Editing IPCC:AR6/WGI/Chapter-2 (section)

=== 2.3.1 Atmosphere and Earth’s Surface ===

<div id="h2-14-siblings" class="h2-siblings"></div>

<div id="2.3.1.1" class="h3-container"></div>

<span id="surface-temperatures"></span>
==== 2.3.1.1 Surface Temperatures ====

<div id="h3-12-siblings" class="h3-siblings"></div>

<div id="2.3.1.1.1" class="h4-container"></div>

<span id="temperatures-of-the-deep-past-65-ma-to-8-ka"></span>
===== 2.3.1.1.1 Temperatures of the deep past (65 Ma to 8 ka) =====

<div id="h4-7-siblings" class="h4-siblings"></div>

This assessment of the paleo reference periods (Cross-Chapter Box 2.1) draws from studies based mostly or entirely on indirect observational evidence from geological archives (i.e., proxy records) rather than reconstructions that rely more heavily on modelled parameters and those based on deep-ocean temperatures (e.g., [[#Köhler--2015|Köhler et al., 2015]] ; [[#Friedrich--2016|Friedrich et al., 2016]] ). In contrast to AR5, temperature estimates from climate models are not included in the assessed values for paleo reference periods in this chapter. The AR5 concluded that the reconstructed GMST during the PETM was 4°C–7°C warmer than pre-PETM mean climate ( ''low confidence'' ), and that the EECO and the MPWP were 9°C–14°C and 1.9°C–3.6°C warmer than pre-industrial, respectively ( ''medium confidence'' ). The GMST during the LIG was assessed at 1°C–2°C warmer than pre-industrial ''(medium confidence'' ), whereas SROCC narrowed the range to 0.5°C–1.0°C warmer, but did not state a confidence level. The AR5 further concluded that it was ''very'' ''likely'' that the LGM was 3°C–8°C colder than pre-industrial, and ''likely'' that the maximum rate of global warming during the subsequent deglacial period was 1°C–1.5°C kyr <sup>–1</sup> .

For the PETM, new reconstructions agree with those assessed by AR5. A major new compilation of proxy temperature data ( [[#Hollis--2019|Hollis et al., 2019]] ) analysed using multiple statistical approaches ( [[#Inglis--2020|Inglis et al., 2020]] ) indicates that GMST was 10°C–25°C (90% range) warmer than 1850–1900, or about 5°C warmer relative to the pre-PETM state. A related synthesis study also estimates that PETM warmed by 5°C (no uncertainty assigned; [[#Zhu--2019|Zhu et al., 2019]] ). A recent benthic isotope compilation ( [[#Westerhold--2020|Westerhold et al., 2020]] ) transformed to GMST based on the formulation by J. Hansen et al. (2013; Cross-Chapter Box 2.1, Figure 1), and adjusted to 1850–1900 by adding 0.36°C, shows an increase of GMST by about 10°C during the PETM. This reflects the expected higher variability at single sites that were used to splice together the composite time series, compared to the globally averaged composite time series of [[#Zachos--2008|Zachos et al. (2008)]] . The latter was originally used by J. [[#Hansen--2013|]] [[#Hansen--2013|Hansen et al. (2013)]] to reconstruct GMST, and is the preferred representation of the global average bottom water conditions, despite its less well-refined chronology.

For the EECO, new GMST reconstructions fall at the high end of the range assessed by AR5. These include estimates of 7°C–18°C (90% range; [[#Inglis--2020|Inglis et al., 2020]] ) and 12°C–18°C (95% range; [[#Zhu--2019|Zhu et al., 2019]] ) warmer than 1850–1900, and 10°C–16°C warmer than 1995–2014 ‘recent past’ conditions (2 standard error range; [[#Caballero--2013|Caballero and Huber, 2013]] ). Together, they indicate that GMST was 10°C–18°C warmer during the EECO compared with 1850–1900 ( ''medium confidence'' ).

The AR5 did not assess the GMST for the MCO. Reconstructions based on data from multiple study sites include estimates of about 4°C (uncertainty range not specified; [[#You--2009|You et al., 2009]] ) and 5°C–10°C (2 standard error range; [[#Goldner--2014|Goldner et al., 2014]] ) warmer than 1850–1900. Together, these studies indicate that GMST was 4°C–10°C warmer during the MCO ( ''medium confidence'' ).

For the MPWP, new proxy-based estimates of global sea surface temperatures (SST) are about 2.0°C–3.5°C warmer than 1850–1900, depending on which proxy types are included in the analysis ( [[#Foley--2019|Foley and Dowsett, 2019]] ; [[#McClymont--2020|McClymont et al., 2020]] ). On the basis of model-derived relationships between land versus sea surface temperatures under different climate states (Figure 3.2b), the increase in GMST is estimated to have been roughly 15% greater than the increase in global SST. Therefore, GMST during the MPWP is estimated to have been 2.5°C–4.0°C warmer than 1850–1900 ( ''medium confidenc'' e).

For the LIG (Cross-Chapter Box 2.1, Figure 1, and Figure 2.11), a major new compilation of marine proxy data ( [[#Turney--2020|Turney et al., 2020]] ) from 203 sites indicates that the average SST from 129–125 ka was 1.0°C ± 0.2°C (2 SD) warmer than 1850–1900 (reported relative to 1981–2010 and adjusted here by 0.8°C). These temperatures represent the time of peak warmth, which may not have been synchronous among these sites. This compares with two other SST estimates for 125 ka of 0.5°C ± 0.3°C (± 2 SD) warmer at 125 ka relative to 1870–1889 ( [[#Hoffman--2017|Hoffman et al., 2017]] ), and about 1.4°C (no uncertainty stated) warmer at 125 ka relative to 1850–1900 ( [[#Friedrich--2020|Friedrich and Timmermann, 2020]] ; reported relative to 10–5 ka and adjusted here by 0.4°C; [[#Kaufman--2020a|Kaufman et al., 2020a]] ). The average of these post-AR5 global SST anomalies is 1°C. Commensurately (Figure 3.2b), GMST is estimated to have been roughly 1.1°C above 1850–1900 values, although this value could be too high if peak warmth was not globally synchronous ( [[#Capron--2017|Capron et al., 2017]] ). A further estimate of peak GMST anomalies of 1.0°C–3.5°C (90% range; adjusted here to 1850–1900 by adding 0.2°C) based on 59 marine sediment cores ( [[#Snyder--2016|Snyder, 2016]] ) is considerably warmer than remaining estimates and is therefore given less weight in the final assessment. The warmest millennium of the LIG GMST reconstruction in J. [[#Hansen--2013|]] [[#Hansen--2013|Hansen et al. (2013)]] is 1.5°C above 1850–1900. In summary, GMST during the warmest millennia of the LIG (within the interval of around 129–125 ka) is estimated to have reached 0.5°C–1.5°C higher values than the 1850–1990 reference period ( ''medium confidence'' ).

<div id="_idContainer030" class="Basic-Text-Frame"></div>

[[File:db18d2c58474b72f5157a77d07c68327 IPCC_AR6_WGI_Figure_2_11.png]]

'''Figure''' '''2.11 |''' '''Earth’s surface temperature history with key findings annotated within each panel. (a)''' GMST over the Holocene divided into three time scales: (i) 12 kyr–1 kyr in 100-year time steps; (ii) 1000–1900 CE, 10-year smooth; and (iii) 1900–2020 CE (from panel (c)). Median of the multi-method reconstruction (bold lines), with 5th and 95th percentiles of the ensemble members (thin lines). Vertical bars are the assessed ''medium confidence'' ranges of GMST for the Last Interglacial and mid-Holocene ( [[#2.3.1.1|Section 2.3.1.1]] ). The last decade value and ''very likely'' range arises from [[#2.3.1.1.3|Section 2.3.1.1.3]] . '''(b)''' Spatially resolved trends (°C per decade) for HadCRUTv5 over (upper map) 1900–1980, and (lower map) 1981–2020. Significance is assessed following AR(1) adjustment after [[#Santer--2008|Santer et al. (2008)]] , ‘×’ marks denote non-significant trends. '''(c)''' Temperature from instrumental data for 1850–2020, including (upper panel) multi-product mean annual time series assessed in [[#2.3.1.1.3|Section 2.3.1.1.3]] for temperature over the oceans (blue line) and temperature over the land (red line) and indicating the warming to the most recent 10 years; and annually (middle panel) and decadally (bottom panel) resolved averages for the GMST datasets assessed in [[#2.3.1.1.3|Section 2.3.1.1.3]] . The grey shading in each panel shows the uncertainty associated with the HadCRUT5 estimate ( [[#Morice--2021|Morice et al., 2021]] ). All temperatures relative to the 1850–1900 reference period. Further details on data sources and processing are available in the chapter data table (Table 2.SM.1).

New GMST reconstructions for the LGM fall near the middle of AR5’s ''very likely'' range, which was based on a combination of proxy reconstructions and model simulations. Two of these new reconstructions use marine proxies to reconstruct global SST that were scaled to GMST based on different assumptions. One indicates that GMST was 6.2 [4.5 to 8.1] °C cooler than the late Holocene average ( [[#Snyder--2016|Snyder, 2016]] ), and the other, 5.7°C ± 0.8°C (2 SD) cooler than the average of the first part of the Holocene (10–5 ka) ( [[#Friedrich--2020|Friedrich and Timmermann, 2020]] ). A third new estimate ( [[#Tierney--2020|Tierney et al., 2020]] ) uses a much larger compilation of marine proxies along with a data-assimilation procedure, rather than scaling, to reconstruct a GMST of 6.1°C ± 0.4°C (2 SD) cooler than the late Holocene. Assuming that the 1850–1900 reference period was 0.2°C and 0.4°C cooler than the late and first part of the Holocene, respectively ( [[#Kaufman--2020a|Kaufman et al., 2020a]] ), the midpoints of these three new GMST reconstructions average –5.8°C relative to 1850–1900. The coldest multi-century period of the LGM in the J. [[#Hansen--2013|]] [[#Hansen--2013|Hansen et al. (2013)]] reconstruction is 4.3°C colder than 1850–1900. This compares to land- and SST-only estimates of about –6.1°C ± 2°C and –2.2°C ± 1°C, respectively (2 SD), which are based on AR5-generation studies that imply a warmer GMST than more recent reconstructions (Figure 1c in [[#Harrison--2015|Harrison et al., 2015]] ; Figure 7 in [[#Harrison--2016|Harrison et al., 2016]] ). A major new pollen-based data-assimilation reconstruction averages 6.9°C cooler over northern extratropical land ( [[#Cleator--2020|Cleator et al., 2020]] ). LGM temperature variability on centennial scales was about four times higher globally than during the Holocene, and even greater at high latitudes ( [[#Rehfeld--2018|Rehfeld et al., 2018]] ). In summary, GMST is estimated to have been 5°C–7°C lower during the LGM (around 23–19 ka) compared with 1850–1900 ( ''medium confidence'' ).

For the LDT (Cross-Chapter Box 2.1, Figure 1), no new large-scale studies have been published since AR5 ( [[#Shakun--2012|Shakun et al., 2012]] ) to further assess the rate of GMST change during this period of rapid global warming (estimated at 1°C–1.5°C per kyr). The reconstruction of [[#Shakun--2012|Shakun et al. (2012)]] was based primarily on SST records and therefore underrepresents the change in GMST during the LDT. Temperature over Greenland increased by about ten times that rate during the centuries of most rapid warming ( [[#Jansen--2020|Jansen et al., 2020]] ).

<div id="2.3.1.1.2" class="h4-container"></div>

<span id="temperatures-of-the-post-glacial-period-past-7000-years"></span>
===== 2.3.1.1.2 Temperatures of the post-glacial period (past 7000 years) =====

<div id="h4-8-siblings" class="h4-siblings"></div>

The AR5 did not include an assessment of large-scale temperature estimates for the MH, although it assigned ''high confidence'' to the long-term cooling trend over mid- to high-latitudes of the Northern Hemisphere (NH) during the 5 kyr that preceded recent warming. For average annual NH temperatures, the period 1983–2012 was assessed as ''very likely'' the warmest 30-year period of the past 800 years ( ''high confidence'' ) and ''likely'' the warmest 30-year period of the past 1.4 kyr ( ''medium confidence'' ); the warm multi-decadal periods prior to the 20th century were unsynchronized across regions, in contrast to the warming since the mid-20th century ( ''high confidence'' ), although only sparse information was available from the SH.

This section concerns the Holocene period prior to industrialization when GMST was overall highest. Whereas SR1.5 focussed upon the ‘Holocene thermal maximum’ when regional temperatures were up to 1°C higher than 1850–1900, though peak warming occurred regionally at different times between around 10 and 5 ka greatly complicating interpretation. A multi-method reconstruction ( [[#Kaufman--2020a|Kaufman et al., 2020a]] ) based on a quality-controlled, multi-proxy synthesis of paleo-temperature records from 470 terrestrial and 209 marine sites globally ( [[#Kaufman--2020b|Kaufman et al., 2020b]] ) indicates that the median GMST of the warmest two-century-long interval was 0.7 [0.3 to 1.8] °C warmer than 1800–1900 (which averaged 0.03°C colder than 1850–1900; [[#PAGES%202k%20Consortium--2019|PAGES 2k Consortium, 2019]] ), and was centred around 6.5 ka. This is similar to [[#Marcott--2013|Marcott et al. (2013)]] , which is based on a smaller dataset (73 sites) and different procedures to estimate a maximum warmth of 0.8°C ± 0.3°C (2 SD) at around 7.0 ka, adjusted here by adding 0.3°C to account for differences in reference periods. These may be underestimates because averaging inherently smoothed proxy records with uncertain chronologies reduces the variability in the temperature reconstruction (e.g., Dolman and Laepple, (2018) for sedimentary archives). However, the general coincidence between peak warmth and astronomically driven boreal summer insolation might reflect a bias toward summer conditions ( [[#Liu--2014|Liu et al., 2014]] ; [[#Hou--2019|Hou et al., 2019]] ; [[#Bova--2021|Bova et al., 2021]] ), suggesting that the estimate is too high. This possibility is supported by AR5-generation proxy data focusing on 6 ka ( [[#Harrison--2014|Harrison et al., 2014]] ), the long-standing MH modelling target (Cross-Chapter Box 2.1), that indicate surface temperatures for land and ocean were indistinguishable from ‘pre-industrial’ climate (Figure 1c in [[#Harrison--2015|Harrison et al., 2015]] ; Figure 7 in [[#Harrison--2016|Harrison et al., 2016]] ). In contrast, the GMST estimate from the multi-method global reconstruction ( [[#Kaufman--2020a|Kaufman et al., 2020a]] ) for the millennium centred on 6 ka is only about 0.1°C colder than the warmest millennium.

Taking all lines of evidence into account, the GMST averaged over the warmest centuries of the current interglacial period (sometime between around 6 and 7 ka) is estimated to have been 0.2°C–1.0°C higher than 1850–1900 ( ''medium confidence'' ). It is therefore ''more likely than not'' that no multi-centennial interval during the post-glacial period was warmer globally than the most recent decade (which was 1.1°C warmer than 1850–1900; [[#2.3.1.1.3|Section 2.3.1.1.3]] ); the LIG (129–116 ka) is the next most recent candidate for a period of higher global temperature. Zonally averaged mean annual temperature reconstructions ( [[#Routson--2019|Routson et al., 2019]] ) indicate that MH warmth was most pronounced north of 30°N latitude, and that GMST subsequently decreased in general, albeit with multi-century variability, with greater cooling in the NH than in the SH ( [[#Kaufman--2020a|Kaufman et al., 2020a]] ).

The temperature history of the last millennium and the methods used to reconstruct it have been studied extensively, both prior to and following AR5, as summarized recently by [[#Smerdon--2016|Smerdon and Pollack (2016)]] and [[#Christiansen--2017|Christiansen and Ljungqvist (2017)]] . New regional (e.g., [[#Shi--2015|Shi et al., 2015]] ; [[#Stenni--2017|Stenni et al., 2017]] ; [[#Werner--2018|Werner et al., 2018]] ), global ocean ( [[#McGregor--2015|McGregor et al., 2015]] ), quasi-hemispheric ( [[#Neukom--2014|Neukom et al., 2014]] ; [[#Schneider--2015|Schneider et al., 2015]] ; [[#Anchukaitis--2017|Anchukaitis et al., 2017]] ), and global ( [[#Tardif--2019|Tardif et al., 2019]] ) temperature reconstructions, and new regional proxy data syntheses ( [[#Lüning--2019a|Lüning et al., 2019a]] , b) have been published, extending back 1–2 kyr. In addition, a major new global compilation of multiproxy, annually resolved paleo-temperature records for the CE ( [[#PAGES%202k%20Consortium--2017|PAGES 2k Consortium, 2017]] ) has been analysed using a variety of statistical methods for reconstructing temperature ( [[#PAGES%202k%20Consortium--2019|PAGES 2k Consortium, 2019]] ). The median of the multi-method GMST reconstruction from this synthesis (Figure 2.11a) generally agrees with the AR5 assessment, while affording more robust estimates of the following major features of GMST during the CE: (i) an overall millennial-scale cooling trend of –0.18 [–0.28 to 0.00] °C kyr <sup>–1</sup> prior to 1850; (ii) a multi-centennial period of relatively low temperature beginning around the 15th century, with GMST averaging –0.03 [–0.30 to 0.06] °C between 1450 and 1850 relative to 1850–1900; (iii) the warmest multi-decadal period occurring most recently; and (iv) the rate of warming during the second half of the 20th century (from instrumental data) exceeding the 99th percentile of all 51-year trends over the past 2 kyr. Moreover, the new proxy data compilation shows that the warming of the 20th century was more spatially uniform than any other century-scale temperature change of the CE ( ''medium confidence'' ) ( [[#Neukom--2019|Neukom et al., 2019]] ). A new independent temperature reconstruction extending back to 1580 is based on an expanded database of subsurface borehole temperature profiles, along with refined methods for inverse modelling ( [[#Cuesta-Valero--2021|Cuesta-Valero et al., 2021]] ). The borehole data, converted to GMST based on the modelled relation between changes in land versus sea surface temperature outlined previously, indicate that average GMST for 1600–1650 was 0.12°C colder than 1850–1900, which is similar to the PAGES 2k reconstruction (0.09°C colder), although both estimates are associated with relatively large uncertainties (0.8°C (95% range) and 0.5°C (90% range), respectively).

To conclude, following approximately 6 ka, GMST generally decreased, culminating in the coldest multi-century interval of the post-glacial period (since 8 ka), which occurred between around 1450 and 1850 ( ''high confidence'' ). This multi-millennial cooling trend was reversed in the mid-19th century. Since around 1950, GMST has increased at an observed rate unprecedented for any 50-year period in at least the last 2000 years ( ''high confidence'' ).

<div id="2.3.1.1.3" class="h4-container"></div>

<span id="temperatures-during-the-instrumental-period-surface"></span>
===== 2.3.1.1.3 Temperatures during the instrumental period – surface =====

<div id="h4-9-siblings" class="h4-siblings"></div>

The AR5 concluded that it was certain that GMST had increased since the late 19th century. Total warming in GMST was assessed as 0.85 [0.65 to 1.06] °C over 1880–2012, while the change from 1850–1900 to 2003–2012 was assessed at 0.78 [0.72 to 0.85] °C, and from 1850–1900 to 1986–2005 at 0.61 [0.55 to 0.67] °C. The SR1.5 reported warming of GMST from 1850–1900 to 2006–2015 of 0.87°C, with an 1880–2012 trend of 0.86°C and an 1880–2015 trend of 0.92°C. The SRCCL concluded that since the pre-industrial period, surface air temperature over land areas has risen nearly twice as much as the global mean surface temperature ( ''high confidence'' ).

Since AR5, there have been substantial improvements in the availability of instrumental archive data both over the ocean and on land. A new version of the International Comprehensive Ocean-Atmosphere Dataset (ICOADS Release 3.0, [[#Freeman--2017|Freeman et al., 2017]] ) comprises over 450 million in situ marine reports and incorporates newly digitized data, increasing coverage in data sparse regions and times (e.g., polar oceans and World War I). The International Surface Temperature Initiative released a much improved collection of fundamental land surface air temperature records ( [[#Rennie--2014|Rennie et al., 2014]] ) comprising more than 35,000 station records. These advances, both of which have substantially improved spatial coverage, have reduced uncertainties in assessments of both land and marine data.

<span id="marine-domain"></span>
====== Marine domain ======

For SST analyses, three products – HadSST4 (1850–present, [[#Kennedy--2019|Kennedy et al., 2019]] ), ERSSTv5 (1850–present, [[#Huang--2017|Huang et al., 2017]] ) and COBE SST2 (1880–present, ( [[#Hirahara--2014|Hirahara et al., 2014]] ) – now have bias adjustments applied throughout the record. The new SST datasets account for two major issues previously identified in AR5: that globally averaged buoy SSTs are about 0.12°C cooler than ship-based SSTs ( [[#Kennedy--2011|Kennedy et al., 2011]] ; [[#Huang--2015|Huang et al., 2015]] ), and that SSTs from ship engine room intakes may have biases for individual ships depending upon the sensor set-up ( [[#Kent--2006|Kent and Kaplan, 2006]] ) but have an overall warm bias when globally aggregated ( [[#Kennedy--2019|Kennedy et al., 2019]] ). The first issue primarily affects data since 1990, when buoys began to increasingly contribute to the observation network ( [[#Woodruff--2011|Woodruff et al., 2011]] ), and the second issue has its largest effect from the 1940s to the 1970s. From the standpoint of uncertainty, ERSSTv4 (W. [[#Liu--2015|]] [[#Liu--2015|Liu et al., 2015]] ; [[#Huang--2016|Huang et al., 2016]] ) and subsequent versions ( [[#Huang--2017|Huang et al., 2017]] ), and HadSST4 have estimates presented as ensembles that sample parametric uncertainty. Comparisons between these independently-derived analyses and the assessed uncertainties ( [[#Kennedy--2014|Kennedy, 2014]] ; [[#Kent--2017|Kent et al., 2017]] ) show unambiguously that global mean SST increased since the start of the 20th century, a conclusion that is insensitive to the method used to treat gaps in data coverage ( [[#Kennedy--2014|Kennedy, 2014]] ).

A number of recent studies also corroborate important components of the SST record ( [[#Hausfather--2017|Hausfather et al., 2017]] ; [[#Kent--2017|Kent et al., 2017]] ; [[#Cowtan--2018|Cowtan et al., 2018]] ; [[#Kennedy--2019|Kennedy et al., 2019]] ). In particular, ATSR SST satellite retrievals ( [[#Merchant--2012|Merchant et al., 2012]] ; [[#Berry--2018|Berry et al., 2018]] ), the near-surface records from hydrographical profiles ( [[#Gouretski--2012|Gouretski et al., 2012]] ; [[#Huang--2018|Huang et al., 2018]] ), and coastal observations ( [[#Cowtan--2018|Cowtan et al., 2018]] ) have all been shown to be broadly consistent with the homogenized SST analyses. [[#Hausfather--2017|Hausfather et al. (2017)]] also confirmed the new estimate of the rate of warming seen in ERSSTv4 since the late 1990s through comparison with independent SST data sources such as Argo floats and satellite retrievals. Nevertheless, dataset differences remain in the mid-20th century when there were major, poorly-documented, changes in instrumentation and observational practices ( [[#Kent--2017|Kent et al., 2017]] ), particularly during World War II, when ship observations were limited and disproportionately originated from US naval sources ( [[#Thompson--2008|Thompson et al., 2008]] ). [[#Kennedy--2019|Kennedy et al. (2019)]] also identify differences between the new HadSST4 dataset and other SST datasets in the 1980s and 1990s, indicating that some level of structural uncertainty remains during this period, whilst [[#Chan--2019|Chan et al. (2019)]] and [[#Davis--2019|Davis et al. (2019)]] document residual uncertainties in the early and later 20th century records respectively.

Historically, SST has been used as a basis for global temperature assessment on the premise that the less variable SST data provides a better estimate of marine temperature changes than marine air temperature (MAT) ( [[#Kent--2021|Kent and Kennedy, 2021]] ). However, MAT products are used to adjust SST biases in the NOAA SST product because they are assessed to be more homogeneous ( [[#Huang--2017|Huang et al., 2017]] ). Observational datasets exist for night-marine air temperature (NMAT) (e.g., [[#Cornes--2020|Cornes et al., 2020]] ; [[#Junod--2020|Junod and Christy, 2020]] ; [[#Rayner--2020|Rayner et al., 2020]] ) and there are methods to adjust daytime MATs ( [[#Berry--2004|Berry et al., 2004]] ), but there is to date no regularly updated dataset which combines MAT with temperatures over land. MAT datasets are more sparse in recent decades than SST datasets as marine datasets have become increasingly dependent on drifting buoys ( [[#Centurioni--2019|Centurioni et al., 2019]] ) which generally measure SST but not MAT, and there are almost no recent winter MAT data south of 40°S ( [[#Swart--2019|Swart et al., 2019]] ). However, the situation reverses in the 19th century with a greater prevalence of MAT than SST measurements available in the ICOADS data repository ( [[#Freeman--2017|Freeman et al., 2017]] , 2019; [[#Kent--2021|Kent and Kennedy, 2021]] ).

<span id="land-domain"></span>
====== Land domain ======

The GHCNMv4 dataset ( [[#Menne--2018|Menne et al., 2018]] ) includes many more land stations than GHCNMv3, arising from the databank efforts of [[#Rennie--2014|Rennie et al. (2014)]] , and calculates a 100-member parametric uncertainty ensemble drawing upon the benchmarking analysis of [[#Williams--2012|Williams et al. (2012)]] , as well as accounting for sampling effects. A new version of the CRUTEM dataset (CRUTEMv5, [[#Osborn--2021|Osborn et al., 2021]] ) has increased data completeness and additional quality control measures. A new global land dataset, the China Land Surface Air Temperature (CLSAT) dataset ( [[#Xu--2018|Xu et al., 2018]] ) has higher network density in some regions (particularly Asia) than previously existing datasets. Global trends derived from CLSAT are generally consistent with those derived from other land datasets through 2014 ( [[#Xu--2018|Xu et al., 2018]] ).

The AR5 identified diurnal temperature range (DTR) as a substantial knowledge gap. The most recent analysis of Thorne et al. (2016a, b) compared a broad range of gridded estimates of change in DTR, including a new estimate derived from the ISTI databank release using the pairwise homogenization algorithm used to create GHCNMv4, and estimates derived from [[#Vose--2005|Vose et al. (2005)]] , HadEX2 ( [[#Donat--2013a|Donat et al., 2013a]] ), HadGHCND ( [[#Donat--2013b|Donat et al., 2013b]] ), GHCNDEX ( [[#Donat--2013b|Donat et al., 2013b]] ), Berkeley Earth ( [[#Rohde--2013|Rohde et al., 2013]] ), and CRU TS ( [[#Harris--2014|Harris et al., 2014]] ). The analysis highlighted substantial ambiguity in pre-1950 estimates arising from sparse data availability. After 1950 estimates agreed that DTR had decreased globally with most of that decrease occurring over the period 1960–1980. A subsequent DTR analysis using CLSAT further confirmed this behaviour (X. [[#Sun--2018|]] [[#Sun--2018|Sun et al., 2018]] ).

No recent literature has emerged to alter the AR5 finding that it is ''unlikely'' that any uncorrected effects from urbanization (Box 10.3), or from changes in land use or land cover ( [[#2.2.7|Section 2.2.7]] ), have raised global Land Surface Air Temperature (LSAT) trends by more than 10%, although larger signals have been identified in some specific regions, especially rapidly urbanizing areas such as eastern China (Y. [[#Li--2013|]] [[#Li--2013|Li et al., 2013]] ; [[#Liao--2017|Liao et al., 2017]] ; Z. [[#Shi--2019|]] [[#Shi--2019|Shi et al., 2019]] ). There is also no clear indication that site-specific data homogeneity issues have had any significant impact on global trends since the early 20th century; there is more uncertainty in the 19th century, mainly arising from a lack of standardization of instrument shelters, which has been largely accounted for in data from central Europe ( [[#Jones--2012|Jones et al., 2012]] ), but less so elsewhere.

<span id="combined-data-products"></span>
====== Combined data products ======

At the time of AR5 a limitation of conventional datasets was the lack of coverage, especially in high latitudes, which although recognized as an issue ( [[#Simmons--2010|Simmons et al., 2010]] ) had not been addressed in most products. Interpolation involves the statistical imputation of values across regions with limited data and can add both systematic and random uncertainties ( [[#Lenssen--2019|Lenssen et al., 2019]] ). [[#Cowtan--2014|Cowtan and Way (2014)]] applied a kriging-based method to extend existing datasets to polar regions, while [[#Kadow--2020|Kadow et al. (2020)]] used an artificial intelligence-based method, and [[#Vaccaro--2021|Vaccaro et al. (2021)]] used gaussian random Markov fields, for the same purpose, although only [[#Kadow--2020|Kadow et al. (2020)]] uses the most recent generation of datasets as its base. The Berkeley Earth merged product ( [[#Rohde--2020|Rohde and Hausfather, 2020]] ), HadCRUT5 ( [[#Morice--2021|Morice et al., 2021]] ) and NOAA GlobalTemp-Interim ( [[#Vose--2021|Vose et al., 2021]] ) all include interpolation over reasonable distances across data sparse regions which results in quasi-global estimates from the late 1950s when continuous Antarctic observations commenced. Interpolated datasets with substantial coverage of high latitudes show generally stronger warming of GMST than those with limited data in polar regions ( [[#Vose--2021|Vose et al., 2021]] ), and their strong warming at high northern latitudes is consistent with independent estimates from reanalyses ( [[#Simmons--2017|Simmons et al., 2017]] ; [[#Lenssen--2019|Lenssen et al., 2019]] ) and satellites ( [[#Cowtan--2014|Cowtan and Way, 2014]] ). Given the spatial scales of surface temperature variations and the verification of the methods, it is ''extremely likely'' that interpolation results in a less-biased estimate of the actual global temperature change than ignoring regions with limited or no data.

In total there are five conventional datasets which meet spatial coverage requirements and draw from the most recent generation of SST analyses, four of which have sufficient data in the 1850–1900 period to allow an assessment of changes from that baseline (Table 2.3). A fifth dataset is added to the assessment for changes over land areas. Datasets share SST and LSAT data products and in several cases differ solely in the post-processing interpolation applied meaning that there are far fewer methodological degrees of freedom than implied by a straight count of the number of available estimates.

<div id="_idContainer033" class="Basic-Text-Frame"></div>

Table 2.3 | '''Principal characteristics of GMST in situ data products considered in AR6 WGI, highlighting interdependencies in underlying land and SST products and whether inclusion criteria are met.'''

{| class="wikitable"
|-
| '''Dataset'''

| '''Period of Record'''

| '''Land Component'''

| '''SST Component'''

| '''Ensemble Uncertainties?'''

| '''Meets all Inclusion Criteria?'''

| '''Principal Reference'''

|-
| '''HadCRUT5'''

| 1850–2020

| CRUTEM5

| HadSST4

| Yes

| Yes

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

|-
| '''NOAA GlobalTemp – Interim'''

| 1850–2020

| GHCNv4

| ERSSTv5

| Yes, on earlier version

| Yes

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

|-
| '''Berkeley Earth'''

| 1850–2020

| Berkeley

| HadSST4

| No

| Yes

| [[#Rohde--2020|Rohde and Hausfather (2020)]]

|-
| '''Kadow et al.'''

| 1850–2020

| CRUTEM5

| HadSST4

| No

| Yes

| [[#Kadow--2020|Kadow et al. (2020)]]

|-
| '''China – MST'''

| 1856–2020

| CLSAT

| ERSSTv5

| No

| Land only

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

|-
| '''GISTEMP'''

| 1880–2020

| GHCNv4

| ERSSTv5

| Yes

| Post-1880 only

| [[#Lenssen--2019|Lenssen et al. (2019)]]

|-
| '''Cowtan and Way'''

| 1850–2020

| CRUTEM4

| HadSST3

| Yes

| No

| [[#Cowtan--2014|Cowtan and Way (2014)]]

|-
| '''Vaccaro et al.'''

| 1850–2020

| CRUTEM4

| HadSST3

| No

| No

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

|}

Estimates of GMST have also benefitted from improved estimation of parametric uncertainties. New versions of three long-standing products from NASA GISTEMP v4 ( [[#Lenssen--2019|Lenssen et al., 2019]] ), NOAA GlobalTempv5 ( [[#Huang--2019b|]] [[#Huang--2019|B. Huang et al., 2019]] b ) and HadCRUT5 ( [[#Morice--2021|Morice et al., 2021]] ) are all now available as ensemble estimates. These ensembles each account for a variety of systematic and random uncertainty effects in slightly different ways, giving broadly similar results, which are incorporated into the present assessment, with the total uncertainty generally declining up until the mid-20th century as data coverage improves.

Another significant development has been the incorporation of reanalysis products ( [[IPCC:Wg1:Chapter:Chapter-1#1.5.2|Section 1.5.2]] ) into operational monitoring of GSAT. It was reported in AR5 that various reanalyses were broadly consistent with conventional surface datasets in the representation of trends since the mid-20th century. Since that time, [[#Simmons--2017|Simmons et al. (2017)]] found that the ERA-Interim ( [[#Dee--2011|Dee et al., 2011]] ) and JRA-55 ( [[#Kobayashi--2015|Kobayashi et al., 2015]] ) reanalyses continued to be consistent, over the last 20 years, with those surface datasets which fully represented the polar regions. GSAT trends from ERA5 reanalysis ( [[#Hersbach--2020|Hersbach et al., 2020]] ) are also broadly consistent with GMST trends from conventional surface datasets. However, the MERRA-2 reanalysis ( [[#Gelaro--2017|Gelaro et al., 2017]] ) GSAT spuriously cooled sharply relative to ERA-Interim and JRA-55 in about 2007 ( [[#Funk--2019|Funk et al., 2019]] ). Since the early 2000s, analyses of surface temperature, from which near-surface temperature may be derived, have also been available from various satellites ( [[#Famiglietti--2018|Famiglietti et al., 2018]] ; [[#Prakash--2018|Prakash et al., 2018]] ; [[#Susskind--2019|Susskind et al., 2019]] ), which have the potential to improve assessments of temperature changes over data-sparse regions.

Most land areas in the extratropical Northern Hemisphere (NH) have warmed faster than the GMST average over both the 1900–2020 and 1980–2020 periods (Figure 2.11b), although at more regional scales, particularly in data sparse regions, considerable uncertainty is introduced by sometimes large differences in trends between different LSAT datasets ( [[#Rao--2018|Rao et al., 2018]] ). Temperatures averaged over land areas globally have warmed by 1.59 [1.34 to 1.83] °C from 1850–1900 to 2011–2020, substantially higher than the SST warming of 0.88 [0.68 to 1.01] °C. The four conventional surface temperature products which meet all criteria to be included in the final assessment (Table 2.4) agree that each of the last four decades has consecutively been the warmest globally since the beginning of their respective records (Figure 2.11c and Table 2.4). Each of the six years 2015 to 2020 has ''very likely'' been at least 0.9°C warmer than the 1850–1900 average.

<div id="_idContainer034" class="Basic-Text-Frame"></div>

'''Table 2.4''' '''|''' '''Observed increase (°C) in GMST and underlying LSAT and SST estimates in various datasets.''' Numbers in square brackets indicate 5–95% confidence ranges. Trend values are calculated with ordinary least squares following [[#Santer--2008|Santer et al. (2008)]] and expressed as a total change over the stated period. Datasets considered in this table are those with data for at least 90% of global grid points in each year from 1960 onwards. GMST and SST are shown only for data sets which use air temperature (as opposed to climatological SST values) over sea ice. Changes from an 1850–1900 baseline are calculated only for those datasets which have data in at least 80% of years over 1850–1900. GMST values for each year are calculated as the mean of hemispheric means for the NH and SH, while LSAT and SST values are calculated from hemispheric means weighted according to the proportion of land (ocean) in the two hemispheres. This may vary from the methods used by individual data set providers in their own reporting. Products which meet all criteria to be included in the final assessment and contribute to the average are shown in italics. Further details on data sources and processing are available in the chapter data table (Table 2.SM.1).

{| class="wikitable"
|-
| Diagnostic/

Dataset

|
| '''1850–1900 to 1995–2014''' (°C)

| '''1850–1900 to 2001–2020''' (°C)

| '''1850–1900 to 2011–2020''' (°C)

| '''Trend'''

'''1880–2020''' (°C)

| '''Trend'''

'''1960–2020''' (°C)

| '''Trend'''

'''1980–2020''' (°C)

|-
| rowspan="3"| '''HadCRUT5'''

| GMST

| ''0.87''

''[0.81 to 0.94]''

| ''1.01''

''[0.94 to 1.09]''

| ''1.12''

''[1.06 to 1.18]''

| ''1.10''

''[0.89 to 1.32]''

| ''1.04''

''[0.93 to 1.14]''

| ''0.76''

''[0.65 to 0.87]''

|-
| LSAT

| ''1.23''

''[1.06 to 1.38]''

| ''1.44''

''[1.26 to 1.59]''

| ''1.55''

''[1.39 to 1.70]''

| ''1.43''

''[1.16 to 1.70]''

| ''1.50''

''[1.33 to 1.67]''

| ''1.20''

''[1.04 to 1.36]''

|-
| SST

| ''0.73''

''[0.69 to 0.78]''

| ''0.85''

''[0.81 to 0.90]''

| ''0.94''

''[0.90 to 0.99]''

| ''1.03''

''[0.80 to 1.25]''

| ''0.90''

''[0.80 to 0.99]''

| ''0.62''

''[0.51 to 0.72]''

|-
| rowspan="3"| '''NOAA GlobalTemp – Interim'''

| GMST

| ''0.76''

| ''0.91''

| ''1.02''

| ''1.06''

''[0.80 to 1.32]''

| ''1.01''

''[0.90 to 1.11]''

| ''0.75''

''[0.63 to 0.87]''

|-
| LSAT

| ''1.34''

| ''1.55''

| ''1.69''

| ''1.58''

''[1.32 to 1.84]''

| ''1.54''

''[1.40 to 1.68]''

| ''1.19''

''[1.04 to 1.35]''

|-
| SST

| ''0.53''

| ''0.65''

| ''0.75''

| ''0.85''

''[0.59 to 1.12]''

| ''0.79''

''[0.69 to 0.89]''

| ''0.57''

''[0.44 to 0.70]''

|-
| rowspan="3"| '''GISTEMP v4'''

| GMST

|
| ''1.07''

''[0.80 to 1.34]''

| ''1.05''

''[0.94 to 1.16]''

| ''0.79''

''[0.67 to 0.90]''

|-
| LSAT

|
| ''1.48''

''[1.19 to 1.78]''

| ''1.56''

''[1.40 to 1.72]''

| ''1.23''

''[1.07 to 1.39]''

|-
| SST

|
| ''0.91''

''[0.65 to 1.17]''

| ''0.84''

''[0.74 to 0.95]''

| ''0.61''

''[0.49 to 0.72]''

|-
| rowspan="3"| '''Berkeley Earth'''

| GMST

| ''0.89''

| ''1.03''

| ''1.14''

| ''1.17''

''[0.94 to 1.40]''

| ''1.09''

''[1.00 to 1.19]''

| ''0.79''

''[0.68 to 0.90]''

|-
| LSAT

| ''1.28''

| ''1.49''

| ''1.60''

| ''1.50''

''[1.25 to 1.76]''

| ''1.51''

''[1.36 to 1.66]''

| ''1.16''

''[1.00 to 1.32]''

|-
| SST

| ''0.73''

| ''0.85''

| ''0.96''

| ''1.04''

''[0.81 to 1.26]''

| ''0.93''

''[0.84 to 1.01]''

| ''0.64''

''[0.54 to 0.74]''

|-
| '''China-MST'''

| LSAT

| ''1.18''

| ''1.38''

| ''1.49''

| ''1.48''

''[1.21 to 1.75]''

| ''1.48''

''[1.31 to 1.65]''

| ''1.16''

''[1.00 to 1.32]''

|-
| rowspan="3"| '''Kadow et al.'''

| GMST

| ''0.86''

| ''1.00''

| ''1.09''

| ''1.15''

''[0.95 to 1.35]''

| ''1.01''

''[0.92 to 1.10]''

| ''0.73''

''[0.63 to 0.82]''

|-
| LSAT

| ''1.29''

| ''1.49''

| ''1.61''

| ''1.60''

''[1.37 to 1.82]''

| ''1.46''

''[1.30 to 1.61]''

| ''1.14''

''[0.99 to 1.30]''

|-
| SST

| ''0.69''

| ''0.80''

| ''0.88''

| ''0.97''

''[0.78 to 1.16]''

| ''0.83''

''[0.76 to 0.90]''

| ''0.56''

''[0.48 to 0.65]''

|-
| rowspan="3"| '''Cowtan-Way'''

| GMST

| 0.82

[0.75 to 0.89]

| 0.96

[0.89 to 1.03]

| 1.04

[0.97 to 1.11]

| 1.03

[0.84 to 1.22]

| 0.94

[0.82 to 1.07]

| 0.77

[0.67 to 0.87]

|-
| LSAT

| 1.23

| 1.43

| 1.54

| 1.42

[1.15 to 1.68]

| 1.48

[1.31 to 1.65]

| 1.20

[1.04 to 1.36]

|-
| SST

| 0.66

| 0.76

| 0.84

| 0.88

[0.71 to 1.05]

| 0.73

[0.61 to 0.84]

| 0.61

[0.52 to 0.69]

|-
| rowspan="3"| '''Vaccaro et al.'''

| GMST

| 0.76

| 0.89

| 0.97

| 0.99

[0.81 to 1.17]

| 0.89

[0.77 to 1.00]

| 0.72

[0.63 to 0.81]

|-
| LSAT

| 1.15

| 1.35

| 1.47

| 1.40

[1.13 to 1.67]

| 1.47

[1.29 to 1.64]

| 1.21

[1.06 to 1.36]

|-
| SST

| 0.60

| 0.70

| 0.77

| 0.82

[0.67 to 0.97]

| 0.66

[0.55 to 0.76]

| 0.53

[0.44 to 0.61]

|-
| rowspan="2"| '''ERA5'''

| GSAT

|
| 0.78

[0.64 to 0.92]

|-
| LSAT

|
| 1.21

[1.02 to 1.40]

|-
| Average – GMST

|
| 0.85

| 0.99

| 1.09

| 1.11

| 1.04

| 0.76

|-
| Average – LSAT

|
| 1.27

| 1.47

| 1.59

| 1.50

| 1.51

| 1.18

|-
| Average – SST

|
| 0.67

| 0.79

| 0.88

| 0.96

| 0.86

| 0.60

|}

To conclude, from 1850–1900 to 1995–2014, GMST increased by 0.85 [0.69 to 0.95] °C, to the first two decades of the 21st century (2001–2020) by 0.99 [0.84 to 1.10] °C, and to the most recent decade (2011–2020) by 1.09 [0.95 to 1.20] °C. Each of the last four decades has in turn been warmer than any decade that preceded it since 1850. Temperatures have increased faster over land than over the oceans since 1850–1900, with warming to 2011–2020 of 1.59 [1.34 to 1.83] °C versus 0.88 [0.68 to 1.01] °C, respectively.

<div id="2.3.1.2" class="h3-container"></div>

<span id="temperatures-during-the-instrumental-period-free-atmosphere"></span>
==== 2.3.1.2 Temperatures During the Instrumental Period – Free Atmosphere ====

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

The AR5 reported that it was ''virtually certain'' that tropospheric temperatures have risen, and stratospheric temperatures fallen, since the mid-20th century, but that assessments of the rate of change and its vertical structure had only ''medium confidence'' in the NH extratropics and ''low confidence'' elsewhere. In particular there was ''low confidence'' in the vertical structure of temperature trends in the upper tropical troposphere.

<div id="2.3.1.2.1" class="h4-container"></div>

<span id="dataset-developments"></span>
===== 2.3.1.2.1 Dataset developments =====

<div id="h4-10-siblings" class="h4-siblings"></div>

There have been updated radiosonde estimates from the University of Vienna (RAOBCORE and RICH; [[#Haimberger--2012|Haimberger et al., 2012]] ) and a new dataset from the State University of New York (UAHRD, [[#Zhou--2020|Zhou et al., 2020]] ). There are new versions of AMSU products from the University of Alabama in Huntsville (UAHv6.0; [[#Spencer--2017|Spencer et al., 2017]] ) and Remote Sensing Systems (RSSv4.0; [[#Mears--2017|Mears and Wentz, 2017]] ). These updates have led to convergence in the lower stratosphere layer ( [[#Maycock--2018|Maycock et al., 2018]] ); in particular, the move to UAHv6.0 has addressed homogeneity issues identified by [[#Seidel--2016|Seidel et al. (2016)]] , although residual differences remain ( [[#Christy--2018|Christy et al., 2018]] ). Reanalyses products had identified limitations near the 300 hPa level where the contribution of aircraft observations has increased rapidly in recent years ( [[#Dee--2011|Dee et al., 2011]] ; [[#Gelaro--2017|Gelaro et al., 2017]] ), leading to identified biases ( [[#Dee--2009|Dee and Uppala, 2009]] ), that have been addressed in ERA5 ( [[#Hersbach--2020|Hersbach et al., 2020]] ). Modern reanalyses are generally well aligned with radiosonde and satellite observations in the middle and lower troposphere and lower stratosphere. A new operational mid- and upper-stratospheric dataset (STAR) has been developed by [[#Zou--2016|Zou and Qian (2016)]] , merging the previous 1979–2006 SSU dataset ( [[#Zou--2014|Zou et al., 2014]] ) with a dataset from 1998 onwards drawn from relevant AMSU channels ( [[#Wang--2014|Wang and Zou, 2014]] ). Further stratospheric satellite-based datasets from various combinations of satellites have been developed by [[#McLandress--2015|McLandress et al. (2015)]] and [[#Randel--2016|Randel et al. (2016)]] .

New assessments of free-atmosphere temperature are available through radio occultation (RO) and Atmospheric Infrared Sounder (AIRS) products which begin in the early 2000s ( [[IPCC:Wg1:Chapter:Chapter-1#1.5.1.1|Section 1.5.1.1]] ). Global Navigation Satellite System (GNSS)-RO datasets have been compared against AMSU data records, finding almost identical trends ( [[#Khaykin--2017|Khaykin et al., 2017]] ). Comparison of RO with collocated radiosondes, Vaisala RS90/92 and GCOS Reference Upper Air Network data (RS92-GDP; [[#Dirksen--2014|Dirksen et al., 2014]] ), show very good correspondence with global annual mean differences of less than 0.2°C in the upper troposphere and lower stratosphere. Radiosonde daytime radiation biases were identified at higher altitudes ( [[#Ladstädter--2015|Ladstädter et al., 2015]] ; [[#Ho--2017|Ho et al., 2017]] ). The stability of RO makes this data a useful comparator for AMSU ( [[#Chen--2014|Chen and Zou, 2014]] ) and radiosondes ( [[#Ho--2017|Ho et al., 2017]] ; [[#Tradowsky--2017|Tradowsky et al., 2017]] ), as well as anchoring post-2006 reanalyses datasets and improving their consistency in the lower and middle stratosphere ( [[#Long--2017|Long et al., 2017]] ; [[#Ho--2020|Ho et al., 2020]] ). The effective vertical resolution of RO measurements in the upper troposphere and lower stratosphere was found to be up to 100 m at the tropical tropopause ( [[#Zeng--2019a|Zeng et al., 2019a]] ), which is favourable for resolving atmospheric variability ( [[#Scherllin-Pirscher--2012|Scherllin-Pirscher et al., 2012]] ; [[#Wilhelmsen--2018|Wilhelmsen et al., 2018]] ; [[#Stocker--2019|Stocker et al., 2019]] ). Temperature trends in RO products are most consistent with each other and with other observations between 8 km and 25 km ( [[#Ho--2012|Ho et al., 2012]] ; [[#Steiner--2013|Steiner et al., 2013]] , 2020a). The uncertainty increases above 25 km for the early RO period, for which data are based on the single-satellite CHAMP mission, but data at higher altitudes become more reliable for later missions based on advanced receivers ( [[#Steiner--2020a|Steiner et al., 2020a]] ), along with the application of corrections for ionospheric effects ( [[#Danzer--2020|Danzer et al., 2020]] ). The uncertainty due to the changing number of observations is reduced by correcting for the sampling uncertainty in RO climatological fields (e.g., [[#Scherllin-Pirscher--2011|Scherllin-Pirscher et al., 2011]] ). For AIRS, thus far, stability of the instrument has been constrained to less than 0.03°C per decade for selected window channels in a comparison to SSTs measured by ocean buoys ( [[#Aumann--2019|Aumann et al., 2019]] ). Trends were inter-compared with trends in RO data and reanalysis data to assess systematic uncertainties ( [[#Leroy--2018|Leroy et al., 2018]] ).

<div id="2.3.1.2.2" class="h4-container"></div>

<span id="assessment-of-trends"></span>
===== 2.3.1.2.2 Assessment of trends =====

<div id="h4-11-siblings" class="h4-siblings"></div>

Warming has continued in the lower troposphere according to all radiosonde, reanalyses and satellite datasets, with a rate over 1980–2019 similar to surface warming rates (Table 2.5; c.f. Table 2.4). Radiosonde-based products generally show greater warming rates for 1980–2019 than satellite-based products and reanalyses. They also extend further back to the 1950s and trends since quasi-global coverage around 1960 also show warming (Table 2.5). Trends in RO and AIRS data, supported by radiosonde datasets, exhibit a warming trend in most of the mid- to upper- troposphere at all non-polar latitudes over 2002–2019. These also exhibit faster warming rates in the tropics in the upper troposphere than those observed at or near the surface (Figure 2.12); with the lowermost stratosphere also warming while above it is cooling. There is some spread between different data types in the tropics near the 15km level, although these differences are reduced to near zero if a subset of radiosonde data, using only high-quality instruments, is used ( [[#Steiner--2020b|Steiner et al., 2020b]] ). AMSU tropical middle troposphere data also show that warming rates are near or above those in the lower troposphere, but they are measuring much broader layers which greatly complicates interpretation ( [[#Steiner--2020b|Steiner et al., 2020b]] ).

Temperatures averaged through the full lower stratosphere (roughly 10–25 km) have decreased over 1980–2019 in all data products, with the bulk of the decrease prior to 2000. The decrease holds even if the influence of the El Chichon (1982) and Pinatubo (1991) volcanic eruptions on the trend, found by [[#Steiner--2020a|Steiner et al. (2020a)]] to have increased the 1979–2018 cooling trend by 0.06°C per decade, is removed. Most datasets show no significant or only marginally significant trends over 2000–2019, and the results of [[#Philipona--2018|Philipona et al. (2018)]] show weak increases over 2000–2015 in the very lowermost stratosphere sampled by radiosondes.

The STAR dataset shows cooling in the middle and upper stratosphere with a trend of –0.56°C ± 0.16°C per decade for the mid-stratosphere and –0.62°C ± 0.29°C per decade for the upper stratosphere over 1980–2019, although both cooling rates have slowed substantially since the mid-1990s. The overall post-1980 trend is reduced in magnitude by about 0.10°C per decade at both levels if the influences of the El Chichon and Pinatubo eruptions, and the solar cycle, are removed ( [[#Zou--2016|Zou and Qian, 2016]] ). The results obtained by [[#McLandress--2015|McLandress et al. (2015)]] for 1980–2012, [[#Randel--2016|Randel et al. (2016)]] for 1979–2015, and [[#Maycock--2018|Maycock et al. (2018)]] for 1979–2016 are broadly consistent with this.

A rise in the tropopause height of 40 to 120 m per decade between 1981 and 2015 was determined from both radiosonde and reanalysis datasets ( [[#Xian--2019|Xian and Homeyer, 2019]] ). Local studies (e.g., [[#Tang--2017|Tang et al., 2017]] ; X. [[#Chen--2019|]] [[#Chen--2019|Chen et al., 2019]] ) found stronger trends in some regions near the subtropical jet linked to tropical expansion ( [[#2.3.1.4.1|Section 2.3.1.4.1]] ). Whilst [[#Seidel--2006|Seidel and Randel (2006)]] found that the tropopause height was more closely coupled with temperatures in the stratosphere than those in the troposphere, it is not yet clear whether the rate of increase in tropopause height has experienced a similar recent slowdown to that of the cooling of the lower stratosphere, as short-period trends are typically inconclusive due to significant natural variability ( [[#Scherllin-Pirscher--2021|Scherllin-Pirscher et al., 2021]] ). RO data ( [[#Gao--2015|Gao et al., 2015]] ) indicate little change in tropopause height over the short period from 2006 to 2014, but a warming below the tropopause is observed over 2002 to 2019 (Figure 2.12).

<div id="_idContainer036" class="Basic-Text-Frame"></div>

[[File:87a7746f1162a91715bb84e29f811c36 IPCC_AR6_WGI_Figure_2_12.png]]

'''Figure 2.1''' '''2 |''' '''Temperature trends in the upper air. (a)''' Zonal cross-section of temperature anomaly trends (2007–2016 baseline) for 2002–2019 in the upper troposphere and lower stratosphere region. The climatological tropopause altitude is marked by a grey line. Significance is not indicated due to the short period over which trends are shown, and because the assessment findings associated to this figure relate to difference between trends at different heights, not the absolute trends. '''(b, c)''' Trends in temperature at various atmospheric heights for 1980–2019 and 2002–2019 for the near-global (70°N–70°S) domain. '''(d, e)''' as for (b, c) but for the tropical (20°N–20°S) region. Further details on data sources and processing are available in the chapter data table (Table 2.SM.1).

In summary, the troposphere has warmed since the mid-20th century. There is ''medium confidence'' that temperatures in the tropical upper troposphere have warmed faster than those at the surface since 2001, but ''low confidence'' in changes prior to 2001. It is ''virtually certain'' that the lower stratosphere has cooled since the mid-20th century. However, most datasets show that lower stratospheric temperatures have stabilized since the mid-1990s with no significant change over the last 20 years. It is ''likely'' that middle and upper stratospheric temperatures have decreased since 1980, but there is ''low confidence'' in the magnitude. It is ''virtually certain'' that the tropopause height has risen over 1980–2019 but there is ''low confidence'' in the magnitude of this rise, or whether the rate of change has reduced commensurate with stabilized lower stratospheric temperatures.

<div id="_idContainer037" class="Basic-Text-Frame"></div>

Table 2.5 '''|''' '''Observed change (°C) in free atmospheric temperatures in various datasets, for the lower tropospheric and lower stratospheric layers.''' Numbers in square brackets indicate 5–95% confidence ranges. Trend values are calculated with ordinary least squares following ( [[#Santer--2008|Santer et al., 2008]] ) and are expressed as a total change over the stated period. Further details on data sources and processing are available in the chapter data table (Table 2.SM.1).

{| class="wikitable"
|-
| '''Diagnostic/Dataset'''

| '''Trend'''

'''1960–2019'''

| '''Trend'''

'''1980–2019'''

| '''Trend'''

'''2000–2019'''

|-
| colspan="4"| '''Lower troposphere'''

|-
| '''RAOBCORE'''

| 1.08

[0.94 to 1.23]

| 0.74

[0.57 to 0.91]

| 0.52

[0.26 to 0.78]

|-
| '''RICH'''

| 1.20

[1.06 to 1.34]

| 0.79

[0.63 to 0.96]

| 0.53

[0.28 to 0.77]

|-
| '''UAHRD'''

| 0.97

[0.80 to 1.13]

| 0.91

[0.76 to 1.05]

| 0.53

[0.35 to 0.72]

|-
| '''UAH'''

|
| 0.51

[0.37 to 0.65]

| 0.29

[0.07 to 0.50]

|-
| '''RSS'''

|
| 0.79

[0.66 to 0.92]

| 0.41

[0.24 to 0.58]

|-
| '''ERA5.1'''

|
| 0.68

[0.52 to 0.84]

| 0.55

[0.34 to 0.75]

|-
| Average

| 1.08

| 0.74

| 0.47

|-
| colspan="4"| '''Lower stratosphere'''

|-
| '''RAOBCORE'''

| –1.37

[–1.80 to –0.93]

| –1.00

[–1.56 to –0.45]

| –0.05

[–0.20 to 0.09]

|-
| '''RICH'''

| –1.45

[–1.99 to –0.92]

| –1.19

[–1.95 to –0.42]

| 0.02

[–0.20 to 0.23]

|-
| '''UAHRD'''

| −1.25

[−1.51 to −0.98]

| −0.79

[−1.16 to −0.43]

| −0.11

[−0.25 to 0.03]

|-
| '''UAH'''

|
| –1.14

[–1.61 to –0.67]

| –0.24

[–0.37 to –0.12]

|-
| '''RSS'''

|
| –0.90

[–1.37 to –0.43]

| –0.14

[–0.26 to –0.03]

|-
| '''STAR'''

|
| –0.97

[–1.45 to –0.49]

| –0.17

[–0.29 to –0.04]

|-
| '''ERA5.1'''

|
| –1.19

[–1.87 to –0.50]

| –0.01

[–0.13 to 0.10]

|-
| Average

| –1.36

| –1.03

| –0.10

|}

<div id="2.3.1.3" class="h3-container"></div>

<span id="global-hydrological-cycle"></span>

==== 2.3.1.3 Global Hydrological Cycle ====

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

This section focuses on large-scale changes in a subset of components of the hydrological cycle (Cross-Chapter Box 2.2). [[IPCC:Wg1:Chapter:Chapter-8|Chapter 8]] undertakes a holistic assessment of changes in the hydrological cycle integrating observations, modelling and theoretical understanding, while ( [[IPCC:Wg1:Chapter:Chapter-11|Chapter 11]] assesses hydrological cycle extremes such as droughts and floods.

<div id="2.3.1.3.1" class="h4-container"></div>

<span id="paleo-perspective-of-the-global-hydrological-cycle"></span>
===== 2.3.1.3.1 Paleo perspective of the global hydrological cycle =====

<div id="h4-12-siblings" class="h4-siblings"></div>

The AR5 assessed large-scale indicators of terrestrial paleo hydroclimate, including as part of its assessment of paleo floods and droughts, but did not assess proxy evidence for paleo hydroclimate indicators over continental and larger scales. The paleoclimate evidence assessed in AR5 was broadly consistent with global hydroclimate scaling with temperature: warmer periods were wetter (e.g., the Pliocene; increased precipitation) with colder periods being drier (e.g., the LGM; decreased precipitation).

Substantial limitations exist in reconstructing the global hydrological cycle prior to the Quaternary, particularly during the Eocene, due to the lack of high-resolution proxy records and their sparsity. Spatial heterogeneity complicates identification of wetting and drying signals during the PETM and the EECO, with paleo data and model simulations suggesting an intensified global hydrological cycle ( [[#Carmichael--2016|Carmichael et al., 2016]] , 2017; [[#Hyland--2017|Hyland et al., 2017]] ; [[#West--2020|West et al., 2020]] ), in particular an increased specific humidity ( [[#Winnick--2015|Winnick et al., 2015]] ; [[#van%20Dijk--2020|van Dijk et al., 2020]] ). Conditions wetter than present were inferred for the MPWP (Cross Chapter Box 2.4), with intensified Asian monsoons ( [[#An--2015|An et al., 2015]] ) but with nevertheless drier conditions over tropical and subtropical SH locations ( [[#Pontes--2020|Pontes et al., 2020]] ). A new global reconstruction of hydroclimate proxies for the LIG points to stronger boreal precipitation compared to 1850–1900 over high latitudes and especially over monsoon areas, with a more heterogeneous signal for the SH ( [[#Scussolini--2019|Scussolini et al., 2019]] ). This heterogeneity is also present in the tropics, characterized by large zonal differences in precipitation change due to the variations in the intensity of Walker circulation ( [[#2.3.1.4.1|Section 2.3.1.4.1]] ). Available records indicate reduced global vegetation cover and abundant atmospheric dust deposition during the LGM (increased aridity), particularly over the tropics and high latitudes ( [[#Lamy--2014|Lamy et al., 2014]] ; [[#Újvári--2017|Újvári et al., 2017]] ). This agrees with models and moisture-sensitive proxies, suggesting an overall decrease in global precipitation during the LGM relative to recent decades, albeit with regional-scale heterogeneity ( [[#Cao--2019|Cao et al., 2019]] ). Despite lower global precipitation amounts, research since AR5 has identified a wetting of mid-latitudes during the LGM ( [[#Putnam--2017|Putnam and Broecker, 2017]] ; [[#Lowry--2018|Lowry and Morrill, 2018]] ; [[#Morrill--2018|Morrill et al., 2018]] ), thereby complicating the characterization of the LGM as a relatively ‘dry’ period. Low evaporation rates and increased top-soil moisture during the LGM may have contributed to elevated levels of large closed-basin lakes located in the 30°–45° latitudinal belts ( [[#Putnam--2017|Putnam and Broecker, 2017]] ; [[#Scheff--2017|Scheff et al., 2017]] ), such as the south-west United States (e.g., [[#Ibarra--2018|Ibarra et al., 2018]] ), southern Australia ( [[#Petherick--2013|Petherick et al., 2013]] ; [[#Fitzsimmons--2015|Fitzsimmons et al., 2015]] ; [[#Sniderman--2019|Sniderman et al., 2019]] ) and Patagonia (e.g., [[#Quade--2017|Quade and Kaplan, 2017]] ).

New analyses suggest that during the Holocene, the NH mid-latitudes became increasingly wet, in phase with the strength of the latitudinal temperature and insolation gradients ( [[#Shuman--2016|Shuman and Marsicek, 2016]] ; [[#Routson--2019|Routson et al., 2019]] ). Nevertheless, there was also considerable spatial heterogeneity and variability on centennial to millennial timescales ( [[#Newby--2014|Newby et al., 2014]] ; [[#Shuman--2016|Shuman and Marsicek, 2016]] ; H. [[#Zhang--2018|]] [[#Zhang--2018|]] [[#Zhang--2018|]] [[#Zhang--2018|Zhang et al., 2018]] ; [[#Liefert--2020|Liefert and Shuman, 2020]] ). The NH tropics and many regions of the SH deep tropics experienced wetting up until the early to mid-Holocene but drying thereafter ( [[#Shanahan--2015|Shanahan et al., 2015]] ; [[#Nash--2016|Nash et al., 2016]] ; [[#Muñoz--2017|Muñoz et al., 2017]] ; [[#Quade--2018|Quade et al., 2018]] ). ''Evidence'' for the SH is ''limited'' , with a wetting trend during the Holocene in low latitudes of South America ( [[#Kanner--2013|Kanner et al., 2013]] ; [[#Mollier-Vogel--2013|Mollier-Vogel et al., 2013]] ) and parts of the African tropics ( [[#Schefuß--2011|Schefuß et al., 2011]] ; [[#Chevalier--2015|Chevalier and Chase, 2015]] ) but a drying tendency over southern Australia and New Zealand ( [[#van%20den%20Bos--2018|van den Bos et al., 2018]] ; [[#Barr--2019|Barr et al., 2019]] ) and South America ( [[#Quade--2017|Quade and Kaplan, 2017]] ; [[#Moreno--2018|Moreno et al., 2018]] ).

For the CE, new proxy records have led to the creation of continental drought atlases ( [[#Cook--2015|Cook et al., 2015]] ; [[#Palmer--2015|Palmer et al., 2015]] ; [[#Stahle--2016|Stahle et al., 2016]] ; [[#Morales--2020|Morales et al., 2020]] ) and millennial reanalyses ( [[#Steiger--2018|Steiger et al., 2018]] ; [[#Tardif--2019|Tardif et al., 2019]] ). These reconstructions highlighted the occurrence of multi-decadal regional mega-droughts in the NH before 1600 CE, particularly during 800–1200 CE, with a predominance of wet periods after 1700 CE ( [[#Cook--2015|Cook et al., 2015]] ; [[#Rodysill--2018|Rodysill et al., 2018]] ; [[#Shuman--2018|Shuman et al., 2018]] ). In the SH, much of South America and the African tropics experienced a reduction of precipitation during 900–1200 CE and a wetting peak during 1500–1800 CE ( [[#Tierney--2015|Tierney et al., 2015]] ; [[#Nash--2016|Nash et al., 2016]] ; [[#Fletcher--2018|Fletcher et al., 2018]] ; [[#Lüning--2018|Lüning et al., 2018]] ; [[#Campos--2019|Campos et al., 2019]] ), with an opposite pattern in southern subtropical Africa ( [[#Woodborne--2015|Woodborne et al., 2015]] ; [[#Lüning--2018|Lüning et al., 2018]] ). Large multi-decadal variability was documented over Australia and New Zealand during the 800–1300 CE period, followed by a well-defined wet period during 1500–1800 CE ( [[#Barr--2014|Barr et al., 2014]] ; [[#Evans--2019|Evans et al., 2019]] ).

To summarize, since AR5 there has been considerable progress in detecting the variations of the global hydrological cycle prior to the instrumental period. There are indications from multiple sources of a wetting trend during the Holocene, particularly for the NH and parts of the SH tropics ( ''medium confidence'' ). Hydroclimate during the CE is dominated by regional variability, generally precluding definitive statements on changes at continental and larger scales, with a general reduction of mega-drought occurrences over the last about 500 years ( ''medium confidence'' ). Availability of proxy data for assessing Holocene hydroclimate variability is biased towards the NH, with ''medium evidence'' but ''low agreement'' for the assessment of SH changes.

<div id="2.3.1.3.2" class="h4-container"></div>

<span id="surface-humidity"></span>
===== 2.3.1.3.2 Surface humidity =====

<div id="h4-13-siblings" class="h4-siblings"></div>

The AR5 reported ''very likely'' widespread increases in near-surface air specific humidity since the 1970s, abating from around 2000 to 2012 ( ''medium confidence'' ). This abatement resulted in a recent decline in relative humidity over the land.

Near surface humidity has been monitored using in-situ data (e.g., NOCSv2.0; [[#Berry--2011|Berry and Kent, 2011]] ), satellite-derived estimations (e.g., HOAPS3, [[#Liman--2018|Liman et al., 2018]] ; J-OFURO3, [[#Tomita--2019|Tomita et al., 2019]] ), global gridded products such as HadISDH ( [[#Willett--2014|Willett et al., 2014]] , 2020), and reanalyses (e.g., ERA5, JRA-55 and 20CRv3). In-situ based humidity products suffer from uncertainties over poorly sampled regions particularly in the SH ( [[#Berry--2011|Berry and Kent, 2011]] ; [[#Kent--2014|Kent et al., 2014]] ; [[#Willett--2014|Willett et al., 2014]] ). There is general consensus in the inter-annual variability and sign of trends implying ''high confidence'' in increasing specific humidity since the 1970s and decreasing relative humidity since 2000, particularly over land ( [[#Simmons--2010|Simmons et al., 2010]] ; [[#Willett--2014|Willett et al., 2014]] , 2020). Since 2012, specific humidity over land and ocean has remained well above the 1973–2019 average and reached record or near-record values (Figure 2.13b), with the strong 2015–2016 El Niño event boosting surface moisture levels ( [[#Byrne--2018|Byrne and O’Gorman, 2018]] ). The abatement from around 2000 to 2012 reported in AR5 has not persisted. This is consistent with increases in total column water vapour ( [[#2.3.1.3.3|Section 2.3.1.3.3]] ) and a resumption of rapid warming in surface temperatures ( [[#2.3.1.1.3|Section 2.3.1.1.3]] ). The global averaged relative humidity however has remained depressed since 2000 (Figure 2.13d; [[#Simmons--2010|Simmons et al., 2010]] ; [[#Willett--2014|Willett et al., 2014]] , 2020; [[#Dunn--2017|Dunn et al., 2017]] ; [[#Vicente-Serrano--2018|Vicente-Serrano et al., 2018]] ).

Since 1973, increases in specific humidity have been widespread and significant across the majority of the land and ocean regions where observations are available (Figure 2.13a). In contrast, trends in relative humidity show distinct spatial patterns with generally increasing trends over the higher latitudes and the tropics and generally decreasing trends over the sub-tropics and mid-latitudes, particularly over land areas (Figure 2.13c). Near-surface specific humidity over the oceans has increased since the 1970s according to several in-situ, satellite and reanalysis data records ( [[#Kent--2014|Kent et al., 2014]] ; [[#Robertson--2020|Robertson et al., 2020]] ; [[#Willett--2020|Willett et al., 2020]] ). According to the HadISDH product, increases in specific humidity and decreases in relative humidity are significant particularly over the NH mid-latitudes (Figure 2.13a,c). Poor data coverage over the SH south of 20°S does not allow for the robust assessment of trends. Sources of uncertainty include the initial measurement accuracy, homogenization over land, observational height at ships and instrument bias adjustment over ocean, and sparse spatio-temporal sampling ( [[#Prytherch--2015|Prytherch et al., 2015]] ; [[#Roberts--2019|Roberts et al., 2019]] ; [[#Willett--2020|Willett et al., 2020]] ).

<div id="_idContainer039" class="Basic-Text-Frame"></div>

[[File:fac3527940cee5a3edb1f1f7ea1185f3 IPCC_AR6_WGI_Figure_2_13.png]]

'''Figure 2.13''' '''|''' '''Changes in surface humidity. (a)''' Trends in surface specific humidity over 1973–2019. Trends are calculated using OLS regression with significance assessed following AR(1) adjustment after [[#Santer--2008|Santer et al. (2008)]] ; ‘×’ marks denote non-significant trends). '''(b)''' Global average surface specific humidity annual anomalies (1981–2010 base period). '''(c)''' as (a) but for the relative humidity. '''(d)''' as (b) but for the global average surface relative humidity annual anomalies. Further details on data sources and processing are available in the chapter data table (Table 2.SM.1).

In summary, observations since the 1970s show a ''very likely'' increase in near surface specific humidity over both land and oceans. A ''very likely'' decrease in relative humidity has occurred over much of the global land area since 2000, particularly over mid-latitude regions of the NH, with increases at northern high latitudes.

<div id="2.3.1.3.3" class="h4-container"></div>

<span id="total-column-water-vapour-tcwv"></span>

===== 2.3.1.3.3 Total column water vapour (TCWV) =====

<div id="h4-14-siblings" class="h4-siblings"></div>

The AR5 concluded that total column water vapour (TCWV) ''very likely'' increased since the 1970s, at a rate that was overall consistent with the Clausius-Clapeyron relationship (about 7% per °C) given the observed increase in atmospheric temperature.

Records prior to the instigation of quasi-global coverage by radiosondes require the use of statistical relationships to infer TCWV from historical SST observations or the evaluation of centennial-scale reanalysis products ( [[#Smith--2015|Smith and Arkin, 2015]] ). These approaches reveal two periods of positive trends, one from 1910 to 1940 and the other from 1975 onwards ( [[#Zhang--2013|Zhang et al., 2013]] ; [[#Mieruch--2014|Mieruch et al., 2014]] ; [[#Shi--2018|Shi et al., 2018]] ), concurrent with periods of positive SST trends (Figure 2.11). Potential sources of errors in the SST-based estimation of TCWV include both uncertainties in historical SST and uncertainties in the parameters that define the relationship between the variables ( [[#Smith--2015|Smith and Arkin, 2015]] ). Trends based on 20CRv2c, ERA-20C and ERA-20CM indicate an increase in TCWV over much of the global ocean since the beginning of the 20th century, particularly over the tropics ( [[#Bordi--2015|Bordi et al., 2015]] ; [[#Smith--2015|Smith and Arkin, 2015]] ; [[#Poli--2016|Poli et al., 2016]] ). TCWV trends estimated since the middle of the 20th century from radiosonde observations show significant increases over North America and large portions of Eurasia, while decreases are restricted to Australia, eastern Asia and the Mediterranean region (Y. [[#Zhang--2018|]] [[#Zhang--2018|]] [[#Zhang--2018|]] [[#Zhang--2018|Zhang et al., 2018]] ). Overall, there is a significant increase in TCWV over global land areas since 1979 ( [[#Chen--2016|Chen and Liu, 2016]] ).

Since the late 1970s a range of satellite missions permit a quasi-global assessment of TCWV. Several satellite products provide water vapour retrievals based upon distinct spectral domains, in addition to products from radiosondes, reanalyses and GNSS radio occultation. The GEWEX Water Vapour Assessment (G-VAP) provided an intercomparison of several TCWV data records, with global coverage but limited timespan ( [[#Schröder--2018|Schröder et al., 2018]] ). The various global products generally exhibit a positive trend since 1979 (Figure 2.14; [[#Allan--2014|Allan et al., 2014]] ; [[#Mieruch--2014|Mieruch et al., 2014]] ; [[#Schröder--2016|Schröder et al., 2016]] ; J. [[#Wang--2016|]] [[#Wang--2016|]] [[#Wang--2016|]] [[#Wang--2016|]] [[#Wang--2016|]] [[#Wang--2016|Wang et al., 2016]] ), most evident over the tropics ( [[#Gu--2013|Gu and Adler, 2013]] ; [[#Chen--2016|Chen and Liu, 2016]] ; [[#Mears--2018|Mears et al., 2018]] ; [[#Wang--2020|Wang and Liu, 2020]] ; [[#Salamalikis--2021|Salamalikis et al., 2021]] ). The existence of apparent breakpoints in several products, which are generally coincident with changes in the observing system, lead to trend estimates that are not in line with theoretical expectations imposed by the Clausius-Clapeyron relationship ( [[#Schröder--2019|Schröder et al., 2019]] ), although other factors such as regional moisture divergence/convergence could account for the observed TCWV-temperature scaling. Substantial potential inhomogeneities affect trend estimates based on satellite, reanalysis and merged products in particular over Central Africa, the Sahara and central South America ( [[#Schröder--2016|Schröder et al., 2016]] , 2019; J. [[#Wang--2016|]] [[#Wang--2016|]] [[#Wang--2016|]] [[#Wang--2016|]] [[#Wang--2016|]] [[#Wang--2016|Wang et al., 2016]] ). Moreover, data gaps in observations from ground-based GNSS receivers and radiosondes lead to ''low confidence'' in TCWV estimation in these regions.

<div id="_idContainer041" class="Basic-Text-Frame"></div>

[[File:100dd2195115423d233bb4cf4b6a71ee IPCC_AR6_WGI_Figure_2_14.png]]

'''Figure 2.1''' '''4 |''' '''Time series of global mean total column water vapour annual anomalies (mm) relative to a 1988–2008 base period.''' Further details on data sources and processing are available in the chapter data table (Table 2.SM.1).

In summary, positive trends in global total column water vapour are ''very likely'' since 1979 when globally representative direct observations began, although uncertainties associated with changes in the observing system imply ''medium confidence'' in estimation of the trend magnitudes. ''Low confidence'' in longer-term trends arises from uncertainties in the SST-TCWV relationship and current centennial scale reanalyses, particularly during the first half of the 20th century.

<div id="2.3.1.3.4" class="h4-container"></div>

<span id="global-precipitation"></span>

===== 2.3.1.3.4 Global precipitation =====

<div id="h4-15-siblings" class="h4-siblings"></div>

The AR5 concluded that there was ''low confidence'' in precipitation change averaged over global land areas prior to 1950, and ''medium confidence'' thereafter with no significant global trends. There was a ''likely'' overall increase in precipitation in the well-sampled NH mid-latitudes, with ''high confidence'' after 1951.

In situ precipitation records over land extend back for centuries in a few locations, and to the early to mid-20th century quasi-globally. Datasets differ in their input data, completeness of records, period covered, and the gridding procedures applied, which, given spatial clustering and the small spatial scales of precipitation, results in differences in global and regional estimates of precipitation changes (Q. [[#Sun--2018|]] [[#Sun--2018|Sun et al., 2018]] ; [[#Nogueira--2020|Nogueira, 2020]] ). The spatial variability of observed long-term trends (1901–2019) based on GPCC V2020 and CRU TS 4.04 (Figure 2.15a,b) indicates significant increases in precipitation mainly over eastern North America, northern Eurasia, southern South America and north-western Australia. Decreases are strongest across tropical western and equatorial Africa and southern Asia. The temporal evolution of global annual land precipitation anomalies exhibits little consistency between GPCC V2020, CRU TS 4.04 and GHCNv4 datasets, especially prior to 1950, that is associated with limitations in data coverage (Figure 2.15c; [[#Wu--2013|Wu et al., 2013]] ; [[#Shen--2014|Shen et al., 2014]] ; [[#Gu--2015|Gu and Adler, 2015]] ). These disagreements between datasets prior to the 1950s result in differences in trend estimates over global land (Table 2.6). A qualitative consistency in decadal and interdecadal variations between the products is only observed since the 1950s, with primarily positive land precipitation anomalies during the 1950s, 1970s and during 2000 to 2019 (Figure 2.15c).

<div id="_idContainer042" class="Basic-Text-Frame"></div>

'''Table 2.''' '''6 |''' '''Globally averaged trend estimates over land and 90% confidence''' '''intervals for annual precipitation for each time series in Figure 2.15c over three periods all ending in 2019.''' Trends are calculated using OLS regression with significance assessed after [[#Santer--2008|Santer et al. (2008)]] . Further details on data sources and processing are available in the chapter data table (Table 2.SM.1).

{| class="wikitable"
|-
| rowspan="2"| '''Dataset'''

| colspan="3"| '''Trends in annual precipitation (mm yr''' <sup>–1</sup> '''per decade)'''

|-
| '''1901–2019'''

| '''1960–2019'''

| '''1980–2019'''

|-
| GPCCv2020

| 1.01 <sup>a</sup> ± 0.99

| 1.67 ± 3.23

| 5.60 ± 6.38

|-
| CRU TS 4.04

| 0.57 ± 2.08

| 0.17 ± 3.12

| 5.75 <sup>a</sup> ± 5.09

|-
| GHCNv4

| 3.19 <sup>a</sup> ± 1.48

| 5.03 <sup>a</sup> ± 4.87

| 11.06 <sup>a</sup> ± 9.17

|-
| GPCPv2.3

|
| 5.41 <sup>a</sup> ± 5.20

|}

<sup>a</sup> Trend values significant at the 10% level.

<div id="_idContainer044" class="Basic-Text-Frame"></div>

[[File:c2addba488a0d6d769c525d313a70c08 IPCC_AR6_WGI_Figure_2_15.png]]

'''Figure 2.15''' '''|''' '''Changes in observed precipitation. (a, b)''' Spatial variability of observed precipitation trends over land for 1901–2019 for two global in-situ products. Trends are calculated using OLS regression with significance assessed following AR(1) adjustment after [[#Santer--2008|Santer et al. (2008)]] (‘×’ marks denote non-significant trends). '''(c)''' Annual time series and decadal means from 1891 to date relative to a 1981–2010 climatology (note that different products commence at distinct times). '''(d, e)''' as (a, b), but for the periods starting in 1980. '''(f)''' is for the same period for the globally complete merged GPCP v2.3 product. Further details on data sources and processing are available in the chapter data table (Table 2.SM.1).

Several satellite-based precipitation datasets improve the representation of the spatio-temporal changes since the late 20th century. Some of these are based exclusively on satellite data (e.g., CMORPH, [[#Joyce--2004|Joyce et al., 2004]] ; GSMaP, [[#Okamoto--2005|Okamoto et al., 2005]] ), with others being combinations of in situ observations, reanalyses and satellite retrievals (e.g., CMAP, [[#Xie--1997|Xie and Arkin, 1997]] ; TRMM 3B43 V7, [[#Huffman--2007|Huffman et al., 2007]] ; PERSIANN-CDR, [[#Ashouri--2015|Ashouri et al., 2015]] ; CHIRPS, [[#Funk--2015|Funk et al., 2015]] ; GPCP V2.3, [[#Adler--2018|Adler et al., 2018]] ). These can be affected by systematic and random uncertainties due to inhomogeneities in the satellite-derived precipitation and station data and the uncertainties of blending algorithms ( [[#Hegerl--2015|Hegerl et al., 2015]] ; Q. [[#Sun--2018|]] [[#Sun--2018|Sun et al., 2018]] ). The spatial coverage of these products is near-global, with available estimations formally covering 60°S–60°N with decreasing quality from low to high latitudes, depending on the sensors and algorithms used ( [[#Hu--2019|Hu et al., 2019]] ). A detailed description of the most relevant satellite products is provided in section 10.2.1.1.

Recent trends (1980–2019) for GPCC V2020, CRU TS 4.04 and GPCP V2.3 show significant increases in land precipitation over tropical Africa, the eastern portions of Europe and North America, central Asia and the Maritime Continent (Figure 2.14d–f). Significant decreases are observed over central South America, western North America, northern Africa and the Middle East. A detailed assessment of the recent regional precipitation trends using the same datasets can be found in the Atlas. Global trends for 1980–2019 show a general increase in annual precipitation over land, which is particularly marked for CRU TS 4.04 and GHCNv4 (Table 2.6). These changes have been accompanied by a strengthening of precipitation seasonality over tropical land areas, although with broad spread between different satellite-based (GPCP, MSWEP_V1.2, PERSIANN-CDR) and in situ gridded datasets (GPCC, CRU TS; [[#Chou--2013|Chou et al., 2013]] ; [[#Li--2016|Li et al., 2016]] ; [[#Tan--2020|Tan et al., 2020]] ). Increasing trends since 1980, in contrast to longer-term declining trends since 1901, are particularly evident over much of Africa, while more widespread negative trends were observed over much of southern South America in the more recent period ( [[IPCC:Wg1:Chapter:Atlas|Atlas]] 7.2; [[#Knutson--2018|Knutson and Zeng, 2018]] ). A faster recent increase in precipitation over global land is inferred comparing the precipitation trends over 1960–2019 with 1980–2019 (Table 2.6). Over the global ocean, the comparison between precipitation datasets is compromised by the different measurement periods, as well as the spatial coverage of the available products ( [[#Adler--2017|Adler et al., 2017]] ; [[#Nguyen--2018|Nguyen et al., 2018]] ; [[#Jaber--2020|Jaber and Abu-Allaban, 2020]] ; [[#Nogueira--2020|Nogueira, 2020]] ), limiting the ability to assess the sign and magnitude of precipitation trends. The GPCPv2.3 database ( [[#Adler--2017|Adler et al., 2017]] , 2018) exhibits an increase of 2.94 mm yr <sup>–1</sup> per decade over 1980–2019, principally due to the trends over the Indian ocean and in the tropical western Pacific (Figure 2.15f). The regional patterns of recent trends are consistent with the documented increase in precipitation over tropical wet regions and the decrease over dry areas, estimated through GPCP v2.2 data ( [[#Liu--2013|Liu and Allan, 2013]] ; [[#Trammell--2015|Trammell et al., 2015]] ; [[#Kao--2017|Kao et al., 2017]] ; [[#Polson--2017|Polson and Hegerl, 2017]] ).

In summary, globally averaged land precipitation has ''likely'' increased since the middle of the <sup></sup> 20th century ( ''medium confidence'' ), with ''low confidence'' in trends prior to 1950. A faster increase in global land precipitation was observed since the 1980s ( ''medium confidence'' ), with large interannual variability and regional heterogeneity. Over the global ocean there is ''low confidence'' in the estimates of precipitation trends, linked to uncertainties in satellite retrievals, merging procedures and limited in situ observations.

<div id="2.3.1.3.5" class="h4-container"></div>

<span id="precipitation-minus-evaporation"></span>

===== 2.3.1.3.5 Precipitation minus evaporation =====

<div id="h4-16-siblings" class="h4-siblings"></div>

The AR5 concluded that the pattern of precipitation minus evaporation (P–E) over the ocean had been enhanced since the 1950s ( ''medium confidence'' ). Saline surface waters had become saltier, while the relatively fresh surface waters had become fresher. The inferred changes in P–E were consistent with the observed increased TCWV, although uncertainties in the available products prevented identifying robust trends.

Estimating global-scale trends in P–E using direct observations alone is challenging due to limited evaporation measurements and inhomogeneities in satellite-derived precipitation and evaporation datasets ( [[#Hegerl--2015|Hegerl et al., 2015]] ; [[#López--2017|López et al., 2017]] ). Hence, the assessment of global P–E trends is generally performed using reanalyses, although changes in the observing system imply considerable uncertainty ( [[#Skliris--2014|Skliris et al., 2014]] ). Since the second half of the <sup></sup> 20th century, several reanalyses and observational datasets have shown increases in P–E over global land, although 75% of land areas exhibit no significant changes and both internal variability and observational uncertainty are substantial ( [[#Greve--2014|Greve et al., 2014]] ; [[#Robertson--2016|Robertson et al., 2016]] ). The recently released ERA5 ( [[#Hersbach--2020|Hersbach et al., 2020]] ) showed improvements in the representation of tropical precipitation, although it overestimates global precipitation trends in comparison to ERA-Interim and GPCP ( [[#Nogueira--2020|Nogueira, 2020]] ), and suffers from temporal changes in the annual balance between precipitation and evaporation ( [[#Hersbach--2020|Hersbach et al., 2020]] ). The spatial pattern of P–E trends over 1980–2019 (Figure 2.16a) are largely consistent with the trends in the GPCP v2.3 precipitation dataset (Figure 2.15f and [[#2.3.1.3.4|Section 2.3.1.3.4]] ) and agrees in sign with the trends from other reanalyses such as JRA-55 and MERRA-2 (L. [[#Yu--2020|]] [[#Yu--2020|Yu et al., 2020]] ).

<div id="_idContainer046" class="Basic-Text-Frame"></div>

[[File:065e14ccefd8d7dafb150d10c9e869fc IPCC_AR6_WGI_Figure_2_16.png]]

'''Figure 2.''' '''16 |''' '''Changes in precipitation minus evaporation. (a)''' Trends in precipitation minus evaporation (P–E) between 1980 and 2019. Trends are calculated using OLS regression with significance assessed following AR(1) adjustment after [[#Santer--2008|Santer et al. (2008)]] (‘×’ marks denote non-significant trends). Time series of '''(b)''' global, '''(c)''' land-only and '''(d)''' ocean-only average annual P–E (mm day <sup>–1</sup> ). Further details on data sources and processing are available in the chapter data table (Table 2.SM.1).

A variety of reanalysis products exhibit diverse temporal evolutions of P–E (Figure 2.16b–d). Globally MERRA-2, ERA20C and ERA20CM exhibit little change whereas JRA-55, ERA5 and 20CRv3 all imply long-term changes (Figure 2.16d). A potential limitation in estimating P–E from some reanalysis products is readily apparent when considering the temporal evolution of global P–E from CFSR and MERRA (Figure 2.16d) which both exhibit strong discontinuities over the global ocean in the late 1990s. Over global land as a whole, precipitation exceeds evaporation (P–E &gt;0) for all the reanalysis products (Figure 2.16c), with decreasing trends in P–E for ERA5 and JRA-55 and increasing trends for MERRA-2 and CFSR. The P–E over the global ocean is negative (evaporation exceeding precipitation) for most reanalyses (Figure 2.16d), with declining trends in ERA5 and MERRA-2 dominated by trends in evaporation ( [[#Bosilovich--2017|Bosilovich et al., 2017]] ; [[#Hersbach--2020|Hersbach et al., 2020]] ) (Figure 2.16d). The recent increase in ocean evaporation was also documented for several reanalyses ( [[#Craig--2017|Craig et al., 2017]] ) and in satellite data ( [[#Andersson--2011|Andersson et al., 2011]] ; [[#Robertson--2014|Robertson et al., 2014]] ), although with considerable differences between available estimates ( [[#Chandanpurkar--2017|Chandanpurkar et al., 2017]] ; L. [[#Yu--2020|]] [[#Yu--2020|Yu et al., 2020]] ). An alternative indirect approach to estimate P–E changes is based on near-surface ocean salinity ( [[#2.3.3.2|Section 2.3.3.2]] ), which is partially driven by the freshwater flux at the ocean surface. The near-surface salinity trends are more spatially coherent compared to those revealed by P–E estimates from reanalyses, with an intensification of the water cycle over oceans, especially in subtropical regions ( [[#Durack--2012|Durack et al., 2012]] ; [[#Skliris--2014|Skliris et al., 2014]] ; L. [[#Yu--2020|]] [[#Yu--2020|Yu et al., 2020]] ). However, the precise rate of water cycle intensification implied by salinity trends is sensitive to methodological choices (e.g., [[#Skliris--2016|Skliris et al., 2016]] ; [[#Zika--2018|Zika et al., 2018]] ).

In conclusion, observational uncertainty yields ''low confidence'' in globally averaged trends in P–E over the 20th century, with a spatial pattern dominated by precipitation changes over land and by evaporation increases over the ocean. Different reanalyses disagree on the sign of long-term changes in the global mean P–E.

<div id="2.3.1.3.6" class="h4-container"></div>

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

===== 2.3.1.3.6 Streamflow =====

<div id="h4-17-siblings" class="h4-siblings"></div>

The AR5 concluded that there was ''low confidence'' in a positive trend in global river discharge during the <sup></sup> 20th century. It noted that many of the largest rivers with long term streamflow records have been impacted by non-climatic human influences such as dam construction or land-use change.

River discharge is monitored widely, although gaps remain at a subcontinental scale over central Asia and Africa ( [[#Wei--2020|Wei et al., 2020]] ). Substantial recent efforts have been made to generate new global streamflow datasets, consolidating observations from many stream gauges to create streamflow indices ( [[#Do--2018|Do et al., 2018]] ; [[#Gudmundsson--2018|Gudmundsson et al., 2018]] ) and gridded products using neural networks ( [[#Barbarossa--2018|Barbarossa et al., 2018]] ) or combinations between observations and reanalyses ( [[#Suzuki--2018|Suzuki et al., 2018]] ; [[#Ghiggi--2019|Ghiggi et al., 2019]] ).

Human intervention on river discharge linked to increases in evapotranspiration and some reduction of intra-annual streamflow variability ( [[#Jaramillo--2015|Jaramillo and Destouni, 2015]] ; [[#Chai--2020|Chai et al., 2020]] ) might affect the detection of trends in extreme daily streamflow events ( [[#Do--2017|Do et al., 2017]] ; [[#Gudmundsson--2019|Gudmundsson et al., 2019]] ). However, these activities have a minor impact on annual streamflow compared to climate variations ( [[#Dai--2009|Dai et al., 2009]] ; [[#Alkama--2013|Alkama et al., 2013]] ). Available global studies post-1950 generally concur that there have been more rivers experiencing decreases than increases in runoff ( [[#Do--2017|Do et al., 2017]] ; [[#Su--2018|Su et al., 2018]] ; [[#Gudmundsson--2019|Gudmundsson et al., 2019]] ; X. [[#Shi--2019|]] [[#Shi--2019|Shi et al., 2019]] ). Most of the rivers have not experienced statistically significant changes in streamflow, and when globally aggregated there is no significant change ( [[#Dai--2017|Dai and Zhao, 2017]] ). Global streamflow variability is strongly modulated by ENSO and PDV, with below-normal global streamflow as a response to El Niño events and vice-versa during La Niña episodes ( [[#Dai--2016|Dai, 2016]] ; [[#Liang--2016|Liang et al., 2016]] ; [[#Kim--2019|Kim, 2019]] ). The response of streamflow to changes in precipitation associated with ENSO and PDV has heterogeneous regional patterns at subcontinental scales (Section 8.3.2.9.1). No significant trends are found for reanalysis-based discharge estimates over 1993 to 2015 ( [[#Chandanpurkar--2017|Chandanpurkar et al., 2017]] ). Uncertainties in global streamflow trends arise predominantly from changes in instrumentation, gauge restoration, recalibration of rating curves, flow regulation or channel engineering ( [[#Alkama--2011|Alkama et al., 2011]] ; [[#Gudmundsson--2018|Gudmundsson et al., 2018]] ; [[#Ghiggi--2019|Ghiggi et al., 2019]] ).

In summary, the sign of global streamflow trends remains uncertain, with slightly more globally gauged rivers experiencing significantly decreasing flows than significantly increasing flows since the 1950s ( ''low confidence'' ).

<div id="2.3.1.4" class="h3-container"></div>

<span id="atmospheric-circulation"></span>
==== 2.3.1.4 Atmospheric Circulation ====

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

This section focuses on large-scale changes in a subset of components of the atmospheric circulation (Cross-Chapter Box 2.2). [[IPCC:Wg1:Chapter:Chapter-8|Chapter 8]] assesses large-scale as well as regional aspects of circulation components and their impact on the hydrological cycle, while ( [[IPCC:Wg1:Chapter:Chapter-11|Chapter 11]] assesses the association of circulation changes and variability with extreme events.

<div id="2.3.1.4.1" class="h4-container"></div>

<span id="the-hadley-and-walker-circulations"></span>
===== 2.3.1.4.1 The Hadley and Walker circulations =====

<div id="h4-18-siblings" class="h4-siblings"></div>

The AR5 reported ''low confidence'' in trends in the strength of the Hadley circulation (HC) and the Walker circulation (WC) due to uncertainties in available reanalysis datasets and the large interannual-to-decadal variability of associated circulation patterns. However, AR5 indicated a ''likely'' widening of the tropical belt since the 1970s, albeit with large uncertainty in the magnitude of this change. There was ''high confidence'' that the post-1990s strengthening of the Pacific WC reversed its weakening observed from the mid-19th century to the 1990s.

Paleo reconstructions of rainfall and trade winds extending over the last 100 kyr show an intensification of the NH HC concurrently with a weakening of the SH HC and a southward shift of the inter tropical convergence zone (ITCZ) during Heinrich stadials ( [[#Deplazes--2013|Deplazes et al., 2013]] ; [[#McGee--2018|McGee et al., 2018]] ; [[#Stríkis--2018|Stríkis et al., 2018]] ; [[#Wendt--2019|Wendt et al., 2019]] ). An intensification of the HC associated with conditions similar to La Niña (northward migrations of both the ITCZ and the SH westerlies) was found in reconstructions for the MH ( [[#McGee--2014|McGee et al., 2014]] ; [[#Mollier-Vogel--2019|Mollier-Vogel et al., 2019]] ). Changes in insolation from the mid to late Holocene favoured a southward migration in the position of the ITCZ and the descending branch of the HC in the NH, approaching its current width and position ( [[#Wirth--2013|Wirth et al., 2013]] ; [[#Thatcher--2020|Thatcher et al., 2020]] ). Tree ring chronologies from the NH mid-latitudes over the last 800 years show that the northern edge of the HC tended to migrate southward during positive phases of ENSO and PDV, with northward shifts during negative phases ( [[#Alfaro-Sánchez--2018|Alfaro-Sánchez et al., 2018]] ). Between 1400 and 1850 CE the HC over both hemispheres and the ITCZ were displaced southward, consistent with occurrence of drought conditions in several NH regions ( [[#Wirth--2013|Wirth et al., 2013]] ; [[#Burn--2014|Burn and Palmer, 2014]] ; [[#Lechleitner--2017|Lechleitner et al., 2017]] ; [[#Alfaro-Sánchez--2018|Alfaro-Sánchez et al., 2018]] ; [[#Flores-Aqueveque--2020|Flores-Aqueveque et al., 2020]] ). Moreover, several proxy records showed not only inter-hemispheric shifts in the ITCZ but a contraction of the tropical belt during 1400–1850 CE, which followed an expansion during 950–1250 CE ( [[#Denniston--2016|Denniston et al., 2016]] ; [[#Griffiths--2016|Griffiths et al., 2016]] ).

From centennial-scale reanalyses, [[#Liu--2012|Liu et al. (2012)]] and [[#D’Agostino--2017|D’Agostino and Lionello (2017)]] found divergent results on HC extent over the last 150 years, although with unanimity upon an intensification of the SH HC. A substantial discrepancy between HC characteristics in centennial-scale reanalyses and in ERA-Interim ( [[#D’Agostino--2017|D’Agostino and Lionello, 2017]] ) since 1979 yields significant questions regarding their ability to capture changes in HC behaviour. Taken together with the existence of apparent non-climatic artefacts in the datasets ( [[#Nguyen--2015|Nguyen et al., 2015]] ), this implies ''low confidence'' in changes in the extent and intensity of HC derived from centennial-scale reanalyses. However, using multiple observational datasets and centennial-scale reanalyses, [[#Bronnimann--2015|Bronnimann et al. (2015)]] identified a southward shift in the NH HC edge from 1945 to 1980 of about 0.25° latitude per decade, consistent with observed changes in global land monsoon precipitation ( [[#2.3.1.4.2|Section 2.3.1.4.2]] ).

Since AR5 several studies based upon a range of metrics and different reanalyses products have suggested that the annual mean HC extent has shifted poleward at an approximate rate of 0.1°–0.5° latitude per decade over the last about 40 years ( [[#Allen--2017|Allen and Kovilakam, 2017]] ; [[#Davis--2017|Davis and Birner, 2017]] ; [[#Grise--2018|Grise et al., 2018]] ; [[#Staten--2018|Staten et al., 2018]] , 2020; [[#Studholme--2018|Studholme and Gulev, 2018]] ; [[#Grise--2020|Grise and Davis, 2020]] ). The observed widening of the annual mean HC, revealed by a variety of metrics, is primarily due to poleward shift of the Northern Hemisphere HC. There have been stronger upward trends in the NH extent of HC after 1992 (Figure 2.17a). The estimated magnitude of the recent changes based on modern-era reanalyses is not as large as that in AR5, due to apparent biases in older-generation reanalyses ( [[#Grise--2019|Grise et al., 2019]] ). Moreover, large interannual variability leads to uncertainties in estimates of long-term changes ( [[#Nguyen--2013|Nguyen et al., 2013]] ; [[#Garfinkel--2015b|Garfinkel et al., 2015b]] ; [[#Seviour--2018|Seviour et al., 2018]] ; [[#Staten--2018|Staten et al., 2018]] ), particularly for the NH given its zonal asymmetries ( [[#Staten--2020|Staten et al., 2020]] ; [[#Wang--2020|Wang et al., 2020]] ). These large-scale features of the HC based on reanalyses agree with estimates revealed from the Integrated Global Radiosonde Archive (IGRA) during 1979–2012 ( [[#Lucas--2015|Lucas and Nguyen, 2015]] ; [[#Mathew--2016|Mathew et al., 2016]] ). Recent trends based on reanalyses indicate a larger seasonal widening in the HC for summer and autumn in each hemisphere, although the magnitude of changes in HC extent is strongly dependent on dataset and metrics used ( [[#Grise--2018|Grise et al., 2018]] ; Y. [[#Hu--2018|]] [[#Hu--2018|Hu et al., 2018]] ; [[#Staten--2018|Staten et al., 2018]] ). The shifts in the HC position were accompanied by a narrowing ITCZ over the Atlantic and Pacific basins, with no significant change in its location and increases in the precipitation intensity ( [[#Byrne--2018|Byrne et al., 2018]] ).

<div id="_idContainer048" class="Basic-Text-Frame"></div>

[[File:edccb51f7c800aa7409e4500c5eec32e IPCC_AR6_WGI_Figure_2_17.png]]

'''Figure''' '''2.17 |''' '''Time series of the annual mean Northern Hemisphere (NH, top curves) and Southern Hemisphere (SH, bottom curves) Hadley cell extent (a) and Hadley cell intensity (b) since 1979.''' Further details on data sources and processing are available in the chapter data table (Table 2.SM.1).

Trends in the HC intensity since 1979 differ between reanalyses, although there is a tendency toward HC intensification (Figure 2.17b; [[#Nguyen--2013|Nguyen et al., 2013]] ; [[#Chen--2014|Chen et al., 2014]] ; [[#D’Agostino--2017|D’Agostino and Lionello, 2017]] ; R. [[#Huang--2019|Huang et al., 2019]] ), which is more marked in the NH than the SH ( [[#Studholme--2018|Studholme and Gulev, 2018]] ). However, the ability of reanalyses to represent the HC strength has been questioned due to inaccurate representation of latent heating distribution, which is directly related to tropical convection and influences the HC dynamics ( [[#Chemke--2019|Chemke and Polvani, 2019]] ; [[#Mathew--2019|Mathew and Kumar, 2019]] ).

Paleo evidence during the LGM indicates a weaker WC over the Indian Ocean ( [[#DiNezio--2018|DiNezio et al., 2018]] ; [[#Windler--2019|Windler et al., 2019]] ) with a stronger Pacific WC ( [[#DiNezio--2013|DiNezio and Tierney, 2013]] ). During the Holocene, a transition from a strong WC located more westward during the Early-to-Mid Holocene towards a weak and eastward shifted WC during the late Holocene was inferred from proxy records from the Pacific Warm Pool and South East Asia ( [[#Barr--2019|Barr et al., 2019]] ; [[#Dang--2020|Dang et al., 2020]] ; [[#Griffiths--2020|Griffiths et al., 2020]] ), in concurrence with changes in ENSO activity ( [[#2.4.2|Section 2.4.2]] ). Reconstructions for the CE showed weakened WC during 1000–1250 and since 1850, with an intensified circulation during 1500–1850 CE ( [[#Xu--2016|Xu et al., 2016]] ; [[#Deng--2017|Deng et al., 2017]] ).

Considering instrumental records, there is considerable interdecadal variability in the strength of the WC, resulting in time-period dependent magnitude and even sign of trends ( [[#Carilli--2015|Carilli et al., 2015]] ; [[#Bordbar--2017|Bordbar et al., 2017]] ; [[#Hou--2018|Hou et al., 2018]] ), with some studies reporting weakening over the 20th century (e.g., [[#Power--2011|Power and Kociuba, 2011]] ; [[#Liu--2019|Liu et al., 2019]] ), while others reported strengthening (Z. [[#Li--2020|]] [[#Li--2020|Li et al., 2020]] ), particularly over the last 30–40 years (e.g., [[#Hu--2013|Hu et al., 2013]] ; [[#L’Heureux--2013|L’Heureux et al., 2013]] ; [[#Yim--2017|Yim et al., 2017]] ). Based on estimation of changes in mid-tropospheric velocity from changes in observed cloud cover, [[#Bellomo--2015|Bellomo and Clement (2015)]] suggest a weakening and eastward shift of the WC over 1920–2010, however the robustness of this signal is questionable due to high uncertainty in the ship-reported cloud data used before 1954. Using centennial-scale 20CR reanalysis [[#Tseng--2019|Tseng et al. (2019)]] showed that the vertical westerly wind shear over the western Pacific does not indicate any long-term change during 1900–1980, but shows a marked increase since the 1980s that is not present in ERA-Interim and JRA-55, again calling into question the ability of centennial-scale reanalyses to capture tropical circulation changes. Recent strengthening together with a westward shift of the WC ( [[#Bayr--2014|Bayr et al., 2014]] ; [[#Ma--2016|Ma and Zhou, 2016]] ) was identified across several reanalysis products and observational datasets, and using different metrics for quantifying WC. Nevertheless, satellite observations of precipitation and analyses of upper tropospheric humidity suggest substantially weaker strengthening of the WC than implied by reanalyses ( [[#Chung--2019|Chung et al., 2019]] ). This recent strengthening in the WC is associated with enhanced precipitation in the tropical western Pacific, anomalous westerlies in the upper troposphere, strengthened downwelling in the central and eastern tropical Pacific, and anomalous surface easterlies in the western and central tropical Pacific ( [[#Dong--2013|Dong and Lu, 2013]] ; [[#McGregor--2014|McGregor et al., 2014]] ; [[#Choi--2016|Choi et al., 2016]] ). Positive trends in sea level pressure over the eastern Pacific and concurrent negative trends over the Indonesian region result in a pattern implying a shift towards a La Niña-like WC regime, with strengthening of the Pacific Trade Winds mainly over 1979–2012 ( [[#L’Heureux--2013|L’Heureux et al., 2013]] ; [[#England--2014|England et al., 2014]] ; [[#Sohn--2016|Sohn et al., 2016]] ; [[#Zhao--2019|Zhao and Allen, 2019]] ). Seasonal assessment of the WC showed significant changes in the vertical westerly wind shear over the Pacific during the austral summer and autumn implying a strengthening ( [[#Clem--2017|Clem et al., 2017]] ).

In summary, there has been a ''likely'' widening of the Hadley circulation since the 1980s, mostly due to its extension in the NH, although there is only ''medium confidence'' in the extent of the changes. This has been accompanied by a strengthening of the Hadley circulation, particularly in the NH ( ''medium confidence'' ). There is ''low confidence'' in the estimation of long-term trends in the strength of the Walker circulation, which are time period dependent and subject to dataset uncertainties. Trends since 1980 are better characterized and consistent with a ''very likely'' strengthening that resembles a La Niña-like Walker circulation and a westward shift of the Walker circulation, although with ''medium confidence'' in the magnitude of the changes, arising from the differences between satellite observations and reanalysis products.

<div id="2.3.1.4.2" class="h4-container"></div>

<span id="global-monsoon-gm-changes"></span>

===== 2.3.1.4.2 Global monsoon (GM) changes =====

<div id="h4-19-siblings" class="h4-siblings"></div>

The AR5 reported a weakening of the global monsoon (GM) circulation as well as a decrease of global land monsoon rainfall over the second half of the 20th century. Nevertheless, there was ''low confidence'' in the observed circulation trends due to uncertainties in reanalysis products and in the definition of the monsoon area. From a paleo perspective, AR5 only assessed regional monsoon changes.

New research based on high-resolution proxies reinforces previous findings on the influence of orbital cycles on GM variability on millennial time scales. The intensity of the monsoon systems is generally out of phase between hemispheres, being associated with the precession cycle (about 21–23 kyr) ( [[#An--2015|An et al., 2015]] ; P.X. [[#Wang--2017|Wang et al., 2017]] ; [[#Seth--2019|Seth et al., 2019]] ), with intensified NH monsoon systems during precession minima ( [[#Toucanne--2015|Toucanne et al., 2015]] ; [[#Wagner--2019|Wagner et al., 2019]] ). The eccentricity forcing (about 100 kyr cycle) shows stronger GM during interglacial periods (P.X. [[#Wang--2014|]] [[#Wang--2014|Wang et al., 2014]] , 2017; [[#An--2015|An et al., 2015]] ; [[#Mohtadi--2016|Mohtadi et al., 2016]] ). Changes in obliquity (about 41 kyr cycle) modify the strength of monsoon systems, with increased summer monsoon rainfall when obliquity is maximal (Y. [[#Liu--2015|]] [[#Liu--2015|Liu et al., 2015]] b; [[#Mohtadi--2016|Mohtadi et al., 2016]] ). Millennial scale variability in GM during the LDT was also linked to the occurrences of Heinrich stadials, resulting in weakened NH monsoons and intensified SH monsoons ( [[#An--2015|An et al., 2015]] ; P.X. [[#Wang--2017|Wang et al., 2017]] ; [[#Margari--2020|Margari et al., 2020]] ).

An intensification of the NH monsoons in the early to mid-Holocene with increased precipitation and regional expansions of rainfall areas identified through a variety of proxy records is shown by [[#Biasutti--2018|Biasutti et al. (2018)]] and P.X. [[#Wang--2017|Wang et al. (2017)]] . The response for the SH monsoons during this period indicates a weakening in both summer and winter precipitation (P.X. [[#Wang--2014|]] [[#Wang--2014|Wang et al., 2014]] , 2017; [[#Sachs--2018|Sachs et al., 2018]] ). A decline in GM precipitation and a retraction of the northern fringes of monsoon areas was inferred from the mid-Holocene onwards, with some regions experiencing wetter conditions during the mid to late Holocene compared with present and a strengthening of the SH monsoons (P.X. [[#Wang--2014|]] [[#Wang--2014|Wang et al., 2014]] , 2017; [[#Sachs--2018|Sachs et al., 2018]] ). For the CE, GM reconstructions exhibit inter-hemispheric contrast during the period 950–1250 CE, with intensified NH monsoons and weakened SH monsoons, and the opposite pattern during 1400–1850 CE (P.X. [[#Wang--2014|]] [[#Wang--2014|Wang et al., 2014]] ; [[#An--2015|An et al., 2015]] ).

Direct observations highlight that the GM land precipitation, particularly over the NH, experienced a slight increase from 1900 through the early 1950s, followed by an overall decrease from the 1950s to the 1980s, and then an increase to present ( [[#Kitoh--2013|Kitoh et al., 2013]] ; [[#Wang--2018|]] [[#Wang--2018|B. Wang et al., 2018]] , 2021; X. [[#Huang--2019b|]] [[#Huang--2019|Huang et al., 2019]] b ). This highlights the existence of multi-decadal variations in the NH monsoon circulation patterns and precipitation intensity ( [[#Wang--2013|Wang et al., 2013]] ; P.X. [[#Wang--2014|]] [[#Wang--2014|Wang et al., 2014]] , 2017; [[#Monerie--2019|Monerie et al., 2019]] ). An overall increase in monsoon precipitation during extended boreal summer (JJAS) over the NH since 1979 is revealed by GPCP ( [[#Deng--2018|Deng et al., 2018]] ; [[#Han--2019|Han et al., 2019]] ) and CMAP for 1980–2010 ( [[#Jiang--2016|Jiang et al., 2016]] ). SH summer monsoon behaviour is dominated by strong interannual variability and large regional differences ( [[#Kitoh--2013|Kitoh et al., 2013]] ; [[#Lin--2014|Lin et al., 2014]] ; [[#Jiang--2016|Jiang et al., 2016]] ; [[#Kamae--2017|Kamae et al., 2017]] ; [[#Deng--2018|Deng et al., 2018]] ; [[#Han--2019|Han et al., 2019]] ), with no significant trends reported by GPCP and CMAP ( [[#Deng--2018|Deng et al., 2018]] ). Uncertainty predominantly arises from the observed increase in tropical precipitation seasonality ( [[#Feng--2013|Feng et al., 2013]] ) and the estimation of GM precipitation over the ocean areas, leading to a large apparent spread across datasets ( [[#Kitoh--2013|Kitoh et al., 2013]] ; [[#Kamae--2017|Kamae et al., 2017]] ).

In summary, observed trends during the last century indicate that the GM precipitation decline reported in AR5 has reversed since the 1980s, with a ''likely'' increase mainly due to a significant positive trend in the NH summer monsoon precipitation ( ''medium confidence'' ). However, GM precipitation has exhibited large multi-decadal variability over the last century, creating ''low confidence'' in the existence of centennial-length trends in the instrumental record. Proxy reconstructions show a ''likely'' NH monsoons weakening since the mid-Holocene, with opposite behaviour for the SH monsoons.

<div id="2.3.1.4.3" class="h4-container"></div>

<span id="extratropical-jets-storm-tracks-and-blocking"></span>

===== 2.3.1.4.3 Extratropical jets, storm tracks, and blocking =====

<div id="h4-20-siblings" class="h4-siblings"></div>

The AR5 reported a ''likely'' poleward shift of storm tracks and jet streams since the 1970s from different datasets, variables and approaches. These trends were consistent with the HC widening and the poleward shifting of the circulation features since the 1970s. There was ''low confidence'' in any large-scale change in blocking.

Proxy records consistent with modelling results imply a southward shift of the storm tracks over the North Atlantic during the LGM ( [[#Raible--2021|Raible et al., 2021]] ). A variety of proxies are available for the changes in the position of the extratropical jets/westerlies during the Holocene. Recent syntheses of moisture-sensitive proxy records indicate drier-than-present conditions over mid-latitudes of western North America ( [[#Hermann--2018|Hermann et al., 2018]] ; [[#Liefert--2020|Liefert and Shuman, 2020]] ) during the MH, which together with a weakened Aleutian Low ( [[#Bailey--2018|Bailey et al., 2018]] ) implies that the winter North Pacific jetstream was shifted northward. A synthesis of lines of evidence from the SH indicates that the westerly winds were stronger over 14–5 ka, followed by regional asymmetry after 5 ka ( [[#Fletcher--2012|Fletcher and Moreno, 2012]] ). There is no consensus on the shifts of the SH westerlies with some studies implying poleward migrations ( [[#Lamy--2010|Lamy et al., 2010]] ; [[#Voigt--2015|Voigt et al., 2015]] ; [[#Turney--2017|Turney et al., 2017]] ; [[#Anderson--2018|Anderson et al., 2018]] ) and others suggesting an equatorward shift ( [[#Kaplan--2016|Kaplan et al., 2016]] ) in the MH.

During 950–1400 CE, hydroclimate indicators suggest a northward shift of Pacific storm tracks over North America ( [[#McCabe-Glynn--2013|McCabe-Glynn et al., 2013]] ; [[#Steinman--2014|Steinman et al., 2014]] ) which was comparable in magnitude to that over 1979–2015 (J. [[#Wang--2017a|]] [[#Wang--2017|Wang et al., 2017]] a ). Storm tracks over the North Atlantic-European sector shifted northward as indicated by multi-proxy indicators over the North Atlantic ( [[#Wirth--2013|Wirth et al., 2013]] ; [[#Orme--2017|Orme et al., 2017]] ) and Mediterranean ( [[#Roberts--2012|Roberts et al., 2012]] ). Reconstructed westerly winds in the SH suggest a poleward shift ( [[#Lamy--2010|Lamy et al., 2010]] ; [[#Schimpf--2011|Schimpf et al., 2011]] ; [[#Goodwin--2014|Goodwin et al., 2014]] ; [[#Koffman--2014|Koffman et al., 2014]] ; [[#Moreno--2018|Moreno et al., 2018]] ), with latitudinal change comparable to that during recent decades ( [[#Swart--2012|Swart and Fyfe, 2012]] ; [[#Manney--2018|Manney and Hegglin, 2018]] ).

Multiple reanalyses show that since 1979 the subtropical jet wind speeds have generally increased in winter and decreased in summer in both hemispheres, but the trends are regionally dependent ( [[#Pena-Ortiz--2013|Pena-Ortiz et al., 2013]] ; [[#Manney--2018|Manney and Hegglin, 2018]] ; S.H. [[#Lee--2019|]] [[#Lee--2019|Lee et al., 2019]] ). Over NH mid-latitudes, the summer zonal wind speeds have weakened in the mid-troposphere ( [[#Francis--2012|Francis and Vavrus, 2012]] ; [[#Coumou--2014|Coumou et al., 2014]] , 2015; [[#Haimberger--2017|Haimberger and Mayer, 2017]] ). Meanwhile there are indications of enhanced jetstream meandering in boreal autumn at the hemispheric scale ( [[#Francis--2015|Francis and Vavrus, 2015]] ; [[#Di%20Capua--2016|Di Capua and Coumou, 2016]] ), whereas the regional arrangement of meandering depends on the background atmospheric state ( [[#Cohen--2020|Cohen et al., 2020]] ). These meandering trends, however, are sensitive to the metrics used ( [[#Screen--2013|Screen and Simmonds, 2013]] ; [[#Hassanzadeh--2014|Hassanzadeh et al., 2014]] ; [[#Cattiaux--2016|Cattiaux et al., 2016]] ; [[#Vavrus--2018|Vavrus, 2018]] ). Hypothesized links to Arctic warming are assessed in Cross-Chapter Box 10.1.

Multiple reanalyses and radiosonde observations show an increasing number of extratropical cyclones over the NH since the 1950s ( [[#Chang--2016|Chang and Yau, 2016]] ; X.L. [[#Wang--2016|]] [[#Wang--2016|]] [[#Wang--2016|]] [[#Wang--2016|]] [[#Wang--2016|]] [[#Wang--2016|Wang et al., 2016]] ). The positive trends are generally consistent among reanalyses since 1979, though with considerable spread ( [[#Tilinina--2013|Tilinina et al., 2013]] ; X.L. [[#Wang--2016|]] [[#Wang--2016|]] [[#Wang--2016|]] [[#Wang--2016|]] [[#Wang--2016|]] [[#Wang--2016|Wang et al., 2016]] ). In recent decades the number of deep extratropical cyclones has increased over the SH (Section 8.3.2.8.1 and Figure 8.12; [[#Reboita--2015|Reboita et al., 2015]] ; X.L. [[#Wang--2016|]] [[#Wang--2016|]] [[#Wang--2016|]] [[#Wang--2016|]] [[#Wang--2016|]] [[#Wang--2016|Wang et al., 2016]] ), while the number of deep cyclones has decreased in the NH in both winter and summer ( [[#Neu--2013|Neu et al., 2013]] ; [[#Coumou--2015|Coumou et al., 2015]] ; [[#Chang--2016|Chang et al., 2016]] ; J. [[#Wang--2017a|]] [[#Wang--2017|Wang et al., 2017]] a ; [[#Gertler--2019|Gertler and O’Gorman, 2019]] ). The regional changes for different intensity extratropical cyclones are assessed in Section 8.3.2.8.1. The assessment of trends is complicated by strong interannual to decadal variability, sensitivity to dataset choice and resolution ( [[#Tilinina--2013|Tilinina et al., 2013]] ; [[#Lucas--2014|Lucas et al., 2014]] ; X.L. [[#Wang--2016|]] [[#Wang--2016|]] [[#Wang--2016|]] [[#Wang--2016|]] [[#Wang--2016|]] [[#Wang--2016|Wang et al., 2016]] ; [[#Pepler--2018|Pepler et al., 2018]] ; [[#Rohrer--2018|Rohrer et al., 2018]] ) and cyclone identification/tracking methods ( [[#Neu--2013|Neu et al., 2013]] ; [[#Grieger--2018|Grieger et al., 2018]] ). Thus there is overall ''low'' ''confidence'' for recent changes in global extratropical storm tracks.

A consistent poleward shift of the tropospheric extratropical jets since 1979 is reported by multiple reanalyses (Figure 2.18; [[#Davis--2012|Davis and Rosenlof, 2012]] ; [[#Davis--2013|Davis and Birner, 2013]] ; [[#Pena-Ortiz--2013|Pena-Ortiz et al., 2013]] ; [[#Manney--2018|Manney and Hegglin, 2018]] ), and radiosonde winds ( [[#Allen--2012|Allen et al., 2012]] ). This is generally consistent with the previously reported shifts retrieved from satellite temperature observations ( [[#Fu--2011|Fu and Lin, 2011]] ; [[#Davis--2012|Davis and Rosenlof, 2012]] ). After the 1960s the magnitude of meridional shifts in extratropical jets over both the North Atlantic and North Pacific in August is enhanced compared to multi-century variability ( [[#Trouet--2018|Trouet et al., 2018]] ). Despite some regional differences ( [[#Woollings--2014|Woollings et al., 2014]] ; [[#Norris--2016|Norris et al., 2016]] ; J. [[#Wang--2017a|]] [[#Wang--2017|Wang et al., 2017]] a ; [[#Xue--2017|Xue and Zhang, 2017]] ; [[#Ma--2018|Ma and Zhang, 2018]] ; [[#Melamed-Turkish--2018|Melamed-Turkish et al., 2018]] ), overall poleward deflection of storm tracks in boreal winter over both the North Atlantic and the North Pacific was identified during 1979–2010 ( [[#Tilinina--2013|Tilinina et al., 2013]] ). Over the SH extra-tropics there is a similarly robust poleward shift in the polar jet since 1979 ( [[#Pena-Ortiz--2013|Pena-Ortiz et al., 2013]] ; [[#Manney--2018|Manney and Hegglin, 2018]] ; [[#WMO--2018|WMO, 2018]] ), although after 2000 the December–January–February (DJF) tendency to poleward shift of the SH jet stream position ceased ( [[#Banerjee--2020|Banerjee et al., 2020]] ). The general poleward movement in midlatitude jet streams ( [[#Lucas--2014|Lucas et al., 2014]] ) is consistent with the expansion of the tropical circulation ( [[#2.3.1.4.1|Section 2.3.1.4.1]] ). The changes of extratropical jets and westerlies are also related to the annular modes of variability ( [[#2.4|Section 2.4]] and Annex IV).

<div id="_idContainer050" class="Basic-Text-Frame"></div>

[[File:fd4b5ce5996815b4bafa896df1e3d7fb IPCC_AR6_WGI_Figure_2_18.png]]

'''Figure 2.1''' '''8 |''' '''Trends in ERA5 zonal-mean zonal wind speed.''' Shown are '''(a)''' DJF (December–January–February); '''(b)''' MAM (March–April–May); '''(c)''' JJA (June–July–August); and '''(d)''' SON (September–October–November). Climatological zonal winds during the data period are shown in solid contour lines for westerly winds and in dashed lines for easterly. Trends are calculated using OLS regression with significance assessed following AR(1) adjustment after [[#Santer--2008|Santer et al. (2008)]] (‘×’ marks denote non-significant trends). Further details on data sources and processing are available in the chapter data table (Table 2.SM.1).

Robust trends in blocking have only been found in certain regions and specific seasons during recent decades. Increases in blocking frequency have occurred over low-latitude regions in the North Atlantic in boreal winter ( [[#Davini--2012|Davini et al., 2012]] ), the South Atlantic in austral summer ( [[#Dennison--2016|Dennison et al., 2016]] ) and the southern Indian Ocean in austral spring ( [[#Schemm--2018|Schemm, 2018]] ). Over the subpolar North Atlantic sustained periods of positive Greenland blocking were identified during 1870–1900 and from the late 1990s to 2015 ( [[#Hanna--2015|Hanna et al., 2015]] ). Further analysis of association of Greenland blocking with the NAM is provided in [[#2.4.1.1|Section 2.4.1.1]] . Meanwhile, a reduced blocking frequency has been found over winter in Siberia ( [[#Davini--2012|Davini et al., 2012]] ) and the south-western Pacific in austral spring ( [[#Schemm--2018|Schemm, 2018]] ). Over eastern European Russia and western Siberia (40°E–100°E) a tendency towards longer blocking events was reported by [[#Luo--2016|Luo et al. (2016)]] for 2000–2013 and by [[#Tyrlis--2020|Tyrlis et al. (2020)]] for 1979–2017. Inter-annual variance in the number of blocking events over the SH ( [[#Oliveira--2014|Oliveira et al., 2014]] ) and North Atlantic ( [[#Kim--2015|Kim and Ha, 2015]] ) has enhanced. Blocking events and their trends are sensitive to choice of datasets, calculation periods and methods ( [[#Cheung--2013|Cheung et al., 2013]] ; [[#Barnes--2014|Barnes et al., 2014]] ; [[#Pepler--2018|Pepler et al., 2018]] ; [[#Rohrer--2018|Rohrer et al., 2018]] ; [[#Woollings--2018b|Woollings et al., 2018b]] ; [[#Kononova--2020|Kononova and Lupo, 2020]] ). As a result, hemispheric and global trends in blocking frequency have overall ''low'' ''confidence.''

In summary, the total number of extratropical cyclones has ''likely'' increased since the 1980s in the NH ( ''low confidence'' ), but with fewer deep cyclones particularly in summer. The number of strong extratropical cyclones has ''likely'' increased in the SH ( ''medium confidence'' ). The extratropical jets and cyclone tracks have ''likely'' been shifting poleward in both hemispheres since the 1980s with marked seasonality in trends ( ''medium confidence'' ). There is ''low confidence'' in shifting of extratropical jets in the NH during the mid-Holocene and over 950–1400 CE to latitudes that ''likely'' were similar to those since 1979. There is ''low confidence'' in observed global-scale changes in the occurrence of blocking events.

<div id="2.3.1.4.4" class="h4-container"></div>

<span id="surface-wind-and-sea-level-pressure"></span>

===== 2.3.1.4.4 Surface wind and sea level pressure =====

<div id="h4-21-siblings" class="h4-siblings"></div>

The AR5 concluded that surface winds over land had generally weakened. The ''confidence'' for both land and ocean surface wind trends was ''low'' owing to uncertainties in datasets and measures used. Sea level pressure (SLP) was assessed to have ''likely'' decreased from 1979–2012 over the tropical Atlantic and increased over large regions of the Pacific and South Atlantic, but trends were sensitive to the period analysed.

Terrestrial in situ wind datasets have been updated and the quality-control procedures have been improved, with particular attention to homogeneity and to better retaining true extreme values ( [[#Dunn--2012|Dunn et al., 2012]] , 2014, 2016). Global mean land wind speed (excluding Australia) from HadISD for 1979–2018 shows a reduction (stilling) of 0.063 m s <sup>–1</sup> per decade ( [[#Azorin-Molina--2019|Azorin-Molina et al., 2019]] ). Trends are broadly insensitive to the subsets of stations used. Although the meteorological stations are unevenly distributed worldwide and sparse in South America and Africa, the majority exhibit stilling particularly in the NH (Figure 2.19). Regionally, strong decreasing trends are reported in central Asia and North America (–0.106 and –0.084 m s <sup>–1</sup> per decade respectively) during 1979–2018 ( [[#McVicar--2012|McVicar et al., 2012]] ; [[#Vautard--2012|Vautard et al., 2012]] ; J. [[#Wu--2018|]] [[#Wu--2018|Wu et al., 2018]] ; [[#Azorin-Molina--2019|Azorin-Molina et al., 2019]] ). This stilling tendency has reversed after 2010 and the global mean surface winds have strengthened ( [[#Zeng--2019b|Zeng et al., 2019b]] ; [[#Azorin-Molina--2020|Azorin-Molina et al., 2020]] ), although the robustness of this reversal is unclear given the short period and interannual variability ( [[#Kousari--2013|Kousari et al., 2013]] ; [[#Kim--2015|Kim and Paik, 2015]] ; [[#Azorin-Molina--2019|Azorin-Molina et al., 2019]] ).

<div id="_idContainer052" class="Basic-Text-Frame"></div>

[[File:26ea185ca5ced7098b8a78c982776822 IPCC_AR6_WGI_Figure_2_19.png]]

'''Figure 2.19''' '''|''' '''Trends in surface wind speed. (a)''' Station observed winds from the integrated surface database (HadISD v2.0.2.2017f); '''(b)''' Cross-Calibrated Multi-Platform wind product; '''(c)''' ERA5; and '''(d)''' wind speed from the Objectively Analyzed Air-Sea Heat Fluxes dataset, release 3 (OAFLUX, release 3). White areas indicate incomplete or missing data. Trends are calculated using OLS regression with significance assessed following AR(1) adjustment after [[#Santer--2008|Santer et al. (2008)]] ; ‘×’ marks denote non-significant trends. Further details on data sources and processing are available in the chapter data table (Table 2.SM.1).

Over the ocean, datasets demonstrate considerable disagreement in surface wind speed trends and spatial features ( [[#Kent--2013|Kent et al., 2013]] ). Global ocean surface winds from NOCv2.0 demonstrate upward trends of about 0.11 m s <sup>–1</sup> per decade (1979–2015) with somewhat smaller trends from WASwind for 1979–2011 ( [[#Azorin-Molina--2017|Azorin-Molina et al., 2017]] , 2019). The trends are consistent until 1998, but diverge thereafter. Both ERA5 and JRA-55 reanalyses show consistently increasing global marine wind speeds over 1979–2015, though flattening since 2000, whereas MERRA-2 agrees until 1998, but then exhibits increased variability and an overall decrease in the last two decades ( [[#Azorin-Molina--2019|Azorin-Molina et al., 2019]] ). This agrees with estimates by [[#Sharmar--2021|Sharmar et al. (2021)]] showing upward ocean wind trends from 1979 to 2000 which are consistent in ERA-Interim, ERA5 and MERRA-2, but disagree with CFSR trends for the same period. Over 2000–2019 all reanalyses show diverging tendencies. An updated multiplatform satellite database (comprising data from altimeters, radiometers, and scatterometers) from 1985–2018 shows small increases in mean wind speed over the global ocean, with the largest increase observed in the Southern Ocean ( [[#Young--2019|Young and Ribal, 2019]] ), consistent with signals in ERA-Interim, ERA5 and MERRA-2 ( [[#Sharmar--2021|Sharmar et al., 2021]] ). Overall, most products suggest positive trends over the Southern Ocean, western North Atlantic and the tropical eastern Pacific since the early 1980s.

The modern era reanalyses exhibit SLP increases over the SH subtropics with stronger increases in austral winter over 1979–2018. Over the NH, SLP increased over the mid-latitude Pacific in boreal winter and decreased over the eastern subtropical and mid-latitude North Atlantic in boreal summer. Discrepancies in the low-frequency variations during the first half of the 20th century exist in the centennial-scale reanalysis products ( [[#Befort--2016|Befort et al., 2016]] ). Overall, modern reanalysis datasets support the AR5 conclusion that there is no clear signal for trends in the strength and position of the permanent and quasi-permanent pressure centres of action since the 1950s. Instead, they highlight multi-decadal variations. Large-scale SLP is strongly associated with the changes in modes of variability ( [[#2.4|Section 2.4]] and Annex IV).

In summary, since the 1970s a worldwide weakening of surface wind has ''likely'' occurred over land, particularly marked in the NH, with ''low confidence'' in a recent partial recovery since around 2010. Differences between available wind speed estimates lead to ''low confidence'' in trends over the global ocean as a whole but with most estimates showing strengthening globally over 1980–2000 and over the last four decades in the Southern Ocean, western North Atlantic and the tropical eastern Pacific.

<div id="2.3.1.4.5" class="h4-container"></div>

<span id="stratospheric-polar-vortex-and-sudden-warming-events"></span>

===== 2.3.1.4.5 Stratospheric polar vortex and sudden warming events =====

<div id="h4-22-siblings" class="h4-siblings"></div>

The AR5 assessed changes in the polar vortices and reported a ''likely'' decrease in the lower-stratospheric geopotential heights over Antarctica in spring and summer at least since 1979.

Multiple definitions for the polar vortex strength and sudden stratospheric warming (SSW) events have been proposed and compared ( [[#Butler--2015|Butler et al., 2015]] ; [[#Palmeiro--2015|Palmeiro et al., 2015]] ; [[#Waugh--2017|Waugh et al., 2017]] ; [[#Butler--2018|Butler and Gerber, 2018]] ), and new techniques identifying daily vortex patterns and SSWs have been developed (D.M. [[#Mitchell--2013|]] [[#Mitchell--2013|Mitchell et al., 2013]] ; [[#Kretschmer--2018|Kretschmer et al., 2018]] ). Errors in reanalysis stratospheric winds were assessed and discrepancies in stratospheric atmospheric circulation and temperatures between reanalyses, satellites and radiosondes have been reported (D.M. [[#Mitchell--2013|]] [[#Mitchell--2013|Mitchell et al., 2013]] ; [[#Duruisseau--2017|Duruisseau et al., 2017]] ).

The northern stratospheric polar vortex has varied intra-seasonally and with altitude during recent decades. Multiple reanalysis and radiosonde datasets show that the midwinter lower stratospheric geopotential height (150 hPa) over the polar region north of 60°N has increased significantly since the early 1980s ( [[#Bohlinger--2014|Bohlinger et al., 2014]] ; [[#Garfinkel--2017|Garfinkel et al., 2017]] ). This signal extends to the middle and upper stratosphere. In January-February zonal winds north of 60°N at 10 hPa have been weakening ( [[#Kim--2014|Kim et al., 2014]] ; [[#Kretschmer--2018|Kretschmer et al., 2018]] ). Daily atmospheric circulation patterns over the northern polar stratosphere exhibit a decreasing frequency of strong vortex events and commensurate increase in more-persistent weak events, which largely explains the observed significant weakening of the vortex during 1979–2015 ( [[#Kretschmer--2018|Kretschmer et al., 2018]] ). The northern polar vortex has weakened in early winter but strengthened during late winter ( [[#Bohlinger--2014|Bohlinger et al., 2014]] ; [[#Garfinkel--2015a|Garfinkel et al., 2015a]] , 2017; [[#Ivy--2016|Ivy et al., 2016]] ; [[#Seviour--2017|Seviour, 2017]] ; [[#Kretschmer--2018|Kretschmer et al., 2018]] ). In the middle and upper stratosphere, a strengthening trend of the northern polar vortex during DJF has occurred since 1998, contrasting the weakening trend beforehand (D. [[#Hu--2018|]] [[#Hu--2018|Hu et al., 2018]] ). The position of the polar vortex also has long-term variations, exhibiting a persistent shift toward Northern Siberia and away from North America in February over the period 1979–2015 ( [[#Zhang--2016|Zhang et al., 2016]] ; J. [[#Zhang--2018|]] [[#Zhang--2018|]] [[#Zhang--2018|]] [[#Zhang--2018|Zhang et al., 2018]] ). Multiple measures show similar location changes ( [[#Seviour--2017|Seviour, 2017]] ).

Sudden stratospheric warming (SSW), a phenomenon of rapid stratospheric air temperature increases (sometimes by more than 50°C in 1–2 days), is tightly associated with the reversal of upper stratospheric zonal winds, and a resulting collapse or substantial weakening of the stratospheric polar vortex ( [[#Butler--2015|Butler et al., 2015]] ; [[#Butler--2018|Butler and Gerber, 2018]] ) and on average occurs approximately 6 times per decade in the NH winter ( [[#Charlton--2007|Charlton et al., 2007]] ; [[#Butler--2015|Butler et al., 2015]] ). The SSW record from all modern reanalyses is very consistent. There is a higher occurrence of major midwinter SSWs in the 1980s and 2000s with no SSW events during 1990–1997 ( [[#Reichler--2012|Reichler et al., 2012]] ; [[#Butler--2015|Butler et al., 2015]] ). An assessment of multi-decadal variability and change in SSW events is sensitive to both chosen metric and methods ( [[#Palmeiro--2015|Palmeiro et al., 2015]] ). Due to the lack of assimilation of upper air data, the centennial-scale reanalyses do not capture SSW events, even for the most recent decades ( [[#Butler--2015|Butler et al., 2015]] , 2017) and hence cannot inform on earlier behaviour. There has been considerably less study of trends in the SH stratosphere polar vortex strength despite the interest in the ozone hole and the potential impact of the SH stratosphere polar vortex strength on it. The occurrence of SSW events in the SH is not as frequent as in the NH, with only 3 documented events in the last 40 years ( [[#Shen--2020|Shen et al., 2020]] ).

In summary, it is ''likely'' that the northern lower stratospheric polar vortex has weakened since the 1980s in midwinter, and its location has shifted more frequently toward the Eurasian continent. The short record and substantial decadal variability yields ''low confidence'' in any trends in the occurrence of SSW events in the NH winter and such events in the SH are rare.

<div id="cross-chapter-box-2.3" class="h2-container box-container"></div>

'''Cross-Chapter Box 2.3 | New Estimates of Global Warming to Date, and Key Implications'''

<div id="h2-15-siblings" class="h2-siblings"></div>

'''Contributing Authors:''' Peter W. Thorne (Ireland/United Kingdom), Blair Trewin (Australia), Richard P. Allan (United Kingdom), Richard Betts (United Kingdom), Lea Beusch (Switzerland), Chris Fairall (United States of America), Piers Forster (United Kingdom), Baylor Fox-Kemper (United States of America), Jan S. Fuglestvedt (Norway), John C. Fyfe (Canada), Nathan P. Gillett (Canada), Ed Hawkins (United Kingdom), Christopher Jones (United Kingdom), Elizabeth Kent (United Kingdom), Svitlana Krakovska (Ukraine), Elmar Kriegler (Germany), Jochem Marotzke (Germany), H. Damon Matthews (Canada), Thorsten Mauritsen (Germany/Denmark), Anna Pirani (Italy), Joeri Rogelj (United Kingdom, Austria/Belgium), Steven K. Rose (United States of America), Bjørn H. Samset (Norway), Sonia I. Seneviratne (Switzerland), Claudia Tebaldi (United States of America), Andrew Turner (United Kingdom), Russell S. Vose (United States of America), Rachel Warren (United Kingdom)

This Cross-Chapter Box presents the AR6 WGI assessment of observed global warming and describes improvements and updates since AR5 and subsequent Special Reports. The revised estimates result from: the availability of new and revised observational datasets; the occurrence of recent record warm years; and the evaluation of the two primary metrics used to estimate global warming in past IPCC reports: ‘Global mean surface temperature’ (GMST) and ‘Global surface air temperature’ (GSAT). Implications for threshold crossing times, remaining carbon budgets and impacts assessments across AR6 WGs are discussed.

Cross-Chapter Box 2.3

'''Dataset innovations'''

Since AR5, all major datasets used for assessing observed temperature change based upon GMST have been updated and improved ( [[#2.3.1.1.3|Section 2.3.1.1.3]] ). A number of new products have also become available, including new datasets (e.g., Berkeley Earth, [[#Rohde--2020|Rohde and Hausfather, 2020]] ) and new interpolations based on existing datasets (e.g., [[#Cowtan--2014|Cowtan and Way, 2014]] and [[#Kadow--2020|Kadow et al., 2020]] ). These various estimates are not fully independent.

Improvements in global temperature datasets since AR5 have addressed two major systematic issues. First, new SST datasets ( [[#Huang--2017|Huang et al., 2017]] ; [[#Kennedy--2019|Kennedy et al., 2019]] ) address deficiencies previously identified in AR5 relating to the shift from predominantly ship-based to buoy-based measurements; these improvements result in larger warming trends, particularly in recent decades. Second, all datasets now employ interpolation to improve spatial coverage. This is particularly important in the Arctic, which has warmed faster than the rest of the globe in recent decades ( [[IPCC:Wg1:Chapter:Atlas|Atlas]] 5.9.2.2); under-sampling of the Arctic leads to a cool bias in recent decades ( [[#Simmons--2017|Simmons et al., 2017]] ; [[#Benestad--2019|Benestad et al., 2019]] ). Some datasets are now spatially complete ( [[#Cowtan--2014|Cowtan and Way, 2014]] ; [[#Kadow--2020|Kadow et al., 2020]] ) while others have expanded spatial coverage ( [[#Lenssen--2019|Lenssen et al., 2019]] ; [[#Rohde--2020|Rohde and Hausfather, 2020]] ; [[#Morice--2021|Morice et al., 2021]] ; [[#Vose--2021|Vose et al., 2021]] ). Several interpolation methods have been benchmarked against test cases (e.g., [[#Lenssen--2019|Lenssen et al., 2019]] ), and comparisons with reanalyses further confirm the value of such interpolation ( [[#Simmons--2017|Simmons et al., 2017]] ). It is ''extremely likely'' that interpolation produces an improved estimate of the changes in GMST compared to ignoring data-void regions.

Overall, dataset innovations and the availability of new datasets have led to an assessment of increased GMST change relative to the directly equivalent estimates reported in AR5 (Cross-Chapter Box 2.3, Table 1 and Figure 1).

'''Effects of warming since AR5 and choice of metrics of global mean temperature change'''

Each of the six years from 2015 to 2020 has ''likely'' been warmer than any prior year in the instrumental record. GMST for the decade 2011–2020 has been 0.19 [0.16 to 0.22] °C warmer than 2003–2012, the most recent decade used in AR5 (Cross-Chapter Box 2.3, Figure 1). A linear trend has become a poorer representation of observed change over time since most of the sustained warming has occurred after the 1970s (Cross-Chapter Box 2.3, Figure 1) and all values since 2012 are at least 0.2°C above a linear trendline for 1850–2020. For this reason, the primary method used to assess observed warming in this report is the change in temperature from 1850–1900 to the most recent decade (2011–2020) or the recent past (1995–2014), replacing the trend-based methods used in AR5 and earlier assessments. The effect of this change from trend-based to change-based metrics is currently relatively minor at –0.03°C (&lt;5%) for the most recent decade, but this may not remain the case in future ( ''high confidence'' ).

<div id="_idContainer032" class="Basic-Text-Frame"></div>

[[File:0eb5647470256cf45e0ca85e9ee7fe91 IPCC_AR6_WGI_CCBox_2_3_Figure_1.png]]

'''Cross-chapter Box 2.3, Figure 1'''

'''|'''

'''Changes in assessed historical surface temperature changes since AR5. (a)''' Summary of the impact of various steps from AR5 assessment warming-to-date number for 1880–2012 using a linear trend fit to the AR6 assessment based upon the difference between 1850–1900 and 2011–2020. Whiskers provide 90% ( ''very likely'' ) ranges. AR6 assessment in addition denotes additional warming since the period around 1750 (Cross-Chapter Box 1.2). '''(b)''' Time series of the average of assessed AR5 series (orange, faint prior to 1880 when only HadCRUT4 was available) and AR6 assessed series (blue) and their differences (offset) including an illustration of the two trend fitting metrics used in AR5 and AR6. Further details on data sources and processing are available in the chapter data table (Table 2.SM.1).

'''Observed changes in global mean temperature since the pre-industrial era'''

AR5 used 1850–1900 as an approximate pre-industrial baseline for global temperature change, whilst using an earlier pre-industrial baseline of 1750 for radiative forcings. Cross-Chapter Box 1.2 assesses that there was an observed GMST change from the period around 1750 to 1850–1900 of around 0.1°C ( ''likely'' range –0.1 to +0.3°C, ''medium confidence'' ). This additional global temperature change before 1850–1900 is not included when making AR6 assessments on global warming to date, global temperature threshold crossing times, or remaining carbon budgets to ensure consistency with previous ARs.

'''Addressing the non-equivalence of GMST and GSAT'''

GMST is a combination of land surface air temperatures (LSAT) and SSTs, whereas GSAT is a combination of LSAT and marine air temperatures (MATs). Although GMST and GSAT are closely related, the two measures are physically distinct. The implications have become more apparent since AR5 ( [[#Merchant--2013|Merchant et al., 2013]] ; [[#Cowtan--2015|Cowtan et al., 2015]] ; [[#Simmons--2017|Simmons et al., 2017]] ; [[#IPCC--2018|IPCC, 2018]] (SR1.5); [[#Richardson--2018|Richardson et al., 2018]] ), and it has been shown ( [[#Rubino--2020|Rubino et al., 2020]] ) that MAT and SST can show distinct multi-decadal-scale trends and patterns of interannual variability. Although SR1.5 used GMST for observational-based and GSAT for model-based headline warming statements, they noted the importance of the difference for their assessment (SR1.5 [[IPCC:Wg1:Chapter:Chapter-1#1.2.1.1|Section 1.2.1.1]] ). The SR1.5 used information from CMIP5 models to estimate a GSAT equivalent from observation-based GMST for certain applications such as remaining carbon budgets. The following subsections assess available lines of evidence related to the equivalence between GMST and GSAT.

''Physical understanding''

A well-understood physical constraint on the vertical gradient between the air and sea surface temperature is that it is approximately proportional to the turbulent sensible heat flux in the atmospheric surface layer ( [[#Chor--2020|Chor et al., 2020]] ). Similarly, the latent heat flux scales with the vertical humidity gradient and, in the global mean and in most oceanic regions, the latent heat flux is substantially larger than the sensible heat flux (Sections 7.2.1 and 9.2.1.3). If GSAT were to warm faster than GMST, the sensible surface heat flux would respond so as to reduce this difference. However, it is the sum of the sensible, latent, and radiative heat fluxes that controls GMST, so the sensible heat flux effect cannot be considered in isolation. Attempts to further constrain the combination of fluxes (e.g., [[#Lorenz--2010|Lorenz et al., 2010]] ; [[#Siler--2019|Siler et al., 2019]] ) rely on parameterizations or output from Earth system models (ESMs) or reanalyses and so are not considered independent. Apart from the above global considerations, regional and seasonal effects such as changes to the frequency and intensity of storms, sea state, cloudiness, sea ice cover, vegetation and land use may all affect the GSAT to GMST difference, either directly or by altering the relationships between gradients and energy fluxes. These changing energy flux relationships are monitored through observing the stratification of the upper ocean (Section 9.2.1.3) and the response of upper ocean processes (Cross-Chapter Box 5.3) in ESMs and reanalyses, but such monitoring tasks rival the observational challenge of directly observing SSTs and 2 m air temperature under a wide range of conditions. In summary, because of the lack of physical constraints and the complexity of processes driving changes in the GSAT to GMST temperature differences, there is no simple explanation based on physical grounds alone for how this difference responds to climate change.

''Direct observational evidence''

There is currently no regularly updated, entirely observation-based dataset for GSAT. The best available observations of near-surface air temperature over ocean are datasets of night-time marine air temperature (NMAT; e.g., [[#Cornes--2020|Cornes et al., 2020]] ; [[#Junod--2020|Junod and Christy, 2020]] ), though spatial coverage is less extensive than for SST. Night-time measurements are used to avoid potential biases from daytime heating of ship superstructures. [[#Kennedy--2019|Kennedy et al. (2019)]] show little difference between HadNMAT2 and HadSST4 between 1920 and 1990, but a warming of SST relative to NMAT manifesting as a step change of 0.05°C–0.10°C in the early 1990s, which may reflect an actual change, the impact of increasingly divergent spatial coverage between SST and MAT measurements, or unresolved structural uncertainties in one or both datasets. This leads to NMAT warming around 10% more slowly than SST over the last century. In contrast, [[#Junod--2020|Junod and Christy (2020)]] find NMAT trends which are 8–17% larger than those for SST in the ERSSTv4 and HadISST datasets for the period 1900 to 2010, but 11–15% smaller than the SST trends for the same datasets from 1979 to 2010. However, ERSSTv4 uses NMAT data as a basis for homogeneity adjustment so is not fully independent. [[#Kent--2021|Kent and Kennedy (2021)]] note sensitivity to methodological choices in comparisons but find that NMAT is warming more slowly than SST products over most periods considered. [[#Rubino--2020|Rubino et al. (2020)]] exploit tropical Pacific moored buoy arrays, available since the early 1980s, and find differences in NMAT and SST anomalies, which are sensitive to the choice of period and show spatio-temporal ENSO-related (Annex IV) signals in the differences.

Overall, with ''medium evidence'' and ''low agreement'' , available observational products suggest that NMAT is warming less than SST by up to 15%. Given that these ocean observations cover roughly two thirds of the globe, this implies that GMST is warming up to at most 10% faster than GSAT. Substantial uncertainty remains and the effect is highly sensitive to the choice of both time period and choice of NMAT and SST observational products to compare. Observed NMAT warming faster than observed SST cannot be precluded.

''CMIP model-based evidence''

CMIP historical simulations and projections agree that GSAT increases faster than GMST, the reverse of what is indicated by many marine observations. Several studies approximate the approach used to derive GMST from observations by blending SST over open ocean and SAT over land and sea ice from model output ( [[#Cowtan--2015|Cowtan et al., 2015]] ; [[#Richardson--2018|Richardson et al., 2018]] ; [[#Beusch--2020|Beusch et al., 2020]] ; [[#Gillett--2021|Gillett et al., 2021]] ). Cowtan et al. found that trends in GSAT are of the order of 9% larger than for GMST in CMIP5, based on data from 1850–2100 (historical + RCP8.5), if anomalies are blended and sea ice is allowed to vary over time ( [[#Cowtan--2015|Cowtan et al., 2015]] ). Broadly consistent numbers are found for both CMIP5 and CMIP6, across a range of SSP and RCP scenarios and time periods ( [[#Richardson--2018|Richardson et al., 2018]] ; [[#Beusch--2020|Beusch et al., 2020]] ; [[#Gillett--2021|Gillett et al., 2021]] ). Blending monthly anomalies and allowing sea ice to vary, the change in GSAT for 2010–2019 relative to 1850–1900 is 2–8% larger than spatially-complete GMST in CMIP6 historical and SSP2-4.5 simulations ( [[#Gillett--2021|Gillett et al., 2021]] ), and 6–12% larger in CMIP5 historical and RCP2.6 and 8.5 simulations for 2007–2016 relative to 1861–1880 ( [[#Richardson--2018|Richardson et al., 2018]] ). However, a true like-for-like comparison to observational products is challenging because methodological choices have a large impact on the relationship between modelled GMST and GSAT and none of these studies fully reproduces the methods used to derive estimates of GMST in recent observational datasets, which use various ways to infill areas lacking in situ observations ( [[#Jones--2020|Jones, 2020]] ).

Marine boundary layer behaviour and parameterizations in all CMIP models are based upon Monin-Obukhov similarity theory (e.g., [[#Businger--1971|Businger et al., 1971]] ), which informs assumptions around gradients in the near-surface boundary layer dependent upon temperature, wind speed and humidity. This leaves open the possibility of a common model bias, while [[#Druzhinin--2019|Druzhinin et al. (2019)]] also point to departures of temperature profiles from theoretical predictions under certain conditions. There remain inadequacies in understanding and modelling of key processes ( [[#Edwards--2020|Edwards et al., 2020]] ), and biases in the representation of the absolute SST-MAT difference have been identified in climate models and reanalyses ( [[#Găinuşă-Bogdan--2015|Găinuşă-Bogdan et al., 2015]] ; [[#Zhou--2020|Zhou et al., 2020]] ).

''Reanalysis-based evidence''

[[#Simmons--2017|Simmons et al. (2017)]] found that in JRA-55 and ERA-Interim (following an adjustment to account for an apparent discontinuity), GSAT increased 2–4% faster than GMST over the period 1979–2016. In atmospheric reanalyses, SST is given as a lower boundary condition from an observed globally interpolated product (such as HadISST; [[#Rayner--2003|Rayner et al., 2003]] ) whereas the air temperature is reliant upon model parameterizations and assimilated observations that do not include MAT observations ( [[#Simmons--2017|Simmons et al., 2017]] ), thereby limiting their capability to constrain differences in GMST and GSAT trends. Furthermore, it is unclear what the lack of dynamic coupling at the ocean-atmosphere interface might imply for the representativeness of reanalysis-based estimates.

''Representation of surface temperatures in sea ice regions''

There is a significant issue in areas where sea ice melts or grows, where the quantity used in observational-based GMST estimates switches between air temperature and sea surface temperature. This primarily affects analyses combining SAT anomalies over land and ice with SST anomalies over ocean. In areas where sea ice has recently melted, the climatological value changes from an air-temperature based estimate to an SST estimate based upon the freezing point of seawater (–1.8°C). This switch in climatology to, in general, a warmer climatology, leads to a bias towards reduced warming in anomalies compared with analyses based on absolute temperatures. [[#Richardson--2018|Richardson et al. (2018)]] found this underestimation to amount to approximately 3% of observed warming in historical model simulations. Given the projected future sea ice losses, the effect will grow in future ( ''low confidence'' ), with potential effects of the order of 0.1°C in the second half of the 21st century under high warming scenarios, although with some uncertainty arising from the large spread of sea ice loss in model projections ( [[#Tokarska--2019|Tokarska et al., 2019]] ).

'''Cross Chapter Box 2.3, Table'''

'''1 |'''

'''Summary of key observationally based global warming estimates (in °C) to various reference periods in the present report and selected prior reports (AR5 WGI and SR1.5) and their principal applications (see [[IPCC:Wg1:Chapter:Chapter-1#1.4.1|Section 1.4.1]] for further information on reference periods).''' Further details on data sources and processing are available in the chapter data table (Table 2.SM.1).
{| class="wikitable"
|-
| '''Reference Period'''

| '''AR6 GMST''' (° '''C)'''

| '''AR6 GSA''' '''T''' <sup>a</sup> '''(''' ° '''C)'''

| '''AR5 and/or''' SR1.5 '''(''' italics ''') – Only Where Reported''' (° '''C)'''

| '''Principal Use of This Period in this Report and Previous Reports'''

|-
| 1850–1900 to 2011–2020

| 1.09 [0.95 to 1.20]

| 1.09 [0.91 to 1.23]

|
| Warming to present in AR6 WGI

|-
| 1850–1900 to 2010–2019

| 1.06 [0.92 to 1.17]

| 1.06 [0.88 to 1.21]

|
| Attributable warming assessment period in AR6 WGI

|-
| 1850–1900 to 2006–2019

| 1.03 [0.89 to 1.14]

| 1.03 [0.86 to 1.18]

|
| AR6 WGI warming estimate as a line of evidence for energy budget constraints to estimate ECS and TCR

|-
| 1850–1900 to 2006–2015

| 0.94 [0.79 to 1.04]

| 0.94 [0.76 to 1.08]

| ''0.87 [0.75 to 0.99] – GMST''

''0.97 [0.85 to 1.09] – GSA'' ''T'' <sup>b</sup>

| Warming to date in SR1.5

|-
| 1850–1900 to 2003–2012

| 0.90 [0.74 to 1.00]

| 0.90 [0.72 to 1.03]

| 0.78 [0.72 to 0.85]

| Warming to date in AR5 WGI

|-
| 1850–1900 to 2001–2020

| 0.99 [0.84 to 1.10]

| 0.99 [0.81to 1.14]

|
| Warming to first two decades of 21st century

|-
| 1850–1900 to 1995–2014

| 0.85 [0.69 to 0.95]

| 0.85 [0.67 to 0.98]

|
| Warming to recent past in AR6 WGI

|-
| 1850–1900 to 1986–2005

| 0.69 [0.54 to 0.79]

| 0.69 [0.52 to 0.82]

| 0.61 [0.55 to 0.67] <sup>c</sup>

| Warming to recent past in AR5 WGI. This difference is used to report in this box the implications of the AR6 historical global surface temperature assessment in a way that is directly comparable to the AR5 estimate.

|-
| 1850–1900 to 1961–1990

| 0.36 [0.23 to 0.44]

| 0.36 [0.22 to 0.45]

|
| Warming to reference period recommended by WMO for national-level data sets used for climate change assessment (included in the AR6 WGI Atlas)

|-
| 1880–2012 OLS trend

| 0.92 [0.68 to 1.17]

|
| 0.85 [0.65 to 1.06]

| Warming trend to date in AR5 WGI Summary for Policymakers and AR5 Synthesis Report

|}

<sup>a</sup> As the uncertainty in the relationship between GMST and GSAT changes is independent of the uncertainty in the assessed change in GMST, these uncertainties are combined in quadrature.

<sup>b</sup> The SR1.5 derived a GSAT estimate by taking the CMIP5 ensemble mean GSAT change of 0.99°C, sub-sampling to HadCRUTv4.6, noting the offset in trends (0.84°C HadCRUT4 observed GMST vs. 0.86°C modelled GMST) and adjusting by this to arrive at an estimate of 0.97°C change in GSAT. The ''likely'' uncertainty range of ±0.12°C was not further adjusted.

<sup>c</sup> Note that the AR5 approach for the change from 1850–1900 to both 1986–2005 and 2003–2012 was based upon one dataset (HadCRUT4) and its parametric uncertainty estimates are known to underestimate the true uncertainty.

''Summary of lines of evidence''

GMST and GSAT are physically distinct. There is ''high confidence'' that long-term changes in GMST and GSAT differ by at most 10% in either direction. However, conflicting lines of evidence from models and direct observations combined with limitations in theoretical understanding lead to ''low confidence'' in the sign of any difference in long-term trends. The ''very likely'' range of estimated historical GMST warming is combined with the assessed ± 10% uncertainty in the relationship between GMST and GSAT changes to infer a GSAT equivalent, accounting for any possible real-world physical difference. Improvements in understanding may yield a robust basis to apply a scaling-factor to account for the difference in future assessments.

'''Mapping between AR5 and AR6 Assessments'''

The AR5 assessed estimate for historical warming between 1850–1900 and 1986–2005 is 0.61 [0.55 to 0.67] °C. The equivalent in AR6 is 0.69 [0.54 to 0.79] °C, and the 0.08 [-0.01 to 0.12] °C difference is an estimate of the contribution of changes in observational understanding alone (Cross-Chapter Box 2.3, Table 1). The exact value of this contribution depends upon the metric being compared (GMST/GSAT, the method used to calculate a trend or change between two periods, the exact reference period used), with the best estimates (with the exception of the SR1.5 GSAT estimate) falling between 0.07°C and 0.12°C. The choice of 1850–1900 to 1986–2005 as the basis is due to the widespread use of this period across AR5 and SR1.5 in several contexts. The AR6-assessed GMST warming between 1850–1900 and 2011–2020 is 1.09 [0.95 to 1.20] °C. An AR5-equivalent assessment using this estimated difference in observational understanding is thus 1.01 [0.94 to 1.08] °C. These updates and improvements in observational datasets affect other quantities that derive from the assessment of GSAT warming, including estimates of remaining carbon budgets and estimates of crossing times of 1.5°C and 2°C of global warming (see Cross Chapter Box 2.3, Table 1).

'''Updates to estimated Global Warming Level (GWL) crossing times'''

The updated estimate of historical warming is one contribution to the revised time of projected crossing of the threshold of 1.5°C global warming in comparison with SR1.5, but is not the only reason for this update. The AR6 assessment of future change in GSAT (Table 4.5) results in the following threshold-crossing times, based on 20-year moving averages. The threshold-crossing time is defined as the midpoint of the first 20-year period during which the average GSAT exceeds the threshold. During the near term (2021–2040), a 1.5°C GSAT increase relative to the average over the period 1850–1900 is ''very likely'' to occur in scenario SSP5-8.5, ''likely'' to occur in scenarios SSP2-4.5 and SSP3-7.0, and ''more likely than not'' to occur in scenarios SSP1-1.9 and SSP1-2.6. In all scenarios assessed here except SSP5-8.5, the central estimate of crossing the 1.5°C global warming level lies in the early 2030s. This is in the early part of the ''likely'' range (2030–2052) assessed in SR1.5, which assumed continuation of the then-reported warming rate; this estimated rate has been confirmed in AR6 ( [[IPCC:Wg1:Chapter:Chapter-3#3.3.1|Section 3.3.1]] ). Roughly half of this difference arises from the higher diagnosed historical warming in AR6. The other half arises because, for central estimates of climate sensitivity, most scenarios show stronger warming over the near term than was assessed as ‘current’ in SR1.5 ( ''medium confidence'' ). When considering scenarios similar to SSP1-1.9 instead of linear extrapolation, the SR1.5 estimate of when 1.5°C global warming is crossed is close to the central estimate reported here (SR1.5, Table 2.SM.12).

'''Implications for assessment of emissions scenarios and remaining carbon budgets'''

To estimate the global warming implications of emissions scenarios, AR5 and SR1.5 combined estimates of observed GMST changes from 1850–1900 to 1986–2005 (Cross-Chapter Box 2.3, Table 1) with GSAT projections of subsequent warming. AR6 undertakes three changes to this approach. First, the AR6 assessment of improved observational records is used. Second, the recent past baseline period is updated from 1986–2005 to 1995–2014, and, third, historical estimates are expressed in GSAT instead of GMST for consistency of historical estimates with future projections. The updated estimates of warming to date in AR6 lead to higher estimates of future warming, all else being equal. The temperature classification of emissions scenarios in the WGIII report adopts the definition of temperature classes as introduced in SR1.5, and assigns emissions scenarios to these classes based on their AR6 assessed GSAT outcomes (Cross-Chapter Box 7.1; WGIII Annex C.II.2.4).

In both AR5 and SR1.5, remaining carbon budgets were expressed as a function of GSAT warming, while also highlighting the implications of using historical warming estimates expressed in GMST. The AR5 reported total carbon budgets for GSAT warming relative to 1861–1880. The AR5 Synthesis Report (SYR) also includes remaining carbon budget estimates based on AR5 WGIII scenario projections that use the method for AR5 scenario projections described above. The SR1.5 integrated several methodological advancements to estimate remaining carbon budgets and reported budgets for additional GSAT warming since the 2006–2015 period, estimating, following the application of an adjustment ( [[#Richardson--2016|Richardson et al., 2016]] , Table 1.1, SR1.5) to GMST, that 0.97°C (± 0.12°C) of GSAT warming occurred historically between 1850–1900 and 2006–2015. The AR6 assessment, above, leads to an estimate of 0.94°C of warming between 1850–1900 and 2006–2015. All other factors considered equal, the AR6 estimate thus implies that 0.03°C more warming is considered for remaining carbon budgets compared to SR1.5. Combining this 0.03°C value with the SR1.5 transient climate response to cumulative emissions of CO <sub>2</sub> (TCRE) translates into remaining carbon budgets about 70 [40–140] GtCO <sub>2</sub> larger compared to SR1.5 on a like-for-like basis. Meanwhile, on the same like-for-like basis, updates to historical observational products would reduce remaining carbon budgets reported in AR5 SYR based on WGIII scenario projections by about 180 [120 to 370] GtCO <sub>2</sub> . Box 5.2 provides a further overview of updates to estimates of the remaining carbon budget since AR5.

'''Implications for assessment of impacts and adaptation'''

The assessment of global warming to date now being larger than previously assessed has no consequence on the assessment of past climate impacts, nor does it generally imply that projected climate impacts are now expected to occur earlier. The implications are mainly that the level of warming associated with a particular impact has been revised. This has very limited practical implications for the assessment of the benefits of limiting global warming to specific levels, as well as for the urgency of adaptation action. For example, impacts that occurred in the period 1986–2005 were previously associated with a GMST increase of 0.61°C relative to 1850–1900, relative to AR5 estimates. These impacts are now instead associated with a GMST increase of 0.69°C, relative to the assessment in this Report. The impacts themselves have not changed. Similarly, the impacts previously associated with a GMST or GSAT increase of 1.5°C will now generally be associated with a slightly different global warming level. This is because projections of future warming and its impacts relative to 1850–1900 are normally made by adding projected warming from a recent past baseline to an estimate of the observed warming from 1850–1900, as in AR5 and SR1.5.

Most of the previously projected impacts and risks associated with global warming of 1.5°C have therefore not changed and are still associated with the same level of future warming (0.89°C) relative to 1986–2005. With this warming now estimated as 0.08°C larger than in AR5, the future impacts previously associated with 1.5°C warming are now associated with 1.58°C warming. Similarly, the impacts now associated with 1.5°C warming would have previously been associated with 1.42°C warming. There are exceptions where impacts studies have used a baseline earlier than 1986–2005 (e.g., [[#King--2017|King et al., 2017]] ), for which the new estimate of the historical warming would mean an earlier occurrence of the projected impacts. However, even in these cases, the ostensible difference in impacts associated with a 0.08°C difference in global mean temperature will be small in comparison with the uncertainties. There are also substantial uncertainties in regional climate changes and the magnitude of climate impact-drivers projected to occur with global warming of 1.5°C ( [[#Betts--2018|Betts et al., 2018]] ; [[#Seneviratne--2018|Seneviratne et al., 2018]] ). Furthermore, the time of reaching global warming of 1.5°C is subject to uncertainties of approximately ±10 years associated with uncertainties in climate sensitivity, and ±3 to 4 years associated with the different SSP forcing scenarios ( [[IPCC:Wg1:Chapter:Chapter-4#4.3.4|Section 4.3.4]] , Table 4.5, and see discussion above).

There is therefore ''high confidence'' that assessment of the magnitude and timing of impacts-related climate quantities at 1.5°C is not substantially affected by the revised estimate of historical global warming. The assessment of the implications of limiting global warming to 1.5°C compared to 2°C will also remain broadly unchanged by the updated estimate of historical warming, as this depends on the relative impacts rather than the absolute impacts at any specific definition of global temperature anomaly ( ''high confidence'' ).

<div id="2.3.2" class="h2-container"></div>

<span id="cryosphere"></span>