Editing IPCC:AR6/WGI/TS (section)

=== TS.1.2 Progress in Climate Science ===

<div id="h2-7-siblings" class="h2-siblings"></div>

<div id="TS.1.2.1" class="h3-container"></div>

<span id="ts.1.2.1-observation-based-products-and-their-assessments"></span>
==== TS.1.2.1 Observation-based Products and their Assessments ====

<div id="h3-1-siblings" class="h3-siblings"></div>

'''Observational capabilities have continued to improve and expand overall since AR5, enabling improved consistency between independent estimates of climate drivers, the combined climate feedbacks, and the observed energy and sea level increase. Satellite climate records and improved reanalyses are used as an additional line of evidence for assessing changes at the global and regional scales. However, there have also been reductions in some observational data coverage or continuity and limited access to data resulting from data policy issues. Natural archives of past climate, such as tropical glaciers, have also been subject to losses (in part due to anthropogenic climate change). Links to chapters 1.5.1, 1.5.2, 10.2.2'''
Earth system observations are an essential driver of progress in our understanding of climate change. Overall, capabilities to observe the physical climate system have continued to improve and expand. Improvements are particularly evident in ocean observing networks and remote-sensing systems. Records from several recently instigated satellite measurement techniques are now long enough to be relevant for climate assessments. For example, globally distributed, high-vertical-resolution profiles of temperature and humidity in the upper troposphere and stratosphere can be obtained from the early 2000s using global navigation satellite systems, leading to updated estimates of recent atmospheric warming. Improved measurements of ocean heat content, warming of the land surface, ice-sheet mass loss and sea level changes allow a better closure of the global energy and sea level budgets relative to AR5. For surface and balloon-based networks, apparent regional data reductions result from a combination of data policy issues, data curation/provision challenges, and real cessation of observations, and are to an extent counter-balanced by improvements elsewhere. Limited observational records of extreme events and spatial data gaps currently limit the assessment of some observed regional climate change. Links to chapters 1.5.1, 2.3.2, 7.2.2, Box 7.2, Cross-Chapter Box 9.1, 9.6.1, 10.2.2, 10.6, 11.2, 12.4

New paleoclimate reconstructions from natural archives have enabled more robust reconstructions of the spatial and temporal patterns of past climate changes over multiple time scales (Box TS.2). However, paleoclimate archives, such as tropical glaciers and modern natural archives used for calibration (e.g., corals and trees), are rapidly disappearing owing to a host of pressures, including increasing temperatures ( ''high confidence'' ). Substantial quantities of past instrumental observations of weather and other climate variables, over both land and ocean, which could fill gaps in existing datasets, remain un-digitized or inaccessible. These include measurements of temperature (air and sea surface), rainfall, surface pressure, wind strength and direction, sunshine amount and many other variables dating back into the 19th century. Links to chapters 1.5.1

Reanalyses combine observations and models (e.g., a numerical weather prediction model) using data assimilation techniques to provide a spatially complete, dynamically consistent estimate of multiple variables describing the evolving climate state. Since AR5, new reanalyses have been developed for the atmosphere and the ocean with various combinations of increased resolution, extended records, more consistent data assimilation and larger availability of uncertainty estimates. Limitations remain, for example, in how reanalyses represent global-scale changes to the water cycle. Regional reanalyses use high-resolution, limited-area models constrained by regional observations and with boundary conditions from global reanalyses. There is ''high confidence'' that regional reanalyses better represent the frequencies of extremes and variability in precipitation, surface air temperature and surface wind than global reanalyses and provide estimates that are more consistent with independent observations than dynamical downscaling approaches. Links to chapters 1.5.2, 10.2.1.2, Annex I

<div id="TS.1.2.2" class="h3-container"></div>

<span id="ts.1.2.2-climate-model-performance"></span>
==== TS.1.2.2 Climate Model Performance ====

<div id="h3-2-siblings" class="h3-siblings"></div>

'''This report assesses results from climate models participating in the Coupled Model Intercomparison Project Phase 6 (CMIP6) of the World Climate Research Programme. These models include new and better representation of physical, chemical and biological processes, as well as higher resolution, compared to climate models considered in previous IPCC Assessment Reports. This has improved the simulation of the recent mean state of most large-scale indicators of climate change and many other aspects across the climate system. Some differences from observations remain, for example in regional precipitation patterns. Projections of the increase in global surface temperature, the pattern of warming, and global mean sea level rise from previous IPCC Assessment Reports and other studies are broadly consistent with subsequent observations, especially when accounting for the difference in radiative forcing scenarios used for making projections and the radiative forcings that actually occurred.''' '''The CMIP6 historical simulations assessed in this report have an ensemble mean global surface temperature change within 0.2°C of the observations over most of the historical period, and observed warming is within the ''very likely'' range of the CMIP6 ensemble. However, some CMIP6 models simulate a warming that is either above or below the assessed ''very likely'' range of observed warming. The information about how well models simulate past warming, as well as other insights from observations and theory, are used to assess projections of global warming (see Cross-Section Box TS.1). Increasing horizontal resolution in global climate models improves the representation of small-scale features and the statistics of daily precipitation ( ''high confidence'' ). Earth system models, which include additional biogeochemical feedbacks, often perform as well as their lower-complexity global climate model counterparts, which do not account for these additional feedbacks ( ''medium confidence'' ). Links to chapters 1.3.6, 1.5.3, 3.1, 3.5.1, 3.8.2, 4.3.1, 4.3.4, 7.5, 8.5.1, 9.6.3.1'''
Climate model simulations coordinated and collected as part of the World Climate Research Programme’s Coupled Model Intercomparison Project Phase 6 (CMIP6), complemented by a range of results from the previous phase (CMIP5), constitute a key line of evidence supporting this Report. The latest generation of CMIP6 models have an improved representation of physical processes relative to previous generations, and a wider range of Earth system models now represent biogeochemical cycles. Higher-resolution models that better capture smaller-scale processes are also increasingly becoming available for climate change research (Figure TS.2, Panels a and b). Results from coordinated regional climate modelling initiatives, such as the Coordinated Regional Climate Downscaling Experiment (CORDEX) complement and add value to the CMIP global models, particularly in complex topography zones, coastal areas and small islands, as well as for extremes. Links to chapters 1.5.3, 1.5.4, 2.8.2, FAQ 3.3, 6.2.2, 6.4, 6.4.5, 8.5.1, 10.3.3, Atlas.1.4

Projections of the increase in global surface temperature and the pattern of warming from previous IPCC Assessment Reports and other studies are broadly consistent with subsequent observations ( ''limited evidence, high agreement'' ), especially when accounting for the difference in radiative forcing scenarios used for making projections and the radiative forcings that actually occurred (Figure TS.3). The AR5 and SROCC projections of GMSL for the 2007–2018 period have been shown to be consistent with observed trends in GMSL and regional weighted mean tide gauges. Links to chapters 1.3.6, 9.6.3.1

For most large-scale indicators of climate change, the simulated recent mean climate from CMIP6 models underpinning this assessment have improved compared to the CMIP5 models used in AR5 ( ''high confidence'' ). This is evident from the performance of 18 simulated atmospheric and land large-scale indicators of climate change between the three generations of models (CMIP3, CMIP5, and CMIP6) when benchmarked against reanalysis and observational data (Figure TS.2, Panel c). Earth system models, characterized by additional biogeochemical feedbacks, often perform at least as well as related, more constrained, lower-complexity models lacking these feedbacks ( ''medium confidence'' ). Links to chapters 3.8.2, 10.3.3.3

The CMIP6 multi-model mean global surface temperature change from 1850–1900 to 2010–2019 is close to the best estimate of the observed warming. However, some CMIP6 models simulate a warming that is below or above the assessed ''very likely'' range. The CMIP6 models also reproduce surface temperature variations over the past millennium, including the cooling that follows periods of intense volcanism ( ''medium confidence'' ). For upper air temperature, an overestimation of the upper tropical troposphere warming by about 0.1°C per decade between 1979 and 2014 persists in most CMIP5 and CMIP6 models ( ''medium confidence'' ), whereas the differences between simulated and improved satellite-derived estimates of change in global mean temperature through the depth of the stratosphere have decreased. Links to chapters 3.3.1

Some CMIP6 models demonstrate an improvement in how clouds are represented. CMIP5 models commonly displayed a negative shortwave cloud radiative effect that was too weak in the present climate. These errors have been reduced, especially over the Southern Ocean, due to a more realistic simulation of supercooled liquid droplets with sufficient numbers and an associated increase in the cloud optical depth. Because a negative cloud optical depth feedback in response to surface warming results from ‘brightening’ of clouds via active phase change from ice to liquid cloud particles (increasing their shortwave cloud radiative effect), the extratropical cloud shortwave feedback in CMIP6 models tends to be less negative, leading to a better agreement with observational estimates ( ''medium confidence'' ). CMIP6 models generally represent more processes that drive aerosol–cloud interactions than the previous generation of climate models, but there is only ''medium confidence'' that those enhancements improve their fitness-for-purpose of simulating radiative forcing of aerosol–cloud interactions. Links to chapters 6.4, 7.4.2, FAQ 7.2

CMIP6 models still have deficiencies in simulating precipitation patterns, particularly in the tropical ocean. Increasing horizontal resolution in global climate models improves the representation of small-scale features and the statistics of daily precipitation ( ''high confidence'' ). There is ''high confidence'' that high-resolution global, regional and hydrological models provide a better representation of land surfaces, including topography, vegetation and land-use change, which can improve the accuracy of simulations of regional changes in the terrestrial water cycle. Links to chapters 3.3.2, 8.5.1, 10.3.3, 11.2.3

There is ''high confidence'' that climate models can reproduce the recent observed mean state and overall warming of temperature extremes globally and in most regions, although the magnitude of the trends may differ. There is ''high confidence'' in the ability of models to capture the large-scale spatial distribution of precipitation extremes over land. The overall performance of CMIP6 models in simulating the intensity and frequency of extreme precipitation is similar to that of CMIP5 models ( ''high confidence'' ). Links to chapters Cross-Chapter Box 3.2, 11.3.3, 11.4.3

The structure and magnitude of multi-model mean ocean temperature biases have not changed substantially between CMIP5 and CMIP6 ( ''medium confidence'' ). Since AR5, there is improved consistency between recent observed estimates and model simulations of changes in upper (&lt;700 m) ocean heat content. The mean zonal and overturning circulations of the Southern Ocean and the mean overturning circulation of the North Atlantic (AMOC) are broadly reproduced by CMIP5 and CMIP6 models. Links to chapters 3.5.1, 3.5.4, 9.2.3, 9.3.2, 9.4.2

CMIP6 models better simulate the sensitivity of Arctic sea ice area to anthropogenic CO <sub>2</sub> emissions, and thus better capture the time evolution of the satellite-observed Arctic sea ice loss ( ''high confidence'' ). The ability to model ice-sheet processes has improved substantially since AR5. As a consequence, there is ''medium confidence'' in the representation of key processes related to surface-mass balance and retreat of the grounding-line (the junction between a grounded ice sheet and an ice shelf, where the ice starts to float) in the absence of instabilities. However, there remains ''low confidence'' in simulations of ice-sheet instabilities, ice-shelf disintegration and basal melting owing to their high sensitivity to both uncertain oceanic forcing and uncertain boundary conditions and parameters. Links to chapters 1.5.3, 2.3.2, 3.4.1, 3.4.2, 3.8.2, 9.3.1, 9.3.2, 9.4.1, 9.4.2

CMIP6 models are able to reproduce most aspects of the spatial structure and variance of the El Niño–Southern Oscillation (ENSO) and Indian Ocean Basin and Dipole modes of variability ( ''medium confidence'' ). However, despite a slight improvement in CMIP6, some underlying processes are still poorly represented. Models reproduce observed spatial features and variance of the Southern Annular Mode (SAM) and Northern Annular Mode (NAM) very well ( ''high confidence'' ). The summertime SAM trend is well captured, with CMIP6 models outperforming CMIP5 models ( ''medium confidence'' ). By contrast, the cause of the NAM trend towards its positive phase is not well understood. In the Tropical Atlantic basin, which contains the Atlantic Zonal and Meridional modes, major biases in modelled mean state and variability remain. Model performance is limited in reproducing sea surface temperature anomalies for decadal modes of variability, despite improvements from CMIP5 to CMIP6 ( ''medium confidence'' ) (see also Section TS.1.4.2.2, Table TS.4). Links to chapters 3.7.3–3.7.7

Earth system models (ESMs) simulate globally averaged land carbon sinks within the range of observation-based estimates ( ''high confidence'' ), but global-scale agreement masks large regional disagreements. There is also ''high confidence'' that the ESMs simulate the weakening of the global net flux of CO <sub>2</sub> into the ocean during the 1990s, as well as the strengthening of the flux from 2000. Links to chapters 3.6

Two important quantities used to estimate how the climate system responds to changes in greenhouse gas (GHG) concentrations are the equilibrium climate sensitivity (ECS) and transient climate response (TCR <sup>[[#footnote-005|16]]</sup> ). The CMIP6 ensemble has broader ranges of ECS and TCR values than CMIP5 (see Section TS.3.2 for the assessed range). These higher sensitivity values can, in some models, be traced to changes in extratropical cloud feedbacks ( ''medium confidence'' ). To combine evidence from CMIP6 models and independent assessments of ECS and TCR, various emulators are used throughout the report. Emulators are a broad class of simple climate models or statistical methods that reproduce the behaviour of complex ESMs to represent key characteristics of the climate system, such as global surface temperature and sea level projections. The main application of emulators in AR6 is to extrapolate insights from ESMs and observational constraints to produce projections from a larger set of emissions scenarios, which is achieved due to their computational efficiency. These emulated projections are also used for scenario classification in WGIII. Links to chapters Box 4.1, 4.3.4, 7.4.2, 7.5.6, Cross-Chapter Box 7.1, FAQ 7.2

[[File:0fc6683b2510cbb49255c2fed4f543bf IPCC_AR6_WGI_TS_Figure_2.png]]

'''Figure TS.2 |''' '''Progress in climate models.''' ''The intent of this figure is to show present improvements in climate models in resolution, complexity and representation of key variables.'' '''(a)''' Evolution of model horizontal resolution and vertical levels (based on Figure 1.19). '''(b)''' Evolution of inclusion of processes and resolution from Coupled Model Intercomparison Project Phase 3 (CMIP3), Phase 5 (CMIP5) and Phase 6 (CMIP6; Annex II). '''(c)''' Centred pattern correlations between models and observations for the annual mean climatology over the period 1980–1999. Results are shown for individual CMIP3 (cyan), CMIP5 (blue) and CMIP6 (red) models (one ensemble member is used) as short lines, along with the corresponding ensemble averages (long lines). The correlations are shown between the models and the primary reference observational data set (from left to right: ERA5, GPCP-SG, CERES-EBAF, CERES-EBAF, CERES-EBAF, CERES-EBAF, JR-55, ERA5, ERA5, ERA5, ERA5, ERA5, ERA5, AIRS, ERA5, ESACCI-Soilmoisture, LAI3g, MTE). In addition, the correlation between the primary reference and additional observational data sets (from left to right: NCEP, GHCN, -, -, -, -, ERA5, HadISST, NCEP, NCEP, NCEP, NCEP, NCEP, NCEP, ERA5, NCEP, -, -, FLUXCOM) are shown (solid grey circles) if available. To ensure a fair comparison across a range of model resolutions, the pattern correlations are computed after regridding all datasets to a resolution of 4º in longitude and 5º in latitude. (Expanded from Figure 3.43; produced with ESMValTool version 2). Links to chapters Figure 3.43

<div id="TS.1.2.3" class="h3-container"></div>

<span id="ts.1.2.3-understanding-climate-variability-and-emerging-changes"></span>
==== TS.1.2.3 Understanding Climate Variability and Emerging Changes ====

<div id="h3-3-siblings" class="h3-siblings"></div>

'''Observed changes in climate are unequivocal at the global scale and are increasingly apparent on regional and local spatial scales. Both the rate of long-term change and the amplitude of year-to-year variations differ between regions and across climate variables, thus influencing when changes emerge or become apparent compared to natural variations (see Emergence in Core Concepts Box). The signal of temperature change has emerged more clearly in tropical regions, where year-to-year variations tend to be small over land, than in regions with greater warming but larger year-to-year variations ( ''high confidence'' ) (Figure TS.3). Long-term changes in other variables have emerged in many regions, such as for some weather and climate extremes and Arctic sea ice area. Links to chapters 1.4.2, Cross-Chapter Box 3.1, 9.3.1, 11.3.2, 12.5.2'''
Observational datasets have been extended and improved since AR5, providing stronger evidence that the climate is changing and allowing better estimates of natural climate variability on decadal time scales. There is ''very high confidence'' that the slower rate of global surface temperature change observed over 1998–2012 compared to 1951–2012 was temporary, and was, with ''high confidence'' , induced by internal variability (particularly Pacific Decadal Variability) and variations in solar irradiance and volcanic forcing that partly offset the anthropogenic warming over this period. Global ocean heat content continued to increase throughout this period, indicating continuous warming of the entire climate system ( ''very high confidence'' ). Hot extremes also continued to increase during this period over land ( ''high confidence'' ). Even in a continually warming climate, periods of reduced and increased trends in global surface temperature at decadal time scales will continue to occur in the 21st century ( ''very high confidence'' ). Links to chapters Cross-Chapter Box 3.1, 3.3.1, 3.5.1, 4.6.2, 11.3.2

Since AR5, the increased use of ‘large ensembles’, or multiple simulations with the same climate model but using different initial conditions, supports improved understanding of the relative roles of internal variability and forced change in the climate system. Simulations and understanding of modes of climate variability, including teleconnections, have improved since AR5 ( ''medium confidence'' ), and larger ensembles allow a better quantification of uncertainty in projections due to internal climate variability. Links to chapters 1.4.2, 1.5.3, 1.5.4, 4.2, 4.4.1, Box 4.1, 8.5.2, 10.3.4, 10.4

Changes in regional climate can be detected even though natural climate variations can temporarily increase or obscure anthropogenic climate change on decadal time scales. While anthropogenic forcing has contributed to multi-decadal mean precipitation changes in several regions, internal variability can delay emergence of the anthropogenic signal in long-term precipitation changes in many land regions ( ''high confidence'' ). Links to chapters 10.4

<div id="_idContainer006" class="Basic-Text-Frame"></div>

[[File:322a7c5d229e62e9e591f4710340336a IPCC_AR6_WGI_TS_Figure_3.png]]

'''Figure TS.3 |''' '''Emergence of changes in temperature over the historical period.''' ''The intent of this figure is to show how observed changes in temperature have emerged and that the emergence pattern agrees with model simulations.'' The observed change in temperature at a global warming level of 1°C (a), and the signal-to-noise ratio (the change in temperature at a global warming level of 1°C, divided by the size of year-to-year variations, ( b) ) using data from Berkeley Earth. The right panels show the zonal means of the maps and include data from different observational datasets (red) and the Coupled Model Intercomparison Project Phase 6 (CMIP6) simulations (black, including the 5–95% range) processed in the same way as the observations. Links to chapters 1.4.2, 10.4.3

Mean temperatures and heat extremes have emerged above natural variability in almost all land regions with ''high confidence'' . Changes in temperature-related variables, such as regional temperatures, growing season length, extreme heat and frost, have already occurred, and there is ''medium confidence'' that many of these changes are attributable to human activities. Several impact-relevant changes have not yet emerged from natural variability but will emerge sooner or later in this century depending on the emissions scenario ( ''high confidence'' ). Ocean acidification and deoxygenation have already emerged over most of the global open ocean, as has a reduction in Arctic sea ice ( ''high confidence'' ). Links to chapters 9.3.1, 9.6.4, 11.2, 11.3, 12.4, 12.5, Atlas.3–Atlas.11

<div id="TS.1.2.4" class="h3-container"></div>

<span id="ts.1.2.4-understanding-of-human-influence"></span>
==== TS.1.2.4 Understanding of Human Influence ====

<div id="h3-4-siblings" class="h3-siblings"></div>

'''The evidence for human influence on recent climate change has strengthened progressively from the IPCC Second Assessment Report to AR5 and is even stronger in this assessment, including for regional scales and for extremes. Human influence in the IPCC context refers to the human activities that lead to or contribute to a climate response, such as the human-induced emissions of greenhouse gases that subsequently alter the atmosphere’s radiative properties, resulting in warming of the atmosphere, ocean and land components of the climate system. Other human activities influencing climate include the emission of aerosols and other short-lived climate forcers, and land-use change such as urbanization. Progress in our understanding of human influence is gained from longer observational datasets, improved paleoclimate information, a stronger warming signal since AR5, and improvements in climate models, physical understanding and attribution techniques (see Core Concepts Box). Since AR5, the attribution to human influence has become possible across a wider range of climate variables and climatic impact-drivers (CIDs, see Core Concepts Box). New techniques and analyses drawing on several lines of evidence have provided greater confidence in attributing changes in regional weather and climate extremes to human influence ( ''high confidence'' ). Links to chapters 1.3, 1.5.1, Appendix 1.A, 3.1–3.8, 5.2, 6.4.2, 7.3.5, 7.4.4, 8.3.1, 10.4, Cross-Chapter Box 10.3, 11.2–11.9, 12.4'''
Combining the evidence from across the climate system increases the level of confidence in the attribution of observed climate change to human influence and reduces the uncertainties associated with assessments based on single variables. Links to chapters Cross-Chapter Box 10.3

Since AR5, the accumulation of energy in the Earth system has become established as a robust measure of the rate of global climate change on interannual-to-decadal time scales. The rate of accumulation of energy is equivalent to Earth’s energy imbalance and can be quantified by changes in the global energy inventory for all components of the climate system, including global ocean heat uptake, warming of the atmosphere, warming of the land and melting of ice. Compared to changes in global surface temperature, Earth’s energy imbalance (see Core Concepts Box) exhibits less variability, enabling more accurate identification and estimation of trends. Links to chapters Box 7.2 and [[IPCC:Wg1:Chapter:Chapter-7#7.2|Section 7.2]]

Identifying the human-induced components contributing to the energy budget provides an implicit estimate of the human influence on global climate change (Sections TS.2 and TS.3.1). Links to chapters Cross-Working Group Box: Attribution in Chapter 1, 3.8, 7.2.2, Box 7.2, Cross-Chapter Box 9.1

Regional climate changes can be moderated or amplified by regional forcing from land-use and land-cover changes or from aerosol concentrations and other short-lived climate forcers (SLCFs). For example, the difference in observed warming trends between cities and their surroundings can partly be attributed to urbanization ( ''very high confidence'' ). While established attribution techniques provide confidence in our assessment of human influence on large-scale climate changes (as described in Section TS.2), new techniques developed since AR5, including attribution of individual events, have provided greater confidence in attributing changes in climate extremes to climate change (Box TS.10). Multiple attribution approaches support the contribution of human influence to several regional multi-decadal mean precipitation changes ( ''high confidence'' ). Understanding about past and future changes in weather and climate extremes has increased due to better observation-based datasets, physical understanding of processes, an increasing proportion of scientific literature combining different lines of evidence, and improved accessibility to different types of climate models ( ''high confidence'' ) (see Sections TS.2 and TS.4). Links to chapters Cross-Working Group Box: Attribution in Chapter 1, 1.5, 3.2, 3.5, 5.2, 6.4.3, 8.3, 9.6, 10.1, 10.2, 10.3.3, 10.4.1, 10.4.2, 10.4.3, 10.5, 10.6, Cross-Chapter Box 10.3, Box 10.3, 11.1.6, 11.2–11.9, 12.4

<div id="TS.1.3" class="h2-container"></div>

<span id="ts.1.3-assessing-future-climate-change"></span>