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

== Cross-Chapter Box 3.1 | Global Surface Warming Over the Early 21st Century ==

<div id="h2-8-siblings" class="h2-siblings"></div>

'''Contributors:''' Christophe Cassou (France), Yu Kosaka (Japan), John C. Fyfe (Canada), Nathan P. Gillett (Canada), Ed Hawkins (United Kingdom), Blair Trewin (Australia)

The AR5 found that the rate of global mean surface temperature (GMST) increase inferred from observations over the 1998–2012 period was lower than the rate of increase over the 1951–2012 period, and lower than the ensemble mean increase in historical simulations from CMIP5 climate models extended by Representative Concentration Pathway (RCP) scenario simulations beyond 2005

([[#Flato--2013|Flato et al., 2013]]). This apparent slowdown of surface global warming compared to the 62-year rate was assessed with ''medium confidence'' to have been caused in roughly equal measure by a cooling contribution from internal variability and a reduced trend in external forcing (particularly associated with solar and volcanic forcing) in AR5 based on expert judgement ([[#Flato--2013|Flato et al., 2013]]). In AR5 it was assessed that almost all CMIP5 simulations did not reproduce the observed slower warming, and that there was ''medium confidence'' that the trend difference from the CMIP5 ensemble mean was to a substantial degree caused by internal variability with possible contributions from forcing error and model response uncertainty. This Cross-Chapter Box assesses new findings from observational products and statistical and physical models on trends over the 1998–2012 period considered in AR5.

'''Updated observational and reanalyses datasets and comparison with model simulations'''

Since AR5, there have been version updates and new releases of most observational GMST datasets (Cross-Chapter Box 2.3). All the updated products now available consistently find stronger positive trends for 1998–2012 than those assessed in AR5 ([[#Cowtan--2014|Cowtan and Way, 2014]] ; [[#Karl--2015|Karl et al., 2015]] ; [[#Hausfather--2017|Hausfather et al., 2017]] ; [[#Medhaug--2017|Medhaug et al., 2017]] ; [[#Simmons--2017|Simmons et al., 2017]] ; [[#Risbey--2018|Risbey et al., 2018]]). [[#Simmons--2017|Simmons et al. (2017)]] reported that the 1998–2012 GMST trends in the updated observational and reanalysis datasets available at that time ranged from 0.06°C to 0.14°C per decade, compared with the 0.05°C per decade on average reported in AR5, while the latest data products reported in [[IPCC:Wg1:Chapter:Chapter-2|Chapter 2]] Table 2.4 show GMST or global mean near-surface air temperature (GSAT) trends over that period ranging from 0.12°C to 0.14°C per decade. The lowest trend in [[#Simmons--2017|Simmons et al. (2017)]] is from HadCRUT4, now superseded by HadCRUT5, which shows a trend of 0.12°C per decade. The upward revision is mainly due to improved sea surface temperature (SST) datasets and infilling of surface temperature in locations with missing records in observational products, mainly in the Arctic (see Cross-Chapter Box 2.3 for details).

With these updates, all the observed trends assessed here lie within the 10th–90th percentile range of the simulated trends in the CMIP5 and CMIP6 simulations (Cross-Chapter Box 3.1, Figure 1a). This result is insensitive to whether model GSAT (based on surface air temperature) or GMST (based on a blend of surface air temperature over land and sea ice and SST over open ocean) is used, and to whether or not masking with the observational data coverage is applied. Therefore, the observed 1998–2012 trend is consistent with both the CMIP5 or CMIP6 multi-model ensemble of trends over the same period (''high confidence'').

<div id="_idContainer028" class="Body-copy_Boxes_Blue-Boxes_•-Box-subhead-H1---no-space-below"></div>

[[File:f47669ecd7c06b9027c182aba74543b2 IPCC_AR6_WGI_CCBox_3_1_Figure_1.png]]

'''Cross-Chapter Box 3.1, Figure 1 | 15-year trends of global surface temperature for 1998–2012 and 2012–2026. (a, b)''' GSAT and GMST trends for 1998–2012 '''(a)''' and 2012–2026 '''(b)''' . Histograms are based on GSAT in historical simulations of CMIP6 (red shading, extended by SSP2-4.5) and CMIP5 (grey shading; extended by RCP4.5). Filled and open diamonds at the top represent multi-model ensemble means of GSAT and GMST trends, respectively. Diagonal lines show histograms of HadCRUT5.0.1.0. Triangles at the top of (a) represent GMST trends from Berkeley Earth, GISTEMP, [[#Kadow--2020|Kadow et al. (2020)]] and NOAAGlobalTemp-Interim, and the GSAT trend from ERA5. Selected CMIP6 members whose 1998–2012 trends are lower than the HadCRUT5.0.1.0 mean trend are indicated by purple shading (a) and (b). In (a), model GMST and GSAT, and ERA5 GSAT are masked to match HadCRUT data coverage. '''(c–d)''' Trend maps of annual near-surface temperature for 1998–2012 based on HadCRUT5.0.1.0 mean '''(c),''' and composited surface air temperature trends of subsampled CMIP6 simulations '''(d)''' ''with GSAT trends in the purple shaded'' area in (a). In (c), cross marks indicate trends that are not significant at the 10% level based on t-tests with serial correlation taken into account. The ensemble size used for each of the histograms and the trend composite is indicated at the top right of each of the panels (a, b, d). Model ensemble members are weighted with the inverse of the ensemble size of the same model, so that each model is equally weighted. Further details on data sources and processing are available in the chapter data table (Table 3.SM.1).

'''Internal variability'''

All the observation-based GMST and GSAT trends are lower than the multi-model mean GMST and GSAT trends of both CMIP5 and CMIP6 for 1998–2012 (Cross-Chapter Box 3.1, Figure 1a). This suggests a possible cooling contribution from internal variability during this period. This is supported by initialized decadal hindcasts, which account for the phase of the multi-decadal modes of variability (Sections [[#_idTextAnchor002|3.7.6]] and [[#_idTextAnchor003|3.7.7]]), and which reproduce observed global mean SST and GSAT trends better than uninitialized historical simulations ([[#Guemas--2013|Guemas et al., 2013]] ; [[#Meehl--2014|Meehl et al., 2014]]).

Studies since AR5 identify Pacific Decadal Variability (PDV) as the leading mode of variability associated with unforced decadal GSAT fluctuations, with additional influence from Atlantic Multi-decadal Variability (Annex IV.2.6, IV.2.7; [[#Brown--2015|Brown et al., 2015]] ; [[#Dai--2015|Dai et al., 2015]] ; [[#Steinman--2015|Steinman et al., 2015]] ; [[#Pasini--2017|Pasini et al., 2017]]). PDV transitioned from positive (El Niño-like) to negative (La Niña-like) phases during the slow warming period (Figure 3.39f and Cross-Chapter Box 3.1, Figure 1c). Model ensemble members that capture the observed slower decadal warming under transient forcing, and time segments of model simulations that show decadal GSAT decreases under fixed radiative forcing, also feature negative PDV trends (Cross-Chapter Box 3.1, Figure 1d; [[#Meehl--2011|Meehl et al., 2011]] , 2013, 2014; [[#Maher--2014|Maher et al., 2014]] ; [[#Middlemas--2016|Middlemas and Clement, 2016]]), suggesting the influence of PDV. This is confirmed by statistical models with the PDV-GSAT relationship estimated from observations and model simulations ([[#Schmidt--2014|Schmidt et al., 2014]] ; [[#Meehl--2016b|Meehl et al., 2016b]] ; [[#Hu--2017|Hu and Fedorov, 2017]]), selected ensemble members and time segments from model simulations where PDV by chance evolves in phase with observations over the slow warming period ([[#Huber--2014|Huber and Knutti, 2014]] ; [[#Risbey--2014|Risbey et al., 2014]]), and coupled model experiments in which PDV evolution is constrained to follow the observations ([[#Kosaka--2013|Kosaka and Xie, 2013]] , 2016; [[#England--2014|England et al., 2014]] ; [[#Watanabe--2014|Watanabe et al., 2014]] ; [[#Delworth--2015|Delworth et al., 2015]]). Part of the PDV trend may have been driven by anthropogenic aerosols ([[#Smith--2016|Smith et al., 2016]]); however, this result is model-dependent, and internally-driven PDV dominates the forced PDV signal in the CMIP6 multi-model ensemble ([[#3.7.6|Section 3.7.6]]). It is also notable that there is large uncertainty in the magnitude of the PDV influence on GSAT across models ([[#Deser--2017a|Deser et al., 2017a]] ; C.-Y. [[#Wang--2017|]] [[#Wang--2017|]] [[#Wang--2017|Wang et al., 2017]]) and among the studies cited above. In addition to PDV, contributions to the reduced warming trend from wintertime Northern Hemisphere atmospheric internal variability, particularly associated with a trend towards the negative phase of the Northern Annular Mode/North Atlantic Oscillation (Annex IV.2.1; [[#Guan--2015|Guan et al., 2015]] ; [[#Saffioti--2015|Saffioti et al., 2015]] ; [[#Iles--2017|Iles and Hegerl, 2017]]) or the Cold Ocean–Warm Land (COWL) pattern ([[#Molteni--2017|Molteni et al., 2017]] ; [[#Yang--2020|Yang et al., 2020]]) have been suggested, leading to regional continental cooling over a large part of Eurasia and North America (Cross-Chapter Box 3.1, Figure 1c; [[#Li--2015|C. Li et al., 2015]] ; [[#Deser--2017a|Deser et al., 2017a]] ; [[#Gan--2019|Gan et al., 2019]]).

Such internally-driven variation of decadal GSAT trends is not unique to the 1998–2012 period ([[IPCC:Wg1:Chapter:Chapter-1#1.4.2.1|Section 1.4.2.1]] ; [[#Lovejoy--2014|Lovejoy, 2014]] ; [[#Roberts--2015|Roberts et al., 2015]] ; [[#Dai--2019|Dai and Bloecker, 2019]]). Due to the nature of internal variability, surface temperature changes over the 1998–2012 period are regionally- and seasonally-varying (Cross-Chapter Box 3.1, Figure 1c; [[#Trenberth--2014|Trenberth et al., 2014]] ; [[#Zang--2019|Zang et al., 2019]]). Further, there was no slowdown in the increasing occurrence of hot extremes over land ([[#Kamae--2014|Kamae et al., 2014]] ; [[#Seneviratne--2014|Seneviratne et al., 2014]] ; [[#Imada--2017|Imada et al., 2017]]). Thus, the internally-driven slowdown of GSAT increase does not correspond to slowdown of warming everywhere on the Earth’s surface.

'''Updated forcing'''

CMIP5 historical simulations driven by observed forcing variations ended in 2005 and were extended with RCP scenario simulations for model-observation comparisons beyond that date. Post AR5 studies based on updated external forcing show that while no net effect of updated anthropogenic aerosols is found on GSAT trends ([[#Murphy--2013|Murphy, 2013]] ; [[#Gettelman--2015|Gettelman et al., 2015]] ; [[#Oudar--2018|Oudar et al., 2018]]), natural forcing by moderate volcanic eruptions in the 21st century ([[#Haywood--2014|Haywood et al., 2014]] ; [[#Ridley--2014|Ridley et al., 2014]] ; [[#Santer--2014|Santer et al., 2014]]) and a prolonged solar irradiance minimum around 2009 compared to the normal 11-year cycle ([[#Lean--2018|Lean, 2018]]) yield a negative contribution to radiative forcing, which was missing in CMIP5 (Figure 2.2). This explains part of the difference between observed and CMIP5 trends, as shown based on EMIC simulations ([[#Huber--2014|Huber and Knutti, 2014]] ; [[#Ridley--2014|Ridley et al., 2014]]), statistical and mathematical models ([[#Schmidt--2014|Schmidt et al., 2014]] ; [[#Lean--2018|Lean, 2018]]), and process-based climate models ([[#Santer--2014|Santer et al., 2014]]). However, in a single climate model study by [[#Thorne--2015|Thorne et al. (2015)]] , updating most forcings (greenhouse gas concentrations, solar irradiance, and volcanic and anthropogenic aerosols) available when the study was done made no significant difference to the 1998–2012 GMST trend from that obtained with original CMIP5 forcing. Potential underestimation of volcanic (negative) forcing may have played a role ([[#Outten--2015|Outten et al., 2015]]). In the multi-model ensemble mean, the 1998–2012 GMST trends are almost equal in CMIP5 and CMIP6 (Cross-Chapter Box 3.1, Figure 1a), suggesting compensation by a higher transient climate response and equilibrium climate sensitivity in CMIP6 than CMIP5 (Section 7.5.6). To summarize, while there is ''medium confidence'' that natural forcing that was missing in CMIP5 contributed to the difference of observed and simulated GMST trends, ''confidence'' remains ''low'' in the quantitative contribution of net forcing updates.

'''Energy budget and heat redistribution'''

The early 21st century slower warming was observed in atmospheric temperatures, but the heat capacity of the atmosphere is very small compared to that of the ocean. Although there is noticeable uncertainty among observational products (H. [[#Su--2017|]] [[#Su--2017|Su et al., 2017]]) and observation quality changes through time, global ocean heat content continued to increase during the slower surface warming period (''very high confidence''), at a rate consistent with CMIP5 and CMIP6 historical simulations (Sections 2.3.3.1, [[#_idTextAnchor001|3.5.1.3]] and 7.2.2.2). There is ''high confidence'' that the Earth’s energy imbalance was larger in the 2000s than in the 1985–1999 period (Section 7.2.2.1), consistent with accelerating ocean heat uptake in the past two decades (Section [[#_idTextAnchor001|3.5.1.3]]). Internal decadal variability is mainly associated with redistribution of heat within the climate system (X.H. [[#Yan--2016|]] [[#Yan--2016|]] [[#Yan--2016|Yan et al., 2016]] ; [[#Drijfhout--2018|Drijfhout, 2018]]) while associated top of the atmosphere radiation anomalies are weak ([[#Palmer--2014|Palmer and McNeall, 2014]]). Heat redistribution in the top 350 m of the Indian and Pacific Oceans has been found to be the main contributor to reduced surface warming during the slower surface warming period ([[#Lee--2015|Lee et al., 2015]] ; [[#Nieves--2015|Nieves et al., 2015]] ; F. [[#Liu--2016|]] [[#Liu--2016|Liu et al., 2016]]), consistent with the simulated signature of PDV ([[#England--2014|England et al., 2014]] ; [[#Maher--2018a|Maher et al., 2018a]] ; [[#Gastineau--2019|Gastineau et al., 2019]]). Below 700 m, enhanced heat uptake over the slower surface warming period was observed mainly in the North Atlantic and Southern Ocean ([[#Chen--2014|Chen and Tung, 2014]]), though whether this was a response to forcing or a unique signature of the slow GMST warming has been questioned (W. [[#Liu--2016|]] [[#Liu--2016|Liu et al., 2016]]).

'''Summary and implications'''

With updated observation-based GMST datasets and forcing, improved analysis methods, new modelling evidence and deeper understanding of mechanisms, there is ''very'' ''high confidence'' that the slower GMST and GSAT increase inferred from observations in the 1998–2012 period was a temporary event induced by internal and naturally-forced variability that partly offset the anthropogenic warming trend over this period. Nonetheless, the heating of the climate system continued during this period, as reflected in the continued warming of the global ocean (''very high confidence'') and in the continued rise of hot extremes over land (''medium confidence''). Considering all the sources of uncertainties, it is impossible to robustly identify a single cause of the early 2000s slowdown ([[#Hedemann--2017|Hedemann et al., 2017]] ; [[#Power--2017|Power et al., 2017]]); rather, it should be interpreted as due to a combination of several factors ([[#Huber--2014|Huber and Knutti, 2014]] ; [[#Schmidt--2014|Schmidt et al., 2014]] ; [[#Medhaug--2017|Medhaug et al., 2017]]).

A major El Niño event in 2014–2016 led to three consecutive years of record annual GMST with unusually strong heat release from the North-western Pacific Ocean ([[#Yin--2018|Yin et al., 2018]]), which marked the end of the slower warming period ([[#Hu--2017|Hu and Fedorov, 2017]] ; J. [[#Su--2017|]] [[#Su--2017|Su et al., 2017]] ; [[#Cha--2018|Cha et al., 2018]]). The past five-year period (2016–2020) is the hottest five-year period in the instrumental record up to 2020 (''high confidence''). This rapid warming was accompanied by a PDV shift toward its positive phase (J. [[#Su--2017|]] [[#Su--2017|Su et al., 2017]] ; [[#Cha--2018|Cha et al., 2018]]). A higher rate of warming following the 1998–2012 period is consistent with the predictions in AR5 Box 9.2 ([[#Flato--2013|Flato et al., 2013]]) and with a statistical prediction system (Sévellec and [[#Drijfhout--2018|Drijfhout, 2018]]). Initialized decadal predictions show higher GMST trends in the early 2020s compared to uninitialized simulations ([[#Thoma--2015|Thoma et al., 2015]] ; [[#Meehl--2016a|Meehl et al., 2016a]]).

While some recent studies find that internal decadal GSAT variability may become weaker under GSAT warming, associated in part with reduced amplitude PDV ([[IPCC:Wg1:Chapter:Chapter-4#4.5.3.5|Section 4.5.3.5]] ; [[#Brown--2017|Brown et al., 2017]]), the weakening is small under a realistic range of warming. A large volcanic eruption would temporarily cool GSAT (Cross-Chapter Box 4.1). Thus, there is ''very high confidence'' that reduced and increased GMST and GSAT trends at decadal time scales will continue to occur in the 21st century ([[#Meehl--2013|Meehl et al., 2013]] ; [[#Roberts--2015|Roberts et al., 2015]] ; [[#Medhaug--2016|Medhaug and Drange, 2016]]). However, such internal or volcanically forced decadal variations in GSAT trend have little effect on centennial warming ([[#England--2015|England et al., 2015]] ; Cross-Chapter Box 4.1).
</div>
<div id="3.3.2" class="h2-container"></div>

<span id="precipitation-humidity-and-streamflow"></span>

=== 3.3.2 Precipitation, Humidity and Streamflow ===

<div id="h2-9-siblings" class="h2-siblings"></div>

<div id="3.3.2.1" class="h3-container"></div>

<span id="paleoclimate-context"></span>
==== 3.3.2.1 Paleoclimate Context ====

<div id="h3-5-siblings" class="h3-siblings"></div>

A fact hindering detection and attribution studies in precipitation and other hydrological variables is the large internal variability of these fields relative to the anthropogenic signal. This low signal-to-noise ratio hinders the emergence of the anthropogenic signal from natural variability. Moreover, the sign of the change depends on location and time of the year. Paleoclimate records provide valuable context for observed trends in the 20th and 21st century and assist with the attribution of these trends to human influence (see also ([[IPCC:Wg1:Chapter:Chapter-2#2.3.1.3.1|Section 2.3.1.3.1]]). By nature, hydrological proxy data represent regional conditions, but taken together can represent large-scale patterns. As an example of how paleorecords have helped assessing the origin of changes, we consider some, mainly subtropical, regions which have experienced systematic drying in recent decades (see also Section 8.3.1.3). Paleoclimate simulations of monsoons are assessed in [[#3.3.3.2|Section 3.3.3.2]] .

Records of tree ring width have provided evidence that recent prolonged dry spells in the Levant and Chile are unprecedented in the last millennium (''high confidence'') ([[#Cook--2016a|Cook et al., 2016a]] ; [[#Garreaud--2017|Garreaud et al., 2017]]). East Africa has also been drying in recent decades ([[#Rowell--2015|Rowell et al., 2015]] ; [[#Hoell--2017|Hoell et al., 2017]]), a trend that is unusual in the context of the sedimentary paleorecord spanning the last millennium ([[#Tierney--2015|Tierney et al., 2015]]). This may be a signature of anthropogenic forcing but cannot yet be distinguished from natural variability ([[#Hoell--2017|Hoell et al., 2017]] ; [[#Philip--2018|Philip et al., 2018]]). Likewise, tree rings indicate that the 2012–2014 drought in the south-western United States was exceptionally severe in the context of natural variability over the last millennium, and may have been exacerbated by the contribution of anthropogenic temperature rise (''medium confidence'') ([[#Griffin--2014|Griffin and Anchukaitis, 2014]] ; [[#Williams--2015|Williams et al., 2015]]). Furthermore, [[#Williams--2020|Williams et al. (2020)]] used a combination of hydrological modelling and tree-ring reconstructions to show that the period from 2000 to 2018 was the driest 19-year span in south-western North America since the late 1500s. Nonetheless, tree rings also indicate the presence of prolonged megadroughts in western North America throughout the last millennium that were more severe than 20th and 21st century events (''high confidence'') ([[#Cook--2004|Cook et al., 2004]] , 2010, 2015). These were associated with internal variability ([[#Coats--2016|Coats et al., 2016]] ; [[#Cook--2016b|Cook et al., 2016b]]) and indicate that large-magnitude changes in the water cycle may occur irrespective of anthropogenic influence (see also [[#McKitrick--2019|McKitrick and Christy, 2019]]).

Paleoclimate records also allow for model evaluation under conditions different from present-day. The AR5 concluded that models can successfully reproduce to first-order patterns of past precipitation changes during the Last Glacial Maximum (LGM) and mid-Holocene, though simulated precipitation changes during the mid-Holocene tended to be underestimated ([[#Flato--2013|Flato et al., 2013]]). Further analysis of CMIP5 models confirmed these results but has also revealed systematic offsets from the paleoclimate record ([[#DiNezio--2013|DiNezio and Tierney, 2013]] ; [[#Hargreaves--2014|Hargreaves and Annan, 2014]] ; [[#Harrison--2014|Harrison et al., 2014]] , 2015; [[#Bartlein--2017|Bartlein et al., 2017]] ; [[#Scheff--2017|Scheff et al., 2017]] ; [[#Tierney--2017|Tierney et al., 2017]]). [[#Harrison--2014|Harrison et al. (2014)]] concluded that CMIP5 models do not perform better in simulating rainfall during the LGM and mid-Holocene than earlier model versions despite higher resolution and complexity. However, prescribing changes in vegetation and dust was found to improve the match to the paleoclimate record ([[#Pausata--2016|Pausata et al., 2016]] ; [[#Tierney--2017|Tierney et al., 2017]]) suggesting that vegetation feedbacks in the CMIP5 models may be too weak (''low confidence'') ([[#Hopcroft--2017|Hopcroft et al., 2017]]). [[#Brierley--2020|Brierley et al. (2020)]] compared the latitudinal gradient of annual precipitation changes in the European–African sector simulated by CMIP6 models for the mid-Holocene with pollen-based reconstructions and showed that models generally reproduce the direction of changes seen in the reconstructions (Figure 3.11). They do not show a robust signal in area averaged rainfall over most European regions where quantitative reconstructions exist, which is not incompatible with reconstructions. Over the Sahara/Sahel and West Africa regions, where reconstructions suggest positive anomalies during the mid-Holocene, both CMIP5 and CMIP6 models also simulate a rainfall increase, but it is much weaker (see also ([[#3.3.3.2|Section 3.3.3.2]]). Overall, however, large discrepancies remain between simulations and reconstructions.

<div id="_idContainer030" class="•-2-columns"></div>

[[File:33efc5a39a5a4f520a1db9ebb7ebbd01 IPCC_AR6_WGI_Figure_3_11.png]]

Figure 3.11 | '''Comparison between simulated annual precipitation changes and pollen-based reconstructions in the mid-Holocene (6000 years ago).''' The area-averaged changes relative to the pre-industrial control simulations over five regions ([[#Iturbide--2020|Iturbide et al., 2020]]) as simulated by CMIP6 models (individually identifiable, one ensemble member per model) and CMIP5 models (blue) are shown, stretching from the tropics to high-latitudes. All regions contain multiple quantitative reconstructions of changes relative to present day; their interquartile range are shown by boxes and with whiskers for their full range excluding outliers. Figure is adapted from [[#Brierley--2020|Brierley et al. (2020)]] . Further details on data sources and processing are available in the chapter data table (Table 3.SM.1).

[[#Liu--2018|Liu et al. (2018)]] evaluated the soil moisture changes that occurred during the LGM and concluded that the multi-model median from CMIP5 is consistent with available paleo-records in some regions, but not in others. CMIP5 models accurately reproduce an increase in moisture in the western United States, related to an intensified winter storm track and decreased evaporative demand ([[#Oster--2015|Oster et al., 2015]] ; [[#Ibarra--2018|Ibarra et al., 2018]] ; [[#Lora--2018|Lora, 2018]]). On the other hand, CMIP5 models show a wide variety of responses in the tropical Indo-Pacific region, with only a few matching the pattern of change inferred from the paleoclimate record ([[#DiNezio--2013|DiNezio and Tierney, 2013]] ; [[#DiNezio--2018|DiNezio et al., 2018]]). The variable response across models is related to the effect of the exposure of the tropical shelves during glacial times, which variously intensifies or weakens convection in the rising branch of the Walker cell, depending on model parameterization ([[#DiNezio--2011|DiNezio et al., 2011]]). For the Last Interglacial, CMIP6 models reproduce the proxy-based increased precipitation relative to pre-industrial in the North African, South Asian and North American regions, but not in Australia ([[#Scussolini--2019|Scussolini et al., 2019]]).

In summary, there is ''medium confidence'' that CMIP5 and CMIP6 models can reproduce broad aspects of precipitation changes during paleo reference periods, but large discrepancies remain. Further assessment of model performance and comparison between CMIP5 and CMIP6 during past climates can be found in [[#3.8.2.1|Section 3.8.2.1]] .

<div id="3.3.2.2" class="h3-container"></div>

<span id="atmospheric-water-vapour"></span>
==== 3.3.2.2 Atmospheric Water Vapour ====

<div id="h3-6-siblings" class="h3-siblings"></div>

The AR5 concluded that an anthropogenic contribution to increases in specific humidity is found with ''medium confidence'' at and near the surface. A levelling off of atmospheric water vapour over land in the last two decades that needed better understanding, and remaining observational uncertainties, precluded a more confident assessment ([[#Bindoff--2013|Bindoff et al., 2013]]). Sections 4.5.1.3 and 8.3.1.4 show that there have been significant advances in the understanding of the processes controlling land surface humidity. In particular, there has been a focus on the role of oceanic moisture transport and land-atmosphere feedbacks in explaining the observed trends in relative humidity.

Water vapour is the most important natural greenhouse gas and its amount is expected to increase in a global warming context leading to further warming. Particularly important are changes in the upper troposphere because there water vapour regulates the strength of the water-vapour feedback (Section 7.4.2.2). CMIP5 models have been shown to have a wet bias in the tropical upper troposphere and a dry bias in the lower troposphere, with the former bias and model spread being larger than the latter ([[#Jiang--2012|Jiang et al., 2012]] ; [[#Tian--2013|Tian et al., 2013]]). [[#Tian--2013|Tian et al. (2013)]] also showed that in comparison to the AIRS specific humidity, CMIP5 models have the well-known double Inter-tropical Convergence Zone (ITCZ) bias in the troposphere from 1000 hPa to 300 hPa, especially in the tropical Pacific. Water vapour biases in models are dominated by errors in relative humidity throughout the troposphere, which are in turn closely related to errors in large scale circulation; temperature errors dominate near the tropopause ([[#Takahashi--2016|Takahashi et al., 2016]]). Section 7.4.2.2 discusses this topic in more detail for CMIP6 models. However, [[#Schröder--2019|Schröder et al. (2019)]] show that the majority of well-established water vapour records are affected by inhomogeneity issues and thus should be used with caution (see also ([[IPCC:Wg1:Chapter:Chapter-2#2.3.1.3.3|Section 2.3.1.3.3]]). A comparison of trends in column water vapour path for 1998–2019 in satellite data, a reanalysis, CMIP5 and CMIP6 simulations averaged over the near-global ocean reveals that while on average model trends are higher than those in observations and a reanalysis, the latter lie within the multi-model range (Figure 3.12).

<div id="_idContainer032" class="•-2-columns"></div>

[[File:dc96a534bd92c51c5bc548b5a349cacd IPCC_AR6_WGI_Figure_3_12.png]]

Figure 3.12 | '''Total column water vapour trends (% per decade) for the period 1988–2019 averaged over the near-global oceans (50°S–50°N).''' The figure shows satellite data (RSS) and ERA5.1 reanalysis, as well as CMIP5 (blue) and CMIP6 (red) historical simulations. All available ensemble members were used (see [[#3.2|Section 3.2]]). Fits to the model trend probability distributions were performed with kernel density estimation. Figure is updated from [[#Santer--2007|Santer et al. (2007)]] . Further details on data sources and processing are available in the chapter data table (Table 3.SM.1).

The detection and attribution of tropospheric water vapour changes can be traced back to [[#Santer--2007|Santer et al. (2007)]] , who used estimates of atmospheric water vapour from the satellite-based Special Sensor Microwave Imager (SSM/I) and from CMIP3 historical climate simulations. They provided evidence of human-induced moistening of the troposphere, and found that the simulated human fingerprint pattern was detectable at the 5% level by 2002 in water vapour satellite data (from 1988 to 2006). The observed changes matched the historical simulations forced by greenhouse gas changes and other anthropogenic forcings, and not those due to natural variability alone. Then, [[#Santer--2009|Santer et al. (2009)]] repeated this study with CMIP5 models, and found that the detection and attribution conclusions were not sensitive to model quality. These results demonstrate that the human fingerprint is governed by robust and basic physical processes, such as the water vapour feedback. Finally, [[#Chung--2014|Chung et al. (2014)]] extended this line of research by focusing on the global-mean water vapour content in the upper troposphere. Using satellite-based observations and sets of CMIP5 climate simulations run under various climate-forcing options, they showed that the observed moistening trend of the upper troposphere over the 1979–2005 period could not be explained by internal variability alone, but is attributable to a combination of anthropogenic and natural forcings. This increase in water vapour is accompanied by a reduction in mid-tropospheric relative humidity and clouds in the subtropics and mid-latitude in both models and observations related to changes in the Hadley cell ([[#3.3.3.1.1|Section 3.3.3.1.1]] ; [[#Lau--2015|Lau and Kim, 2015]]).

[[#Dunn--2017|Dunn et al. (2017)]] confirmed earlier findings that global mean surface relative humidity increased between 1973 and 2000, followed by a steep decline (also reported in [[#Willett--2014|Willett et al., 2014]]) until 2013, and specific humidity correspondingly increased and then remained approximately constant (see also ([[IPCC:Wg1:Chapter:Chapter-2#2.3.1.3.2|Section 2.3.1.3.2]]), with none of the CMIP5 models capturing this behaviour. They noted biases in the mean state of the CMIP5 models’ surface relative humidity (and ascribed the failure to the representation of land surface processes and their response to CO <sub>2</sub> forcing), concluding that these biases preclude any detection and attribution assessment. On the other hand, [[#Byrne--2018|Byrne and O’Gorman (2018)]] showed that the positive trend in specific humidity continued in recent years and can be detected over land and ocean from 1979 to 2016. Moreover, they provided a theory suggesting that the increase in annual surface temperature and specific humidity as well as the decrease in relative humidity observed over land are linked to warming over the neighbouring ocean. They also pointed out that the negative trend in relative humidity over land regions is quite uncertain and requires further investigation. A recent study has also identified an anthropogenically-driven decrease in relative humidity over the Northern Hemisphere mid-latitude continents in summer between 1979 and 2014, which was underestimated by CMIP5 models ([[#Douville--2017|Douville and Plazzotta, 2017]]). Furthermore, in a modelling study [[#Douville--2020|Douville et al. (2020)]] showed that this decrease in boreal summer relative humidity over mid-latitudes is related not only to global ocean warming, but also to the physiological effect of CO <sub>2</sub> on plants in the land surface model.

In summary, we assess that it is ''likely'' that human influence has contributed to moistening in the upper troposphere since 1979. Also, there ''is medium confidence'' that human influence contributed to a global increase in annual surface specific humidity, and ''medium confidence'' that it contributed to a decrease in surface relative humidity over mid-latitude Northern Hemisphere continents during summertime.

<div id="3.3.2.3" class="h3-container"></div>

<span id="precipitation"></span>
==== 3.3.2.3 Precipitation ====

<div id="h3-7-siblings" class="h3-siblings"></div>

AR5 concluded that there was ''medium confidence'' that human influence had contributed to large-scale precipitation changes over land since 1950, including an increase in the Northern Hemisphere mid- to high latitudes. Moreover, AR5 concluded that observational uncertainties and challenges in precipitation modelling precluded a more confident assessment ([[#Bindoff--2013|Bindoff et al., 2013]]). Overall, it found that large-scale features of mean precipitation in CMIP5 models were in modest agreement with observations, but there were systematic errors in the tropics ([[#Flato--2013|Flato et al., 2013]]).

Since AR5, X. [[#Li--2016b|Li et al. (2016b)]] found that CMIP5 models simulate the large scale patterns of annual mean land precipitation and seasonality well, as well as reproducing qualitatively the observed zonal mean land precipitation trends for the period 1948–2005: models capture the drying trends in the tropics and at 45°S and the wetting trend in the Northern Hemisphere mid- to high latitudes, but the amplitudes of the changes are much smaller than observed. Land precipitation was found to show enhanced seasonality in observations ([[#Chou--2013|Chou et al., 2013]]), qualitatively consistent with the simulated response to anthropogenic forcing ([[#Dwyer--2014|Dwyer et al., 2014]]). However, models do not appear to reproduce the zonal mean trends in the magnitude of the seasonal cycle over the period 1948–2005, nor the two-dimensional distributions of trends of annual precipitation and seasonality over land, but differences may be explainable by internal variability (X. [[#Li--2016|]] [[#Li--2016|Li et al., 2016]] b). However, observed trends in seasonality depend on data set used (X. [[#Li--2016|]] [[#Li--2016|Li et al., 2016]] b; [[#Marvel--2017|Marvel et al., 2017]]), and [[#Marvel--2017|Marvel et al. (2017)]] found that observed changes in the annual cycle phase are consistent with model estimates of forced changes. These phase changes are mainly characterized by earlier onset of the wet season on the equatorward flanks of the extratropical storm tracks, particularly in the Southern Hemisphere. Box 8.2 assesses regional changes in water cycle seasonality.

The CMIP5 models have also been shown to adequately simulate the mean and interannual variability of the global monsoon ([[#3.3.3.2|Section 3.3.3.2]]), but maintain the double ITCZ bias in the equatorial Pacific ([[#Lee--2014|Lee and Wang, 2014]] ; [[#Tian--2015|Tian, 2015]] ; [[#Ni--2018|Ni and Hsu, 2018]]). Despite the ITCZ bias, CMIP5 models have been used to detect in reanalysis a southward shift in the ITCZ prior to 1975, followed by a northward shift in the ITCZ after 1975, in response to forced changes in inter-hemispheric temperature contrast (Sections 3.3.1.1 and 8.3.2.1, and Figure 8.11; [[#Bonfils--2020|Bonfils et al., 2020]] ; [[#Friedman--2020|Friedman et al., 2020]]). CMIP5 models perform better than CMIP3 models, in particular regarding the global monsoon domain and intensity ([[#Lee--2014|Lee and Wang, 2014]]).

In observations at time scales less than a day intermittent rainfall fluctuations dominate variability, but CMIP5 models systematically underestimate them ([[#Covey--2018|Covey et al., 2018]]). Moreover, as noted in previous generation models, CMIP5 models produce rainfall too early in the day ([[#Covey--2016|Covey et al., 2016]]). Also, models overpredict precipitation frequency but have weaker intensity, although comparison with observed datasets is complex as there are large differences in intensity among them ([[#Herold--2016|Herold et al., 2016]] ; [[#Pendergrass--2017|Pendergrass and Deser, 2017]] ; [[#Trenberth--2017|Trenberth et al., 2017]]). Regarding trends in precipitation intensity, models have also been shown to reproduce the compensation between increasing heavy precipitation and decreasing light to moderate rainfall ([[#Thackeray--2018b|Thackeray et al., 2018b]]), a characteristic found in the observational record ([[#Gu--2018|Gu and Adler, 2018]]). Regional model performance is further assessed in [[IPCC:Wg1:Chapter:Chapter-8|Chapter 8]] and the Atlas, while precipitation extremes are considered in Chapter 11.

The simulation of annual mean rainfall patterns in the CMIP6 models reveals minor improvements compared to those in CMIP5 models (Figure 3.13). The persistent biases include the double ITCZ in the tropical Pacific (seen as bands of excessive rainfall on both sides of the equatorial Pacific in Figure 3.13b,d) and the southward-shifted ITCZ in the equatorial Atlantic, which have been linked to the meridional pattern of SST bias (S. [[#Zhou--2020|]] [[#Zhou--2020|Zhou et al., 2020]]) and the reduced sensitivity of precipitation to local SST ([[#Good--2021|Good et al., 2021]]). [[#Tian--2020|Tian and Dong (2020)]] also found that all three generations of CMIP models share similar systematic annual mean precipitation errors in the tropics, but that the double ITCZ bias is slightly reduced in CMIP6 models in comparison to CMIP3 and CMIP5 models. They also found some improvement in the overly intense Indian ocean ITCZ and the too dry South American continent except over the Andes. [[#Fiedler--2020|Fiedler et al. (2020)]] identified improvements in the tropical mean spatial correlations and root mean square error of the climatology as well as in the day-to-day variability, but found little change across CMIP phases in the double ITCZ bias and diurnal cycle. The CMIP6 models reproduce better the domain and intensity of the global monsoon (see [[#3.3.3.2|Section 3.3.3.2]]). Moreover, CMIP6 models better represent the storm tracks ([[#Priestley--2020|Priestley et al., 2020]] ; also ([[#3.3.3.3|Section 3.3.3.3]]), thereby reducing the precipitation biases in the North Atlantic and mid-latitudes of the Southern Hemisphere (Figure 3.13b,d). As a result, pattern correlations between simulated and observed annual mean precipitation range between 0.80 and 0.92 for CMIP6 models, compared to a range of 0.79 to 0.88 for CMIP5 ([[#Bock--2020|Bock et al., 2020]]). This relative improvement may be related to increased model resolution, as found when comparing biases in the mean of the HighResMIP models with the mean of the corresponding lower-resolution versions of the same models (see Figure 3.13e,f), particularly in the tropics and extratropical storm tracks. Consistent with this, a recent study using several coupled models showed that increasing the atmospheric resolution leads to a strong decrease in the precipitation bias in the tropical Atlantic ITCZ (see further discussion in [[#3.8.2.2|Section 3.8.2.2]] ; [[#Vannière--2019|Vannière et al., 2019]]). Based on these results we assess that despite some improvements, CMIP6 models still have deficiencies in simulating precipitation patterns, particularly over the tropical ocean (''high confidence'').

<div id="_idContainer034" class="•-2-columns"></div>

[[File:0cec6193260a8480eb69f4fad725ce73 IPCC_AR6_WGI_Figure_3_13.png]]

Figure 3.13 | '''Annual-mean precipitation rate (mm day''' '''–1''' ''') for the period 1995–2014. (a)''' Multi-model (ensemble) mean constructed with one realization of the CMIP6 historical experiment from each model. '''(b)''' Multi-model mean bias, defined as the difference between the CMIP6 multi-model mean and the precipitation analysis from the Global Precipitation Climatology Project (GPCP) version 2.3 ([[#Adler--2003|Adler et al., 2003]]). '''(c)''' Multi-model mean of the root mean square error calculated over all months separately and averaged with respect to the precipitation analysis from GPCP version 2.3. '''(d)''' Multi-model mean bias, calculated as the difference between the CMIP6 multi-model mean and the precipitation analysis from GPCP version 2.3. Also shown is the multi-model mean bias as the difference between the multi-model mean of '''(e)''' high resolution and '''(f)''' low-resolution simulations of four HighResMIP models and the precipitation analyses from GPCP version 2.3. Uncertainty is represented using the advanced approach. No overlay indicates regions with robust signal, where ≥66% of models show change greater than the variability threshold and ≥80% of all models agree on sign of change; diagonal lines indicate regions with no change or no robust signal, where &lt;66% of models show a change greater than the variability threshold; crossed lines indicate regions with conflicting signal, where ≥66% of models show change greater than the variability threshold and &lt;80% of all models agree on the sign of change. For more information on the advanced approach, please refer to the Cross-Chapter Box Atlas.1. Dots in panel (e) mark areas where the bias in high resolution versions of the HighResMIP models is not lower in at least three out of four models than in the corresponding low-resolution versions. Further details on data sources and processing are available in the chapter data table (Table 3.SM.1).

Recent studies comparing observations and CMIP5 simulations have shown that tropical volcanic eruptions induce a significant reduction in global precipitation, particularly over the wet tropics, including the global monsoon regions ([[#Iles--2014|Iles and Hegerl, 2014]] ; [[#Paik--2017|Paik and Min, 2017]] ; [[#Paik--2020a|Paik et al., 2020a]]). Reconstructions and modelling studies also suggest a distinct remote influence of volcanic forcing such that large volcanoes erupting in one hemisphere can enhance monsoon precipitation in the other hemisphere (F. [[#Liu--2016|]] [[#Liu--2016|Liu et al., 2016]] ; [[#Zuo--2019|Zuo et al., 2019]]). The climatic effect of volcanic eruptions is further assessed in Cross-Chapter Box 4.1.

An intensification of the wet–dry zonal mean patterns, consisting of the wet tropical and mid-latitude bands becoming wetter, and the dry subtopics becoming drier is expected in response to greenhouse gas and ozone changes (Section 8.2.2.2). However, detecting these changes is complicated by model errors in locating the main features of rainfall patterns. To deal with this issue, [[#Marvel--2013|Marvel and Bonfils (2013)]] identified in each CMIP5 historical simulation the latitudinal peaks and troughs of the rainfall latitudinal patterns, measured the amplification and shift of these patterns in a pattern-based fingerprinting study, and found that the simultaneous amplification and shift in zonal precipitation patterns are detectable in Global Precipitation Climatology Project (GPCP) observations over the 1979–2012 period. Similarly, [[#Bonfils--2020|Bonfils et al. (2020)]] found that the intensification of wet–dry zonal patterns identified in CMIP5 historical simulations is detectable in reanalyses over the 1950–2014 period (see also Figure 8.11).

Based on long-term island precipitation records, [[#Polson--2016|Polson et al. (2016)]] identified significant increases in precipitation in the tropics and decreases in the subtropics, which are consistent with those simulated by the CMIP5 models. Moreover, results from [[#Polson--2017|Polson and Hegerl (2017)]] give support to an intensification of the water cycle according to the wet-gets-wetter, dry-gets-drier paradigm over tropical land areas as well. Other studies suggest that this paradigm does not necessarily hold over dry regions where moisture is limited (see also Section 8.2.2.1; [[#Greve--2014|Greve et al., 2014]] ; [[#Kumar--2015|Kumar et al., 2015]]). [[#Polson--2017|Polson and Hegerl (2017)]] explained this discrepancy by taking into account the seasonal and interannual movement of the regions ([[#Allan--2014|Allan, 2014]]). A follow-up study using CMIP6 models also found that the observed strengthening contrast of precipitation over wet and dry regions was detectable, although the increase was significantly larger in observations than in the multi-model mean. The change was attributed to a combination of anthropogenic and natural forcings, with anthropogenic forcings detectable in multi-signal analyses (Figure 3.14; [[#Schurer--2020|Schurer et al., 2020]]).

<div id="_idContainer036" class="•-2-columns"></div>

[[File:8946db0b175de9fa41f10aaefd9df3d4 IPCC_AR6_WGI_Figure_3_14.png]]

Figure 3.14 | '''Wet (a) and dry (b) region tropical mean (30°S–30°N) annual precipitation anomalies.''' Observed data are shown with black lines (GPCP), ERA5 reanalysis is shown in grey, single model simulations are shown with light blue/red lines (CMIP6), and multi-model mean results are shown with dark blue/red lines (CMIP6). Wet and dry region annual anomalies are calculated as the running mean over 12 months relative to a 1988–2020 base period. The regions are defined as the wettest third and driest third of the surface area, calculated for the observations and for each model separately for each season (following [[#Polson--2017|Polson and Hegerl, 2017]]). Scaling factors '''(c, d)''' are calculated for the combination of the wet and dry region mean, where the observations, reanalysis and all the model simulations are first standardized using the mean standard deviation of the pre-industrial control simulations. Two total least squares regression methods are used: noise in variables (following [[#Polson--2017|Polson and Hegerl, 2017]]) which estimates a best estimate and a 5–95% confidence interval using the pre-industrial controls (circle and thick green line) and the pre-industrial controls with double the variance (thin green line); and a bootstrap method ([[#DelSole--2019|DelSole et al., 2019]]) (5–95% confidence interval shown with a purple line and best estimate with a circle). Panel (c) shows results for GPCP and panel (d) for ERA5. Figure is adapted from [[#Schurer--2020|Schurer et al. (2020)]] . Further details on data sources and processing are available in the chapter data table (Table 3.SM.1).

Global land precipitation has ''likely'' increased since the middle of the 20th century (''medium confidence''), while there is ''low confidence'' in trends in land data prior to 1950 and over the ocean during the satellite era due to disagreement between datasets ([[IPCC:Wg1:Chapter:Chapter-2#2.3.1.3.4|Section 2.3.1.3.4]]). Figure 3.15a shows the time evolution of the global mean land precipitation since 1950, as well as the trend during the period. [[#Adler--2017|Adler et al. (2017)]] found no significant trend in the global mean precipitation during the satellite era, consistent with model simulations ([[#Wu--2013|Wu et al., 2013]]) and physical understanding of the energy budget (Section 8.2.1). This has been suggested to be due to the negative effect of anthropogenic sulphate aerosol that opposed the positive influence of rising global mean temperatures due to greenhouse gases ([[#Salzmann--2016|Salzmann, 2016]] ; [[#Richardson--2018|Richardson et al., 2018]]). The precipitation change expected from ocean warming is also partly offset by the fast atmospheric adjustment to increasing greenhouse gases (Section 8.2.1). Over the ocean, the negligible trend may be due to the cancelling effects of CO <sub>2</sub> and aerosols ([[#Richardson--2018|Richardson et al., 2018]]).

<div id="_idContainer038" class="•-2-columns"></div>

[[File:6965d17c0b5e1e5bdf038a871bbd85eb IPCC_AR6_WGI_Figure_3_15.png]]

Figure 3.15 | '''Observed and simulated time series of anomalies in zonal average annual mean precipitation. (a), (c–f)''' Evolution of global and zonal average annual mean precipitation (mm day <sup>–1</sup>) over areas of land where there are observations, expressed relative to the base period of 1961–1990, simulated by CMIP6 models (one ensemble member per model) forced with both anthropogenic and natural forcings (brown) and natural forcings only (green). Multi-model means are shown in thick solid lines and shading shows the 5–95% confidence interval of the individual model simulations. The data is smoothed using a low pass filter. Observations from three different datasets are included: gridded values derived from Global Historical Climatology Network (GHCN version 2) station data, updated from [[#Zhang--2007|Zhang et al. (2007)]] , data from the Global Precipitation Climatology Product (GPCP L3 version 2.3, [[#Adler--2003|Adler et al. (2003)]]) and from the Climate Research Unit (CRU TS4.02, [[#Harris--2014|Harris et al. (2014)]]). Also plotted are boxplots showing interquartile and 5–95% ranges of simulated trends over the period for simulations forced with both anthropogenic and natural forcings (brown) and natural forcings only (blue). Observed trends for each observational product are shown as horizontal lines. Panel (b) shows annual mean precipitation rate (mm day <sup>–1</sup>) of GHCN version 2 for the years 1950–2014 over land areas used to compute the plots. Further details on data sources and processing are available in the chapter data table (Table 3.SM.1).

A gridpoint based analysis of annual precipitation trends over land regions since 1901 ([[#Knutson--2018|Knutson and Zeng, 2018]]) comparing observed and simulated trends found that detectable anthropogenic increasing trends have occurred prominently over many mid- to high-latitude regions of the Northern Hemisphere and subtropics of the Southern Hemisphere. The observed trends in many cases are significantly stronger than modelled in the CMIP5 historical runs for the 1901–2010 period (though not for 1951–2010), which may be due to disagreement between observed datasets ([[IPCC:Wg1:Chapter:Chapter-2#2.3.1.3.4|Section 2.3.1.3.4]]), and/or suggest possible deficiencies in models.

The observed precipitation increase in the Northern Hemisphere high latitudes over the period 1966–2005 was attributed to anthropogenic forcing by a study using CMIP5 models ([[#Wan--2015|Wan et al., 2015]]) supporting the AR5 assessment. Initial results from CMIP6 also support the role of anthropogenic forcing in the precipitation increase observed in Northern Hemisphere high latitudes (see Figure 3.15c): the observed positive trend detected for the band 60°N–90°N can only be reproduced when anthropogenic forcing is included, although models tend to simulate overall a larger positive trend. A similar positive trend, but less significant, is also detected between 30°N–60°N, while in the southern mid-latitudes no trend is simulated (see Figure 3.15d, f).

For the Southern Hemisphere extratropics, [[#Solman--2016|Solman and Orlanski (2016)]] found that the observed summertime rainfall increase over high latitudes and decrease over mid-latitudes over the period 1979–2010 are quasi-zonally symmetric and related to changes in eddy activity. The latter were in turn found to be associated with the poleward shift of the westerlies due mostly to ozone depletion. Positive rainfall trends in the subtropics, particularly over south-eastern South America (see also Section 10.4.2.2) and northern and central Australia, have been also attributed to stratospheric ozone depletion ([[#Kang--2011|Kang et al., 2011]] ; [[#Gonzalez--2014|Gonzalez et al., 2014]]) and greenhouse gases ([[#Vera--2015|Vera and Díaz, 2015]] ; [[#Saurral--2019|Saurral et al., 2019]]). During austral winter, wetting at high latitudes and drying at mid-latitudes are not zonally homogenous, due to both changes in eddy activity and increased lower troposphere humidity. [[#Solman--2016|Solman and Orlanski (2016)]] associated these climate changes with increases in greenhouse gas concentration levels. Recently, [[#Blazquez--2017|Blazquez and Solman (2017)]] have shown that CMIP5 models represent very well the dynamical forcing and the frequency of frontal precipitation in the Southern Hemisphere winter extratropics, but the amount of precipitation due to fronts is overestimated. Chapters 10 and 11 validate in more detail the simulation of fronts in climate models (Sections 10.3.3.4.4 and 11.7.2.3).

Over the ocean, observations show coherent large-scale patterns of fresh ocean regions becoming fresher and salty ocean regions saltier across the globe, which has been related through modelling studies to changes in precipitation minus evaporation and is consistent with the wet-gets-wetter, dry-gets-drier paradigm (see Sections 3.5.2.2 and 8.2.2.1; [[#Durack--2012|Durack et al., 2012]] , 2013; [[#Skliris--2014|Skliris et al., 2014]] ; [[#Durack--2015|Durack, 2015]] ; [[#Hegerl--2015|Hegerl et al., 2015]] ; [[#Levang--2015|Levang and Schmitt, 2015]] ; [[#Zika--2015|Zika et al., 2015]] ; [[#Grist--2016|Grist et al., 2016]] ; [[#Cheng--2020|Cheng et al., 2020]]).

Overall, studies published since AR5 provide further evidence of an anthropogenic influence on precipitation, and therefore we now assess that it is ''likely'' that human influence has contributed to large-scale precipitation changes observed since the mid-20th century. New attribution studies strengthen previous findings of a detectable increase in mid to high latitude land precipitation over the Northern Hemisphere (''high confidence''). There is ''medium confidence'' that human influence has contributed to a strengthening of the zonal mean wet tropics-dry subtropics contrast, and that tropical rainfall changes follow the wet-gets-wetter, dry-gets-drier paradigm. There is also ''medium confidence'' that ozone depletion has increased precipitation over the southern high latitudes and decreased it over southern mid-latitudes during austral summer. Owing to observational uncertainties and inconsistent results between studies, we conclude that there is ''low confidence'' in the attribution of changes in the seasonality of precipitation.

<div id="3.3.2.4" class="h3-container"></div>

<span id="streamflow"></span>
==== 3.3.2.4 Streamflow ====

<div id="h3-8-siblings" class="h3-siblings"></div>

Streamflow is to-date the only variable of the terrestrial water cycle with enough in-situ observations to allow for detection and attribution analysis at continental to global scales. Based on evidence from a few formal detection and attribution studies, particularly on the timing of peak streamflow, and the qualitative evaluation of studies reporting on observed and simulated trends, AR5 concluded that there is ''medium confidence'' that anthropogenic influence on climate has affected streamflow in some middle and high latitude regions ([[#Bindoff--2013|Bindoff et al., 2013]]). The AR5 also noted that observational uncertainties are large and that often only a limited number of models were considered.

([[IPCC:Wg1:Chapter:Chapter-2#2.3.1.3.6|Section 2.3.1.3.6]] assesses that there have not been significant trends in global average streamflow over the last century, though regional trends have been observed, driven in part by internal variability. Only a limited number of studies have systematically compared observed streamflow trends at continental to global scales with changes simulated by global circulation models in a detection and attribution setting. H. [[#Yang--2017|]] [[#Yang--2017|Yang et al. (2017)]] did not find a significant correlation between observed runoff changes and changes simulated in CMIP5 models in most grid cells, consistent with the assessment that observed changes are dominated by internal variability. In a pan-European assessment, [[#Gudmundsson--2017|Gudmundsson et al. (2017)]] attributed the spatio-temporal pattern of decreasing streamflow in southern Europe and increasing streamflow in northern Europe to anthropogenic climate change, but also concluded that additional effects of human water withdrawals could not be excluded. Focussing on continental runoff between 1958 and 2004, [[#Alkama--2013|Alkama et al. (2013)]] found a significant change only when using reconstructed data over all rivers, and a large uncertainty in the estimate of the global streamflow trend due to opposing changes over different continents. [[#Gedney--2014|Gedney et al. (2014)]] detected the influence of aerosols on streamflow in North America and Europe, with aerosols having driven an increase in streamflow due to reduced evaporation (see Section 8.3.1.5 for details on processes). There is also evidence for a detectable anthropogenic contribution toward earlier winter-spring streamflows in the north-central US ([[#Kam--2018|Kam et al., 2018]]) and in western Canada ([[#Najafi--2017|Najafi et al., 2017]]). From a model evaluation perspective, [[#Sheffield--2013|Sheffield et al. (2013)]] reported that CMIP5 models reproduce spatial variations in runoff in North America well, though they tend to underestimate it.

Recently, [[#Gudmundsson--2021|Gudmundsson et al. (2021)]] performed a global detection and attribution study on streamflow and found that some regions are drying and others are wetting. Moreover, the simulated streamflow trends are consistent with observations only if externally forced climate change is considered, and the simulated effects of water and land management cannot reproduce the observed trends. The effects of volcanic eruptions in driving reduced streamflow have also been detected in the wet tropics ([[#Iles--2015|Iles and Hegerl, 2015]] ; [[#Zuo--2019|Zuo et al., 2019]]).

In summary, there is ''medium confidence'' that anthropogenic climate change has altered local and regional streamflow in various parts of the world and that the associated global-scale trend pattern is inconsistent with internal variability. Moreover, human interventions and water withdrawals, while affecting streamflow, cannot explain the observed spatio-temporal trends (''medium confidence'').

<div id="cross-chapter-box-3.2" class="h2-container box-container"></div>
<div class="container-box col-cross">