014]] ; [[#Moffa-Sánchez--2014|Moffa-Sánchez et al., 2014]] ; J. [[#Wang--2017|]] [[#Wang--2017|]] [[#Wang--2017|Wang et al., 2017]] ; [[#Malik--2018|Malik et al., 2018]] ; [[#Mann--2021|Mann et al., 2021]] ), but other studies question the role of natural forcings over this period ( [[#Zanchettin--2014|Zanchettin et al., 2014]] ; [[#Lapointe--2020|Lapointe et al., 2020]] ).

Model evaluation of the AMV phenomenon remains difficult because of short observational records (especially of detailed process-based observations), the lack of stationarity in the variance, spatial patterns and frequency of the AMV assessed from modelled SST ( [[#Qasmi--2017|Qasmi et al., 2017]] ), difficulties in estimating the forced signals in both historical simulations and observations ( [[#Tandon--2015|Tandon and Kushner, 2015]] ), and because of probable interplay between internally and externally-driven processes ( [[#Watanabe--2019|Watanabe and Tatebe, 2019]] ). Furthermore, models simulate a large range of historical anthropogenic aerosol forcing ( [[#Smith--2020|Smith et al., 2020]] ) and questions often referred to as signal-to-noise paradox have been raised concerning the models’ ability to correctly simulate the magnitude of the response of AMV-related atmospheric circulation phenomena, such as the NAO ( [[#3.7.1|Section 3.7.1]] ), to both internally and externally generated changes ( [[#Scaife--2018|Scaife and Smith, 2018]] ). Related methodological and epistemological uncertainties also call into question the relevance of the traditional basin-average SST index to assessing the AMV phenomenon ( [[#Zanchettin--2014|Zanchettin et al., 2014]] ; [[#Frajka-Williams--2017|Frajka-Williams et al., 2017]] ; [[#Haustein--2019|Haustein et al., 2019]] ; [[#Wills--2019|Wills et al., 2019]] ).

To summarize, results from CMIP5 and CMIP6 models together with new statistical techniques to evaluate the forced component of modelled and observed AMV, provide ''robust evidence'' that external forcings have modulated AMV over the historical period. In particular, anthropogenic and volcanic aerosols are thought to have played a role in the timing and intensity of the negative (cold) phase of AMV recorded from the mid-1960s to mid-1990s and subsequent warming ( ''medium confidence'' ). However, there is ''low confidence'' in the estimated magnitude of the human influence. The limited level of confidence is primarily explained by difficulties in accurately evaluating model performance in simulating AMV. The evaluation is severely hampered by short instrumental records but also, equally importantly, by the lack of detailed and coherent long-term process-based observations (for example of the AMOC, aerosol optical depth, surface fluxes and cloud changes), which limit our process understanding. In addition, studies often rely solely on simplistic SST indices that may be hard to interpret ( [[#Zhang--2016|Zhang et al., 2016]] ) and may mask critical physical inconsistencies in simulations of the AMV compared to observations ( [[#Zhang--2017|Zhang, 2017]] ).

<div id="3.8" class="h1-container"></div>

<span id="synthesis-across-earth-system-components"></span>