Editing IPCC:AR6/SR15/Chapter-3 (section)

== Box 3.3: Lessons from Past Warm Climate Episodes ==

<div id="section-3-3-9-block-1"></div>

Climate projections and associated risk assessments for a future warmer world are based on climate model simulations. However, Coupled Model Intercomparison Project Phase 5 (CMIP5) climate models do not include all existing Earth system feedbacks and may therefore underestimate both rates and extents of changes (Knutti and Sedláček, 2012) <sup>[[#fn:r324|324]]</sup> . Evidence from natural archives of three moderately warmer (1.5°C–2°C) climate episodes in Earth’s past help to assess such long-term feedbacks (Fischer et al., 2018) <sup>[[#fn:r325|325]]</sup> .

While evidence over the last 2000 years and during the Last Glacial Maximum (LGM) was discussed in detail in the IPCC Fifth Assessment Report (Masson-Delmotte et al., 2013) <sup>[[#fn:r326|326]]</sup> , the climate system response during past warm intervals was the focus of a recent review paper (Fischer et al., 2018) <sup>[[#fn:r327|327]]</sup> summarized in this Box. Examples of past warmer conditions with essentially modern physical geography include the Holocene Thermal Maximum (HTM; broadly defined as about 10–5 kyr before present (BP), where present is defined as 1950), the Last Interglacial (LIG; about 129–116 kyr BP) and the Mid Pliocene Warm Period (MPWP; 3.3-3.0 Myr BP).

Changes in insolation forcing during the HTM (Marcott et al., 2013) <sup>[[#fn:r328|328]]</sup> and the LIG (Hoffman et al., 2017) <sup>[[#fn:r329|329]]</sup> led to a global temperature up to 1°C higher than that in the pre-industrial period (1850–1900); high-latitude warming was 2°C-4°C (Capron et al., 2017) <sup>[[#fn:r330|330]]</sup> , while temperature in the tropics changed little (Marcott et al., 2013) <sup>[[#fn:r331|331]]</sup> . Both HTM and LIG experienced atmospheric CO <sub>2</sub> levels similar to pre-industrial conditions (Masson-Delmotte et al. 2013). During the MPWP, the most recent time period when CO <sub>2</sub> concentrations were similar to present-day levels, the global temperature was &gt;1°C and Arctic temperatures about 8°C warmer than pre-industrial (Brigham-Grette et al., 2013) <sup>[[#fn:r332|332]]</sup> .

Although imperfect as analogues for the future, these regional changes can inform risk assessments such as the potential for crossing irreversible thresholds or amplifying anthropogenic changes (Box 3.3, Figure 1). For example, HTM and LIG greenhouse gas (GHG) concentrations show no evidence of runaway greenhouse gas releases under limited global warming. Transient releases of CO <sub>2</sub> and CH <sub>4</sub> may follow permafrost melting, but these occurrences may be compensated by peat growth over longer time scales (Yu et al., 2010) <sup>[[#fn:r333|333]]</sup> . Warming may release CO <sub>2</sub> by enhancing soil respiration, counteracting CO <sub>2</sub> fertilization of plant growth (Frank et al., 2010) <sup>[[#fn:r334|334]]</sup> . Evidence of a collapse of the Atlantic Meridional Overturning Circulation (AMOC) during these past events of limited global warming could not be found (Galaasen et al., 2014) <sup>[[#fn:r335|335]]</sup> .

The distribution of ecosystems and biomes (major ecosystem types) changed significantly during past warming events, both in the ocean and on land. For example, some tropical and temperate forests retreated because of increased aridity, while savannas expanded (Dowsett et al., 2016) <sup>[[#fn:r336|336]]</sup> . Further, poleward shifts of marine and terrestrial ecosystems, upward shifts in alpine regions, and reorganizations of marine productivity during past warming events are recorded in natural archives (Williams et al., 2009; Haywood et al., 2016) <sup>[[#fn:r337|337]]</sup> . Finally, past warming events are associated with partial sea ice loss in the Arctic. The limited amount of data collected so far on Antarctic sea ice precludes firm conclusions about Southern Hemisphere sea ice losses (de Vernal et al., 2013) <sup>[[#fn:r338|338]]</sup> .

Reconstructed global sea level rise of 6–9 m during the LIG and possibly &gt;6 m during the MPWP requires a retreat of either the Greenland or Antarctic ice sheets or both (Dutton et al., 2015) <sup>[[#fn:r339|339]]</sup> . While ice sheet and climate models suggest a substantial retreat of the West Antarctic ice sheet (WAIS) and parts of the East Antarctic ice sheet (DeConto and Pollard, 2016) <sup>[[#fn:r340|340]]</sup> during these periods, direct observational evidence is still lacking. Evidence for ice retreat in Greenland is stronger, although a complete collapse of the Greenland ice sheet during the LIG can be excluded (Dutton et al., 2015) <sup>[[#fn:r341|341]]</sup> . Rates of past sea level rises under modest warming were similar to or up to two times larger than rises observed over the past two decades (Kopp et al., 2013) <sup>[[#fn:r342|342]]</sup> . Given the long time scales required to reach equilibrium in a warmer world, sea level rise will ''likely'' continue for millennia even if warming is limited to 2°C.

Finally, temperature reconstructions from these past warm intervals suggest that current climate models underestimate regional warming at high latitudes (polar amplification) and long-term (multi-millennial) global warming. None of these past warm climate episodes involved the high rate of change in atmospheric CO <sub>2</sub> and temperatures that we are experiencing today (Fischer et al., 2018) <sup>[[#fn:r343|343]]</sup> .

<div id="section-3-3-9-block-2"></div>

<span id="box-3.3-figure-1"></span>
<!-- START IMG -->
<!-- IMG TITLE -->
'''Box 3.3, Figure 1'''

<span id="impacts-and-responses-of-components-of-the-earth-system."></span>
<!-- IMG CAPTION -->
'''Impacts and responses of components of the Earth System.'''
<!-- IMG FILE -->
[[File:fc4e9d430a967a9155c24261904c9124 box-3.3-figure-1024x746.jpg]]

Summary of typical changes found for warmer periods in the paleorecord, as discussed by Fischer et al. (2018) <sup>[[#fn:r344|344]]</sup> . All statements are relative to pre-industrial conditions. Statements in italics indicate that no conclusions can be drawn for the future. Note that significant spatial variability and uncertainty exists in the assessment of each component, and this figure therefore should not be referred to without reading the publication in detail. HTM: Holocene Thermal Maximum, LIG: Last Interglacial, MPWP: Mid Pliocene Warm Period. (Adapted from Fischer et al., 2018)

<!-- END IMG -->
<span id="ocean-chemistry"></span>
=== 3.3.10 Ocean Chemistry ===

<div id="section-3-3-10-block-1"></div>

Ocean chemistry includes pH, salinity, oxygen, CO <sub>2</sub> , and a range of other ions and gases, which are in turn affected by precipitation, evaporation, storms, river runoff, coastal erosion, up-welling, ice formation, and the activities of organisms and ecosystems (Stocker et al., 2013) <sup>[[#fn:r345|345]]</sup> . Ocean chemistry is changing alongside increasing global temperature, with impacts projected at 1.5°C and, more so, at 2°C of global warming (Doney et al., 2014) <sup>[[#fn:r346|346]]</sup> ( ''medium to high confidence'' ). Projected changes in the upper layers of the ocean include altered pH, oxygen content and sea level. Despite its many component processes, ocean chemistry has been relatively stable for long periods of time prior to the industrial period (Hönisch et al., 2012) <sup>[[#fn:r347|347]]</sup> . Ocean chemistry is changing under the influence of human activities and rising greenhouse gases ( ''virtually certain'' ; Rhein et al., 2013; Stocker et al., 2013) <sup>[[#fn:r348|348]]</sup> . About 30% of CO <sub>2</sub> emitted by human activities, for example, has been absorbed by the upper layers of the ocean, where it has combined with water to produce a dilute acid that dissociates and drives ocean acidification ( ''high confidence'' ) (Cao et al., 2007; Stocker et al., 2013) <sup>[[#fn:r349|349]]</sup> . Ocean pH has decreased by 0.1 pH units since the pre-industrial period, a shift that is unprecedented in the last 65 Ma ( ''high confidence'' ) (Ridgwell and Schmidt, 2010) <sup>[[#fn:r350|350]]</sup> or even 300 Ma of Earth’s history ( ''medium confidence'' ) (Hönisch et al., 2012) <sup>[[#fn:r351|351]]</sup> .

Ocean acidification is a result of increasing CO <sub>2</sub> in the atmosphere ( ''very high confidence'' ) and is most pronounced where temperatures are lowest (e.g., polar regions) or where CO <sub>2</sub> -rich water is brought to the ocean surface by upwelling (Feely et al., 2008) <sup>[[#fn:r352|352]]</sup> . Acidification can also be influenced by effluents from natural or disturbed coastal land use (Salisbury et al., 2008) <sup>[[#fn:r353|353]]</sup> , plankton blooms (Cai et al., 2011) <sup>[[#fn:r354|354]]</sup> , and the atmospheric deposition of acidic materials (Omstedt et al., 2015) <sup>[[#fn:r355|355]]</sup> . These sources may not be directly attributable to climate change, but they may amplify the impacts of ocean acidification (Bates and Peters, 2007; Duarte et al., 2013) <sup>[[#fn:r356|356]]</sup> . Ocean acidification also influences the ionic composition of seawater by changing the organic and inorganic speciation of trace metals (e.g., 20-fold increases in free ion concentrations of metals such as aluminium) – with changes expected to have impacts although they are currently poorly documented and understood ( ''low confidence'' ) (Stockdale et al., 2016) <sup>[[#fn:r357|357]]</sup> .

Oxygen varies regionally and with depth; it is highest in polar regions and lowest in the eastern basins of the Atlantic and Pacific Oceans and in the northern Indian Ocean (Doney et al., 2014; Karstensen et al., 2015; Schmidtko et al., 2017) <sup>[[#fn:r358|358]]</sup> . Increasing surface water temperatures have reduced oxygen in the ocean by 2% since 1960, with other variables such as ocean acidification, sea level rise, precipitation, wind and storm patterns playing roles (Schmidtko et al., 2017) <sup>[[#fn:r359|359]]</sup> . Changes to ocean mixing and metabolic rates, due to increased temperature and greater supply of organic carbon to deep areas, has increased the frequency of ‘dead zones’, areas where oxygen levels are so low that they no longer support oxygen dependent life (Diaz and Rosenberg, 2008) <sup>[[#fn:r360|360]]</sup> . The changes are complex and include both climate change and other variables (Altieri and Gedan, 2015) <sup>[[#fn:r361|361]]</sup> , and are increasing in tropical as well as temperate regions (Altieri et al., 2017) <sup>[[#fn:r362|362]]</sup> .

Ocean salinity is changing in directions that are consistent with surface temperatures and the global water cycle (i.e., precipitation versus evaporation). Some regions, such as northern oceans and the Arctic, have decreased in salinity, owing to melting glaciers and ice sheets, while others have increased in salinity, owing to higher sea surface temperatures and evaporation (Durack et al., 2012) <sup>[[#fn:r363|363]]</sup> . These changes in salinity (i.e., density) are also potentially contributing to large-scale changes in water movement (Section 3.3.8).

<span id="global-synthesis"></span>
=== 3.3.11 Global Synthesis ===

<div id="section-3-3-11-block-1"></div>

Table 3.2 features a summary of the assessments of global and regional climate changes and associated hazards described in this chapter, based on the existing literature. For more details about observation and attribution in ocean and cryosphere systems, please refer to the upcoming IPCC Special Report on the Ocean and Cryosphere in a Changing Climate (SROCC) due to be released in 2019.

<div id="section-3-3-11-block-2"></div>

<span id="table-3.2"></span>
<!-- START TABLE -->
'''Table 3.2'''

<span id="summary-of-assessments-of-global-and-regional-climate-changes-and-associated-hazards"></span>
'''Summary of assessments of global and regional climate changes and associated hazards'''

Confidence and likelihood statements are quoted from the relevant chapter text and are omitted where no assessment was made, in which case the IPCC Fifth Assessment Report (AR5) assessment is given where available. GMST: global mean surface temperature, AMOC: Atlantic Meridional Overturning Circulation, GMSL: global mean sea level.

<!-- TABLE -->
{| class="wikitable"
|-
!
! Observed change (recent past versus pre-industrial)
! Attribution of observed change to human-induced forcing (present-day versus pre-industrial)
! Projected change at 1.5°C of global warming compared to pre-industrial (1.5°C versus 0°C)
! Projected change at 2°C of global warming compared to pre-industrial (2°C versus 0°C)
! Differences between 2°C and 1.5°C of global warming
|-
! GMST<br />
anomaly
| GMST anomalies were 0.87°C (±0.10°C ''likely'' range) above pre-industrial (1850–1900) values in the 2006–2015 decade, with a recent warming of about 0.2°C (±0.10°C) per decade ( ''high confidence'' )
[Chapter 1]

| The observed 0.87°C GMST increase in the 2006–2015 decade compared to<br />
pre-industrial (1850–1900) conditions was mostly human-induced ( ''high confidence'' )
<br/><br/>
Human-induced warming reached about 1°C (±0.2°C ''likely'' range) above pre-industrial levels in 2017

[Chapter 1]

| 1.5°C
| 2°C
| 0.5°C
|-
! Temperature extremes
| Overall decrease in the number of cold days and nights and overall increase in the number of warm days and nights at the global scale on land ( ''very likely'' )
Continental-scale increase in intensity and frequency of hot days and nights, and decrease in intensity and frequency of cold days and nights, in North America, Europe and Australia ( ''very likely'' )

Increases in frequency or duration of warm spell lengths in large parts of Europe, Asia and Australia ( ''high confidence'' ( ''likely'' )), as well as at the global scale ( ''medium confidence'' )

[Section 3.3.2]

| Anthropogenic forcing has contributed to the observed changes in frequency and intensity of daily temperature extremes on the global scale since the mid-20th century ( ''very likely'' )
[Section 3.3.2]

| Global-scale increased intensity and frequency of hot days and nights, and decreased intensity and frequency of cold days and nights ( ''very likely'' )
Warming of temperature extremes highest over land, including many inhabited regions ( ''high confidence'' ), with increases of up to 3°C in the mid-latitude warm season and up to 4.5°C in the high-latitude cold season ( ''high confidence'' )

Largest increase in frequency of unusually hot extremes in tropical regions ( ''high confidence'' )

[Section 3.3.2]

| Global-scale increased intensity and frequency of hot days and nights, and decreased intensity and frequency of cold days and nights ( ''very likely'' )
Warming of temperature extremes highest over land, including many inhabited regions ( ''high confidence'' ), with increases of up to 4°C in the mid-latitude warm season and up to 6°C in the high-latitude cold season ( ''high confidence'' )

Largest increase in frequency of unusually hot extremes in tropical regions ( ''high confidence'' )

[Section 3.3.2]

| Global-scale increased intensity and frequency of hot days and nights, and decreased intensity and frequency of cold days and nights ( ''high confidence'' )
Global-scale increase in length of warm spells and decrease in length of cold spells ( ''high confidence'' )

Strongest increase in frequency for the rarest and most extreme events ( ''high confidence'' )

Particularly large increases in hot extremes in inhabited regions ( ''high confidence'' )

[Section 3.3.2]

|-
! Heavy precipitation
| More areas with increases than decreases in the frequency, intensity and/or amount of heavy precipitation ( ''likely'' )
[Section 3.3.3]

| Human influence contributed to the global-scale tendency towards increases in the frequency, intensity and/or amount of heavy precipitation events ( ''medium confidence'' )
[Section 3.3.3; AR5 Chapter 10 (Bindoff et al., 2013a) <sup>[[#fn:r364|364]]</sup> ]

| Increases in frequency, intensity and/or amount heavy precipitation when averaged over global land, with positive trends in several regions ( ''high confidence'' )
[Section 3.3.3]

| Increases in frequency, intensity and/or amount heavy precipitation when averaged over global land, with positive trends in several regions ( ''high confidence'' )
[Section 3.3.3]

| Higher frequency, intensity and/or amount of heavy precipitation when averaged over global land, with positive trends in several regions ( ''medium confidence'' )
Several regions are projected to experience increases in heavy precipitation at 2°C versus 1.5°C ( ''medium confidence'' ), in particular in high-latitude and mountainous regions, as well as in eastern Asia and eastern North America ( ''medium confidence'' )

[Section 3.3.3]

|-
! Drought and dryness
| ''High confidence'' in dryness trends in some regions, especially drying in the Mediterranean region (including southern Europe, northern Africa and the Near East)
''Low confidence'' in drought and dryness trends at the global scale

[Section 3.3.4]

| ''Medium confidence'' in attribution of drying trends in southern Europe (Mediterranean region)
''Low confidence'' elsewhere, in part due to large interannual variability and longer duration (and thus lower frequency) of drought events, as well as to dependency on the dryness index definition applied

[Section 3.3.4]

| ''Medium confidence'' in drying trends in the Mediterranean region
''Low confidence'' elsewhere, in part due to large interannual variability and longer duration (and thus lower frequency) of drought events, as well as to dependency on the dryness index definition applied

Increases in drought, dryness or precipitation deficits projected in some regions compared to the pre-industrial or present-day conditions, but substantial variability in signals depending on considered indices or climate model ( ''medium confidence'' )

[Section 3.3.4]

| ''Medium confidence'' in drying trends in the Mediterranean region and Southern Africa
''Low confidence'' elsewhere, in part due to large interannual variability and longer duration (and thus lower frequency) of drought events, as well as to dependency on the dryness index definition applied

Increases in drought, dryness or precipitation deficits projected in some regions compared to the pre-industrial or present-day conditions, but substantial variability in signals depending on considered indices or climate model ( ''medium confidence'' ).

[Section 3.3.4]

| ''Medium confidence'' in stronger drying trends in the Mediterranean region and Southern Africa
''Low confidence'' elsewhere, in part due to large interannual variability and longer duration (and thus lower frequency) of drought events, as well as to dependency on the dryness index definition applied

[Section 3.3.4]

|-
! Runoff and river flooding
| Streamflow trends mostly not statistically significant ( ''high confidence'' )
Increase in flood frequency and extreme streamflow in some regions ( ''high confidence'' )

[Section 3.3.5]

| Not assessed in this report
| Expansion of the global land area with a significant increase in runoff ( ''medium confidence'' )
Increase in flood hazard in some regions ( ''medium confidence'' )

[Section 3.3.5]

| Expansion of the global land area with a significant increase in runoff ( ''medium confidence'' )
Increase in flood hazard in some regions ( ''medium confidence'' )

[Section 3.3.5]

| Expansion of the global land area with significant increase in runoff ( ''medium confidence'' )
Expansion in the area affected by flood hazard ( ''medium confidence'' )

[Section 3.3.5]

|-
! Tropical and extra-tropical cyclones
| ''Low confidence'' in the robustness of observed changes
[Section 3.3.6]

| Not meaningful to assess given ''low confidence'' in changes, due to large interannual variability, heterogeneity of the observational record and contradictory findings regarding trends in the observational record
| Increases in heavy precipitation associated with tropical cyclones ( ''medium confidence'' )
| Further increases in heavy precipitation associated with tropical cyclones ( ''medium confidence'' )
| Heavy precipitation associated with tropical cyclones is projected to be higher at 2°C compared to 1.5°C global warming ( ''medium confidence'' ). ''Limited evidence'' that the global number of tropical cyclones will be lower under 2°C of global warming compared to under 1.5°C of warming, but an increase in the number of very intense cyclones ( ''low confidence'' )
|-
! Ocean circulation and temperature
| Observed warming of the upper ocean, with slightly lower rates than global warming ( ''virtually certain'' )
Increased occurrence of marine heatwaves ( ''high confidence'' )

AMOC has been weakening over recent decades ( ''more likely than not'' )

[Section 3.3.7]

| ''Limited evidence'' attributing the weakening of AMOC in recent decades to anthropogenic forcing
[Section 3.3.7]

| colspan="3"| Further increases in ocean temperatures, including more frequent marine heatwaves ( ''high confidence'' )
AMOC will weaken over the 21st century and substantially so under high levels (more than 2°C) of global warming ( ''very likely'' )

[Section 3.3.7]

|-
! rowspan="2"| Sea ice
| rowspan="2"| Continuing the trends reported in AR5, the annual Arctic sea ice extent decreased over the period 1979–2012. The rate of this decrease was ''very likely'' between 3.5 and 4.1% per decade (0.45 to 0.51 million km <sup>2</sup> per decade)
[AR5 Chapter 4 (Vaughan et al., 2013) <sup>[[#fn:r365|365]]</sup> ]

| rowspan="2"| Anthropogenic forcings are ''very likely'' to have contributed to Arctic sea ice loss since 1979
[AR5 Chapter 10<br />
(Bindoff et al., 2013a) <sup>[[#fn:r366|366]]</sup> ]

| At least one sea-ice-free Arctic summer after about 100 years<br />
of stabilized warming ( ''medium confidence'' )[Section 3.3.8]
| At least one sea-ice-free<br />
Arctic summer after about<br />
10 years of stabilized warming ( ''medium confidence'' )[Section 3.3.8]
| Probability of sea-ice-free Arctic summer greatly reduced at 1.5°C versus 2°C of global warming ( ''medium confidence'' )
[Section 3.3.8]

|-
| colspan="3"| Intermediate temperature overshoot has no long-term consequences for Arctic sea ice cover<br />
( ''high confidence'' )
[3.3.8]

|-
! Sea level
| It is ''likely'' that the rate of GMSL rise has continued to increase since the early 20th century, with estimates that range from 0.000 [–0.002 to 0.002] mm yr <sup>–2</sup> to 0.013 [0.007 to 0.019] mm yr <sup>–2</sup>
[AR5 Chapter 13<br />
(Church et al., 2013) <sup>[[#fn:r367|367]]</sup> ]

| It is very ''likely'' that there is a substantial contribution from anthropogenic forcings to the global mean sea level rise since the 1970s
[AR5 Chapter 10 (Bindoff et al., 2013a) <sup>[[#fn:r368|368]]</sup> ]

| Not assessed in this report
| GMSL rise will be about 0.1 m (0.00–0.20 m) less at 1.5°C versus 2°C global warming ( ''medium confidence'' )[Section 3.3.9]
|-
! Ocean<br />
chemistry
| Ocean acidification due to increased CO <sub>2</sub> has resulted in a 0.1 pH unit decrease since the pre-industrial period, which is unprecedented in the last 65 Ma ( ''high confidence'' )
[Section 3.3.10]

| The oceanic uptake of anthropogenic CO <sub>2</sub> has resulted in acidification of surface waters ( ''very high confidence'' ).
[Section 3.3.10]

| colspan="3"| Ocean chemistry is changing with global temperature increases, with impacts projected at 1.5°C and, more so, at 2°C of warming ( ''high confidence'' )
[Section 3.3.10]

|}
<!-- END TABLE -->

<span id="observed-impacts-and-projected-risks-in-natural-and-human-systems"></span>