Editing IPCC:AR6/WGII/Chapter-16 (section)

== 16.2 Synthesis of observed impacts of changes in climate related systems ==

<div id="h1-3-siblings" class="h1-siblings"></div>

This section synthesises the observed impacts of changes in climate-related systems ( [[#16.2.1|Section 16.2.1]] ) on different natural, human and managed systems (outlined in Chapters 2–8) and regions (outlined in Chapters 9–15). To stay as specific as possible given the required level of aggregation, we decided in favour of a summary along specific prominent indicators such as ‘crop yields’ or ‘areas burned by wildfires’ instead of an assessment across broad categories such as ‘food production’ which could include a broad range of measures ranging from climate-induced changes in growing seasons to impacts on livestock and fisheries, etc., or ‘wildfires’ which could also cover impacts on the frequency, intensity, timing, or emissions and health impacts of wildfires. This decision for specificity certainly implies a decision against comprehensiveness. In addition, the level of specificity has to be adjusted given the literature basis which is quite broad regarding crop yields but still limited and less harmonised regarding indicators when it comes to, for example, conflicts. A broader discussion can be found in the sectoral or regional chapters that all cover ‘observed impacts’ individually. [[#16.2.1|Section 16.2.1]] provides key definitions, followed by recent advances in available methods and data for climate impact attribution ( [[#16.2.2|Section 16.2.2]] ), and the assessment of observed impacts of changes in climate related systems ( [[#16.2.3|Section 16.2.3]] ). It is important to note that the assessment is primarily based on peer-reviewed literature, that is, it is limited to the regions and phenomena for which such studies are available. So ‘no assessment’ in a certain region does not imply that the considered type of impact did not occur in this region.

<div id="16.2.1" class="h2-container"></div>

<span id="definitions"></span>
=== 16.2.1 Definitions ===

<div id="h2-5-siblings" class="h2-siblings"></div>

The section adopts the general definition of '''detection''' as ‘demonstration that a considered system has changed without providing reasons for the change’ and '''attribution''' as ‘identifying the causes of the observed long-term change in an impact indicator or of the change in the temporal or spatial extent, the intensity or frequency of a specific event’ (see Glossary (Annex II)).

Based on these general definitions and following the approach applied in WGII AR5 [[IPCC:Wg2:Chapter:Chapter-18|Chapter 18]] ( [[#Cramer--2014|Cramer et al., 2014]] ), we define an '''observed impact''' as the difference between the observed state of a '''natural, human or managed system''' and a counterfactual baseline that characterises the system’s state in the absence of changes in the '''climate-related systems''' , defined here as climate system including the ocean and the cryosphere as physical or chemical systems.

The difference between the observed and the counterfactual baseline state is considered the change in the natural, human or managed system that is attributed to the changes in the climate-related systems ( '''impact attribution''' ). The counterfactual baseline may be stationary or may change over time, for example due to direct human influences such as changes in land use patterns and agricultural or water management affecting exposure and vulnerability to climate-related hazards (see [[#16.2.3|Section 16.2.3]] for methods on how to construct the counterfactual baseline).

In line with the AR5 definition, ‘changes in climate-related systems’ here refer to any long-term trend, irrespective of the underlying causes; thus, an observed impact is not necessarily an observed impact of anthropogenic climate forcing. For example, in this section, sea level rise is defined as relative sea level rise measured against a land-based reference frame (tide gauge measurements), meaning that it is driven not only by thermal expansion and loss of land ice influenced by anthropogenic climate forcing, but also by vertical land movements. As attribution of coastal damages to sea level rise does not distinguish between these components, it does not imply attribution to anthropogenic forcing. Where the literature does allow attribution of changes in natural, human or managed systems to anthropogenic climate forcing (‘joint attribution’, [[#Rosenzweig--2007|Rosenzweig et al., 2007]] ), this is highlighted in the assessment. Often the attribution of changes in the natural, human or managed systems to anthropogenic forcing can be done in a two-step approach where (i) an observed change in a climate-related system is attributed to anthropogenic climate forcing (‘climate attribution’) and (ii) changes in natural, human or managed systems are attributed to this change in the climate-related system (‘impact attribution’).

For climate attribution, the main challenge is the separation of externally human forced changes in the climate-related systems from their internal variability, while for impact attribution it often is the separation of the effects of other external forcings (i.e., direct human influences or natural disturbances) from the impacts of the changes in the climate-related systems. Direct influences not related to changes in the climate-related systems could, for example, be pollution and land use changes amplifying biodiversity losses, intensification of fishing reducing fish stocks, and increasing protection reducing losses due to river floods. The direct human or natural influences may counter the impacts of climate change (e.g., climate change may have reduced flood hazards, but exposure may have increased as people have moved to flood-prone areas, resulting in no change in observed damages). Given the definition of impact attribution, this means that there may be an observed impact of climate change without the detection of a change in the natural, human or managed system. This is different from ‘climate attribution’, where detection and attribution usually are consecutive steps.

Changes in climate-related systems can certainly also affect natural, human and managed systems through indirect effects on land use, pollution or exposure. However, these indirect effects are barely addressed in existing studies.

In addition to impact attribution, there is research on the identification of natural, human or managed systems’ response to short-term (typically daily, monthly or annual) weather fluctuations or individual ''extreme weather events'' . As different from impact attribution, we separately define:

'''‘Identification of weather sensitivity’''' refers to the attribution of the response of a system to fluctuations in weather and short-term changes in the climate-related systems including individual ''extreme weather events'' (e.g., a heatwave or storm surge).

Typical questions addressed include: ‘How much of the observed variability of crop yields is due to variations in weather conditions compared to contributions from management changes?’ (e.g., [[#Ray--2015|Ray et al., 2015]] ; [[#Müller--2017|Müller et al., 2017]] ) and ‘Can weather fluctuations explain part of the observed variability in annual national economic growth rates?’ (e.g., [[#Burke--2015|Burke et al., 2015]] ). Identification of weather sensitivity may also address the effects of individual ''climate extremes'' , for example asking, ‘Was the observed outbreak of cholera triggered by an associated flood event?’ (e.g., [[#Rinaldo--2012|Rinaldo et al., 2012]] ; [[#Moore--2017b|Moore et al., 2017b]] ). It is important to note that sensitivity could be described in diverse ways and that, for example, the fraction of the observed variability in a system explained by weather variability differs from the strength of the systems’ response to a specific change in a weather variable. Nevertheless, all these different measures are integrated in the ‘identification of weather sensitivity’ assessment, where ‘sensitivity’ should not be considered a quantitative one-dimensional mathematical measure.

In this chapter, we explicitly distinguish between assessment statements related to ‘climate attribution’ (listed in Table SM16.21), ‘impact attribution’ (listed in Table SM16.22) and ‘identification of weather sensitivity’ (listed in Table SM16.23). The identification of ‘weather sensitivity’ does not necessarily imply that there also is an impact of long-term changes in the climate-related systems on the considered system. However, if the probability or intensity of an ''extreme weather event'' has increased due to anthropogenic forcing (‘climate attribution’) ( [[#NASEM--2016|NASEM, 2016]] ; WGI AR6 [[IPCC:Wg2:Chapter:Chapter-11|Chapter 11]] [[#Seneviratne--2021|Seneviratne et al., 2021]] ) and the event is also identified as an important driver of an observed fluctuation in a natural, human or managed system (‘identification of weather sensitivity’), then the observed fluctuation is considered (partly) attributed to long-term climate change (‘impact attribution’) and even to anthropogenic forcing.

<div id="16.2.2" class="h2-container"></div>

<span id="methods-and-data-for-impact-attribution-including-recent-advances"></span>
=== 16.2.2 Methods and Data for Impact Attribution Including Recent Advances ===

<div id="h2-6-siblings" class="h2-siblings"></div>

By definition, the counterfactual baseline required for impact attribution cannot be observed. However, it may be approximated by impact model simulations forced by a stationary climate, for example derived by de-trending the observed climate ( [[#Diffenbaugh--2017|Diffenbaugh et al., 2017]] ; [[#Mengel--2021|Mengel et al., 2021]] ), while other relevant drivers (e.g., land use changes or application of pesticides) of changes in the system of interest (e.g., a bird population) evolve according to historical conditions. To attribute to anthropogenic climate forcing, the anthropogenic trends in climate are estimated from a range of different climate models and subtracted from the observed climate (e.g., [[#Abatzoglou--2016|Abatzoglou and Williams, 2016]] , for changes in the extent of forest fires or [[#Diffenbaugh--2019|Diffenbaugh and Burke, 2019]] , for effects on economic inequality) or the ‘no anthropogenic climate forcing’ baseline is directly derived from a large ensemble of climate model simulations not accounting for anthropogenic forcings (e.g., Kirchmeier-Young et al.., 2019b, for the extent of forest fires). In any case, it has to be demonstrated that the applied impact models are able to explain the observed changes in natural, human or managed systems by, for example, reproducing the observations when forced by observed changes in climate-related systems and other relevant drivers.

In a situation where an influence of other direct human drivers can be excluded (e.g., by restriction to remote areas not affected by direct human interventions), the ‘no climate-change’ baseline can also be approximated by data from early observational periods with no or minor levels of climate change. In particular, the contribution of climate change to the observed changes in ecosystems is often also determined by a ‘multiple lines of evidence’ approach where the baseline is not formally quantified but the observed changes are identified as a signal of climate change compared with a no-climate-change situation based on process understanding from, for example, palaeo data and laboratory or field experiments in combination with individual long-term observational records and the large-scale spatial or temporal pattern of observed changes that can hardly be explained by alternative drivers ( [[#Parmesan--2013|Parmesan et al., 2013]] ).

To date, explicit accounting for direct human or natural influences is often hampered by an incomplete understanding of the processes and limited observational data. There are, however, first studies demonstrating the potential of detailed process-based or empirical modelling that explicitly accounts for known variations in direct human or natural drivers and separate their effects from the ones induced by changes in the climate-related systems. Examples are Butler et al. (2018) for the separation of growing season adjustments from within growing season climate effects on US crop yields; [[#Wang--2019|Wang and Hijmans (2019)]] separating effects of shifts in land use from climate effects; [[#Jongman--2015|Jongman et al. (2015)]] ; [[#Formetta--2019|Formetta and Feyen (2019)]] and Tanoue et al. (2016) for the separation of changes in exposure and vulnerability from climate effects on river floods; [[#Kirchmeier-Young--2019b|Kirchmeier-Young et al. (2019b)]] for wildfire attribution; and Venter et al. (2018) for the attribution of ecosystem structural changes to climate change versus other disturbances.

There also has been significant progress in the compilation of fragmented and distributed observational data (e.g., [[#Cohen--2018|Cohen et al., 2018]] , for phenological ecosystem changes; [[#Poloczanska--2013|Poloczanska et al., 2013]] , for distributional shifts in marine ecosystems; [[#Andela--2019|Andela et al., 2019]] , with the new global fire atlas including information about individual fire size, duration, speed and direction) as well as in the regional disaggregation (e.g., [[#Ray--2015|Ray et al., 2015]] , for crop yields ) allowing for the identification of an overall picture of the impacts of progressing climate change. Given the ever-increasing body of literature on observed changes in natural, human and managed systems, there also is a first machine learning approach for an automated identification of relevant literature that could complement or support expert assessments as the one provided here ( [[#Callaghan--2021|Callaghan et al., 2021]] ).

<div id="16.2.3" class="h2-container"></div>

<span id="observed-impacts-of-changes-in-climate-related-systems"></span>
=== 16.2.3 Observed Impacts of changes in climate-related systems ===

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

In this section, we synthesise observed impacts of changes in climate-related systems across a range of ecosystems, sectors and regions. Figure 16.2 summarises the attribution of observed (regional) changes in natural, human or managed systems (orange symbols and confidence ratings), the quantification of weather sensitivity of those systems (blue symbols and confidence ratings) and the attribution of underlying changes in the climate-related systems to anthropogenic forcing (grey symbols and confidence ratings). The figure can be read as a summary and table of content for the underlying Tables SM16.21 on climate attribution, SM16.22 on impact attribution and SM16.23 on identification of weather sensitivity that provide the more detailed explanations behind each regional or global assessment, including all references. The synthesis was generated in collaboration with ‘detection and attribution contact persons’ from the individual chapters that each includes its own assessment of observed impacts, and contributing authors on individual topics. The synthesis of ‘climate attribution’ studies in Table SM16.21 was particularly informed by the WGI assessment.

<div id="_idContainer007" class="Figure"></div>

[[File:7f4d733ca1dfd7d8756f05a0677fbd10 IPCC_AR6_WGII_Figure_16_002.png]]

'''Figure 16.2 |''' '''Impact of climate change or weather fluctuations.'''

If Figure 16.2 only provides an assessment of attributed impacts on a given system (e.g., phenology shifts in terrestrial ecosystems) but does not include an associated ‘identification of weather sensitivity’ that does not mean that the system is not sensitive to weather fluctuations. The focus of our assessment was on ‘impact attribution’, and we only provide an assessment of ‘weather sensitivities’ if the literature has turned out to provide only ''limited evidence'' on impacts of long-term changes in climate-related systems but rather addressed the responses of natural, human or managed systems to short-term weather fluctuations in the climate-related ones.

<div id="16.2.3.1" class="h3-container"></div>

<span id="ecosystems"></span>
==== 16.2.3.1 Ecosystems ====

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

The collapse or transformation of ecosystems is one of the most abrupt potential tipping points associated with climate change. Climate change has started to induce such tipping points, with the first examples including mass mortality in coral reef ecosystems (e.g., [[#Donner--2017|Donner et al., 2017]] ; [[#Hughes--2018|Hughes et al., 2018]] ; [[#Hughes--2019|Hughes et al., 2019]] ) ( ''high confidence'' ), and changes in vegetation cover triggered by wildfires with climate change suppressing the recovery of the former cover ( [[#Tepley--2017|Tepley et al., 2017]] ; [[#Davis--2019|Davis et al., 2019]] ) ( ''low confidence'' because of the still limited number of studies). Another example of an abrupt change in an ecosystem triggered by a climate extreme is the shift from kelp- to urchin-dominated communities along parts of the Western North America coast ( [[#Rogers-Bennett--2019|Rogers-Bennett and Catton, 2019]] ; [[#McPherson--2021|McPherson et al., 2021]] , see ‘Marine ecosystems—Kelp forest’, Table SM16.22). The loss of kelp forests was induced by a marine heatwave where anthropogenic climate forcing has been shown to have increased the probability for an event of that duration by at least a factor of 33 ( [[#Laufkötter--2020|Laufkötter et al., 2020]] ). Many terrestrial ecosystems on all continents show evidence of significant structural transformation, including woody thickening and ‘greening’ in more water-limited ecosystems, with a significant role played by rising atmospheric CO 2 fertilisation in these trends ( ''high confidence'' ) ( [[#Fang--2017|Fang et al., 2017]] ; [[#Stevens--2017|Stevens et al., 2017]] ; [[#Burrell--2020|Burrell et al., 2020]] ). Climate change is identified as a major driver of increases in burned areas in the Western USA ( ''high confidence'' , see ‘Terrestrial ecosystems—Burned areas’, Table SM16.22).

There is also a clear footprint of climate change on species distribution, with appreciable proportions of tropical species expanding into the ranges of temperate species, and boreal species moving into Arctic regions ( ''high confidence'' , see ‘Marine ecosystems—Range reduction and shift’ and ‘Terrestrial ecosystems—Range reduction and shift’, Table SM16.22). Climate change has also shifted the phenology of animals and plants on land and in the ocean ( ''high confidence'' , see ‘Marine ecosystems—Phenology shift’ and ‘Terrestrial ecosystems—Phenology shifts’, Table SM16.22). Both processes have led to emerging hybridisation, competition, temporal or spatial mismatches in predator–prey, guest–host relationships, and the invasion of alien plant pests or pathogens ( [[#Edwards--2004|Edwards and Richardson, 2004]] ; [[#Bebber--2013|Bebber et al., 2013]] ; [[#Parmesan--2013|Parmesan et al., 2013]] ; [[#Millon--2014|Millon et al., 2014]] ; [[#Thackeray--2016|Thackeray et al., 2016]] ).

<div id="16.2.3.2" class="h3-container"></div>

<span id="water-distributionriver-flooding-and-reduction-in-water-availability"></span>
==== 16.2.3.2 Water Distribution—River Flooding and Reduction in Water Availability ====

<div id="h3-9-siblings" class="h3-siblings"></div>

Observed trends in high river flows strongly vary across regions but also with the considered time period ( [[#Gudmundsson--2019|Gudmundsson et al., 2019]] ; [[#Gudmundsson--2021|Gudmundsson et al., 2021]] ) as influenced by climate oscillations such as the El Niño-Southern Oscillation ( [[#Ward--2014|Ward et al., 2014]] ). On the global scale, the spatial pattern of observed trends is largely explained by observed changes in climate conditions as demonstrated by multi-model hydrological simulations forced by observed weather, while the considered direct human influences play only a minor role on global scale ( [[#Gudmundsson--2021|Gudmundsson et al., 2021]] , see ‘Water distribution—Flood induced economic damages’, Table SM16.22). The annual total number of reported fatalities from flooding shows a positive trend (1.5% yr −1 from 1960 to 2013, [[#Tanoue--2016|Tanoue et al., 2016]] ) which appears to be primarily driven by changes in exposure dampened by a reduction in vulnerability, while climate-induced increases in affected areas show only a weak positive trend on the global scale. However, the signal of climate change in flood-induced fatalities may be lost in the regional aggregation, where effects of increasing and decreasing hazards may cancel out. Thus, a climate-driven increase in flood-induced damages becomes detectable in continental subregions with increasing discharge, while the signal of climate change may not be detectable without disaggregation ( [[#Sauer--2021|Sauer et al., 2021]] , see ‘Water distribution—Flood induced economic damages’, Table SM16.22). Compared with river floods, the analysis of impacts of long-term changes in the climate-related systems on the reduction in water availability is much more fragmented and reduced to individual case studies regarding associated societal impacts (see ‘Water distribution—Reductions in water availability + induced damages and fatalities’, Table SM16.22). At the same time, weather fluctuations have led to reductions in water availability with severe societal consequences and high numbers of drought-induced fatalities and damages in particular in Africa and Asia (see ‘Water distribution—Reductions in water availability + induced damages and fatalities’, Table SM16.23) and impacts on malnutrition (see ‘Food system—Malnutrition’, Table SM16.23). Although anthropogenic climate forcing has increased droughts’ intensity or probability in many regions of the world ( ''medium confidence'' ), (see ‘Atmosphere—Droughts’, Table SM16.21) the existing knowledge has not yet been systematically linked to attribute long-term trends in malnutrition, fatalities and damages induced by reduced water availability to anthropogenic climate forcing or long-term climate change. For impacts of individual attributable drought events, see Table 4.5 and ‘Water distribution—Reductions in water availability + induced damages and fatalities’, Table SM16.23.

<div id="16.2.3.3" class="h3-container"></div>

<span id="coastal-systems"></span>
==== 16.2.3.3 Coastal systems ====

<div id="h3-10-siblings" class="h3-siblings"></div>

With their enormous destructive power, tropical cyclones represent a major risk for coastal systems (see ‘Coastal systems—Damages’, Table SM16.23). Despite its relevance, confidence in the influence of anthropogenic climate forcing on the strength and occurrence probability of tropical storms themselves is still low (see ‘Coastal systems—Tropical cyclone activity’, Table SM16.21). However, anthropogenic climate forcing has become the dominant driver of sea level rise ( ''high confidence'' ) (see ‘Coastal systems—Mean and extreme sea levels’, Table SM16.21) and has increased the risk of coastal flooding, including inundation induced by tropical cyclones. In addition, anthropogenic climate forcing has increased the amount of rainfall associated with tropical cyclones ( ''high confidence)'' ( [[#Risser--2017|Risser and Wehner, 2017]] ; [[#Van%20Oldenborgh--2017|Van Oldenborgh et al., 2017]] ; [[#Wang--2018|Wang et al., 2018]] , for Hurricane Harvey in 2017; [[#Patricola--2018|Patricola and Wehner, 2018]] , for hurricane Katrina in 2005, Irma in 2017 and Maria in 2017, see ‘Atmosphere—Heavy precipitation’, Table SM16.21). Assuming that the extreme rainfall is a major driver of the total damages induced by the tropical cyclone, the contribution of anthropogenic climate forcing to the occurrence probability of the observed rainfall (fraction of attributable risk) can also be considered the fraction of attributable risk of the hurricane-induced damages or fatalities ( [[#Frame--2020|Frame et al., 2020]] ; [[#Clarke--2021|Clarke et al., 2021]] , see ‘Coastal systems—Damages’, Table SM16.22). However, first studies do not only quantify the change in occurrence probabilities but translate the actual change in climate-related systems into the additional area affected by flooding in a process-based way ( [[#Strauss--2021|Strauss et al., 2021]] for the contribution of anthropogenic sea level rise (SLR) to damages induced by Hurricane Sandy; [[#Wehner--2021|Wehner and Sampson, 2021]] for the contribution increased precipitation to damages induced by Hurricane Harvey) and attribute a considerable part of the observed damage to anthropogenic climate forcing. In addition, disruption of local economic activity in Annapolis, Maryland and loss of areas and settlements in Micronesia and Solomon Islands have been attributed to relative SLR ( [[#Nunn--2017|Nunn et al., 2017]] ; [[#Albert--2018|Albert et al., 2018]] ; [[#Hino--2019|Hino et al., 2019]] ), while permafrost thawing and sea ice retreat are additional drivers of observed coastal damages in Alaska ( [[#Albert--2016|Albert et al., 2016]] ; [[#Smith--2016|Smith and Sattineni, 2016]] ; [[#Fang--2017|Fang et al., 2017]] ).

<div id="16.2.3.4" class="h3-container"></div>

<span id="food-system"></span>
==== 16.2.3.4 Food system ====

<div id="h3-11-siblings" class="h3-siblings"></div>

Crop yields respond to weather variations but also to increasing atmospheric CO 2 , changes in management (e.g., fertilizer input, changes in varieties), diseases and pests. However, the weather signal is clearly detectable in national and subnational annual yield statistics in main production regions (see ‘Food system—Crop yields’, Table SM16.23). Over the last decades, crop yields have increased nearly everywhere mainly due to technological progress (e.g., [[#Lobell--2007|Lobell and Field, 2007]] [global]; [[#Butler--2018|Butler et al., 2018]] [USA]; [[#Hoffman--2018|Hoffman et al., 2018]] [Sub-Saharan Africa]; [[#Agnolucci--2019|Agnolucci and De Lipsis, 2019]] [Europe]), with only minor areas not experiencing improvements in maize, wheat, rice and soy yields. However, meanwhile, stagnation or decline in yields is also observed in parts of the harvested areas ( ''high confidence'' ) (~20–40% of harvested areas of maize, wheat, rice and soy with wheat being most affected) ( [[#Ray--2012|Ray et al., 2012]] ; [[#Iizumi--2018|Iizumi et al., 2018]] ). Evidence on the contribution of climate change to recent trends is still limited (see ‘Food system—Crop yields’, Table SM16.22). Current global-scale process-based simulations forced by simulated historical and pre-industrial climate lack an evaluation to what degree simulations reproduce observed yields ( [[#Iizumi--2018|Iizumi et al., 2018]] ). Global-scale empirical approaches do not explicitly account for extreme weather events but growing season average temperatures and precipitation (e.g., [[#Lobell--2011|Lobell et al., 2011]] ; [[#Ray--2019|Ray et al., 2019]] ). In addition, studies are constrained by only fragmented information about changes in agricultural management such as growing season adjustments. Some of these limitations have been overcome in regional studies indicating a climate-induced increase (28% of observed trend since 1981) in maize yields in the USA ( [[#Butler--2018|Butler et al., 2018]] , based on a detailed accounting of impacts of extreme temperatures and growing season adjustments) and a climate-induced decrease in millet and sorghum yields (10–20% for millet and 5–15% for sorghum in 2000–2009 compared with pre-industrial conditions) in Africa and a negative effect of historical climate change on potential wheat yields (27% reduction from 1990 to 2015) in Australia ( [[#Hochman--2017|Hochman et al., 2017]] ; [[#Sultan--2019|Sultan et al., 2019]] based on detailed process-based modelling including a dedicated evaluation against observed yield fluctuations). These findings need additional support by independent studies. Results are relatively convergent that climate change has been an important driver of the recent declines in wheat yields in Europe ( ''medium confidence'' ) ( [[#Moore--2015|Moore and Lobell, 2015]] ; [[#Agnolucci--2019|Agnolucci and De Lipsis, 2019]] ; [[#Ray--2019|Ray et al., 2019]] ).

Due to complex interactions with socioeconomic conditions, climate-induced trends in crop yields and production do not directly transmit to crop prices, availability of food, or nutrition status. This complexity, in addition to the limited availability of long-term data, has so far impeded the detection and attribution of a long-term impact of climate change on associated food security indicators. However, in a few cases, observed crop prices (e.g., domestic grain price in Russia and Africa, [[#Götz--2016|Götz et al., 2016]] ; [[#Mawejje--2016|Mawejje, 2016]] ; [[#Baffes--2019|Baffes et al., 2019]] ) are shown to be sensitive to fluctuations in local weather through its impact on production (see ‘Food system—Food prices’, Table SM16.23). In addition, there is growing evidence that ''climate extremes'' (in particular, droughts) have led to malnutrition (in particular, stunting of children) in the historical period ( ''medium confidence'' , see ‘Food system—Malnutrition’, Table SM16.23) but without an attribution of changes to long-term climate change.

<div id="16.2.3.5" class="h3-container"></div>

<span id="temperature-related-mortality"></span>
==== 16.2.3.5 Temperature-Related Mortality ====

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

There is nearly universal evidence that non-optimal ambient temperatures increase mortality ( ''high confidence'' ), with notable heterogeneity only in the shape of the temperature–mortality relationship across geographical regions but often sharply growing relative risks at the outer 5% of the local historical temperature distributions ( [[#Gasparrini--2015|Gasparrini et al., 2015]] ; [[#Guo--2018|Guo et al., 2018]] ; [[#Carleton--2020|Carleton et al., 2020]] ; [[#Zhao--2021|Zhao et al., 2021]] ; see ‘Other societal impacts—Heat-related mortality’, Table SM16.23). Significant advances have been made since AR5 regarding the analysis of temperature-related excess mortality in previously under-researched regions, such as developing countries and (sub)tropical climates (e.g South-East Asia: [[#Dang--2016|Dang et al., 2016]] ; [[#Ingole--2017|Ingole et al., 2017]] ; [[#Mazdiyasni--2017|Mazdiyasni et al., 2017]] ; South Africa: [[#Wichmann--2017|Wichmann, 2017]] , [[#Scovronick--2018|Scovronick et al., 2018]] ; the Middle East: [[#Alahmad--2019|Alahmad et al., 2019]] , [[#Gholampour--2019|Gholampour et al., 2019]] ; and Latin America: [[#Péres--2020|Péres et al., 2020]] ). Progress has also been made with regard to temporal changes in temperature-related excess mortality and underlying population vulnerability over time. Heat-attributable mortality fractions have declined over time in most countries owing to general improvements in health care systems, increasing prevalence of residential air conditioning, and behavioural changes. These factors, which determine the susceptibility of the population to heat, have predominated over the influence of temperature change (see ‘Other societal impacts—Heat-related mortality’, Table SM16.22, [[#De’Donato--2015|De’Donato et al., 2015]] ; [[#Arbuthnott--2016|Arbuthnott et al., 2016]] ; [[#Vicedo-Cabrera--2018a|Vicedo-Cabrera et al., 2018a]] ). Important exceptions exist, for example, where unprecedented heatwaves have occurred recently. No conclusive evidence emerges regarding recent temporal trends in excess mortality attributable to cold exposure ( [[#Vicedo-Cabrera--2018b|Vicedo-Cabrera et al., 2018b]] ). Quantitative detection and attribution studies of temperature-related mortality are still rare. One study (Vicedo-Cabrera et al. 2021), using data from 43 countries, found that 37% (range 20.5–76.3%) of average warm-season heat-related mortality during recent decades can be attributed to anthropogenic climate change ( ''medium confidence'' , see ‘Other societal impacts—Heat-related mortality’, Table SM16.22). Studying excess mortality associated with past heatwaves, such as the 2003 or 2018 events in Europe, even higher proportions of deaths attributable to anthropogenic climate change have been reported for France and the UK ( [[#Mitchell--2016|Mitchell et al., 2016]] ; [[#Clarke--2021|Clarke et al., 2021]] ). Formal attribution studies encompassing cold-related mortality are quasi non-existent. The very few studies from Europe and Australia ( [[#Christidis--2010|Christidis et al., 2010]] ; [[#Åström--2013|Åström et al., 2013]] ; [[#Bennett--2014|Bennett et al., 2014]] ) find weak impacts of climate change on cold-associated excess mortality, with contradictory outcomes both towards higher and lower risks ( ''low confidence'' , see ‘Other societal impacts—Heat-related mortality’, Table SM16.22).

<div id="16.2.3.6" class="h3-container"></div>

<span id="waterborne-diseases"></span>
==== 16.2.3.6 Waterborne Diseases ====

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

Infectious diseases with water-associated transmission pathways constitute a large burden of disease globally. Since the AR5, the evidence has strengthened that waterborne diseases, and especially gastrointestinal infections, are highly to moderately sensitive to weather variability ( ''medium confidence'' , see ‘Water distribution—Waterborne diseases’, Table SM16.23). Increased temperature and high precipitation, with associated flooding events, have been shown to generally increase the risk of diarrhoeal diseases. There are, however, a number of studies that describe important exceptions and modifications to this general observation. While high temperatures favour bacterial diarrhoeal diseases, virally transmitted diarrhoea is on the contrary mostly associated with low temperatures ( [[#Carlton--2016|Carlton et al., 2016]] ; [[#Chua--2021|Chua et al., 2021]] ). Socioeconomic determinants, such as the existence of single-household water supplies ( [[#Herrador--2015|Herrador et al., 2015]] ) or combined sewer overflows ( [[#Jagai--2017|Jagai et al., 2017]] ), have been shown to critically increase the risk of gastrointestinal infections linked to heavy rainfall in high-income countries. Also, for both low- and high-income countries it has been found that gastrointestinal diseases increase following a heavy rainfall event only if preceded by a dry period ( [[#Carlton--2014|Carlton et al., 2014]] ; [[#Setty--2018|Setty et al., 2018]] ). Yet, so far there is no consistent evidence on the role of droughts in favouring waterborne disease transmission ( [[#Levy--2016|Levy et al., 2016]] ). As exemplified by the large cholera outbreak following the 2010 earthquake in Haiti, the existence of functioning sanitation systems is critical for preventing waterborne disease outbreaks, while climatic factors (especially rainfall) are important in driving the transmission dynamics once the outbreak has started ( [[#Rinaldo--2012|Rinaldo et al., 2012]] ). Other socioeconomic factors, such as human mobility and water management projects (e.g., dam constructions), also modify the strength of the association between climatic factors and waterborne diseases, as shown by recent studies in Africa ( [[#Perez-Saez--2015|Perez-Saez et al., 2015]] ; [[#Finger--2016|Finger et al., 2016]] ).

Whereas the weather sensitivity of waterborne diseases is well established for all world regions (see ‘Water distribution—Water-borne diseases’, Table SM16.23), studies attempting to attribute recent trends in waterborne disease to climate change are non-existent, except for investigations on the distribution of marine ''Vibrio'' bacteria and associated disease outbreaks in the coastal North Atlantic and the Baltic Sea regions ( [[#Baker-Austin--2013|Baker-Austin et al., 2013]] ; Baker- [[#Austin--2016|Austin et al., 2016]] ; [[#Vezzulli--2016|Vezzulli et al., 2016]] ; [[#Ebi--2017|Ebi et al., 2017]] ). These investigations provide evidence that increases in sea surface temperatures over recent decades as well as during recent summer heatwaves are linked to increased concentrations of ''Vibrio'' bacteria in coastal waters and an associated rise in environmentally acquired ''Vibrio'' infections in humans.

<div id="16.2.3.7" class="h3-container"></div>

<span id="vector-borne-diseases"></span>
==== 16.2.3.7 Vector-Borne Diseases ====

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

Vector-borne diseases constitute a large burden of infectious diseases worldwide and are highly sensitive to fluctuations of weather conditions including extreme events. Thus, both extreme rainfall and droughts have increased infections ( ''high confidence'' , see ‘Other societal impacts—Vector-borne diseases’, Table SM16.23). For example, in Sudan, anomalous high rainfall increased ''Anopheles'' mosquito breeding sites, leading to malaria outbreaks ( [[#Elsanousi--2018|Elsanousi et al., 2018]] ), while in Barbados and Brazil, drought conditions in urban areas have enhanced dengue incidence due to changes in water storage behaviour creating breeding sites for ''Aedes'' mosquitoes around human dwellings ( [[#Lowe--2018|Lowe et al., 2018]] ; [[#Lowe--2021|Lowe et al., 2021]] ) . In the Caribbean and Pacific Island nations, weather extremes, such as storms and flooding, have led to outbreaks of dengue due to disruption to water and sanitation services, leading to increased exposure to ''Aedes'' mosquito breeding sites ( [[#Descloux--2012|Descloux et al., 2012]] ; [[#Sharp--2014|Sharp et al., 2014]] ; [[#Uwishema--2021|Uwishema et al., 2021]] ). In South and Central America, and Asia, dengue incidence has been shown to be sensitive to variations in temperature and the monsoon season in addition to variations induced by urbanisation and population mobility ( ''high confidence'' [South and Central America] ''; medium confidence'' [Asia]; see ‘Other societal impacts—Vector-borne diseases’, Table SM16.23).

The attribution of changes in disease incidence to long-term climate change is often limited by relatively short reporting periods often only covering 10–15 years. Most studies then attribute trends in the occurrence of vector-borne diseases to the trends in climate across the same observational period and do not refer to an early ‘no climate change’ baseline climate. This means that they also capture trends induced by longerterm climate oscillations. Nevertheless, we list them in Table SM16.22 on ‘impact attribution’ to clearly distinguish them from the analysis of interannual fluctuations. The overall consistency of their findings across regions and time windows indicates that climate change is an important driver of the observed latitudinal or altitudinal range expansions of vector-borne diseases into previously colder areas ( ''medium'' to ''high confidence'' , see ‘Other societal impacts—Vector-borne diseases’, Table SM16.22). In highland areas of Africa and South America, epidemic outbreaks of malaria have become more frequent due to warming trends that allow ''Anopheles'' mosquitoes to persist at higher elevations ( [[#Pascual--2006|Pascual et al., 2006]] ; [[#Siraj--2014|Siraj et al., 2014]] ). In the USA, ticks that transmit Lyme disease have expanded their range northwards because of warmer temperatures ( ''high confidence'' ; [[#Kugeler--2015|Kugeler et al., 2015]] ; [[#McPherson--2017|McPherson et al., 2017]] ; [[#Lin--2019|Lin et al., 2019]] ; [[#Couper--2020|Couper et al., 2020]] ; see ‘Other societal impacts—Vector-borne diseases’, Table SM16.22). In Southern Europe, climate suitability for ''Aedes'' mosquitoes, which transmit dengue and chikungunya, and ''Culex'' mosquitoes, which transmit West Nile virus, has also increased and contributed to unprecedented outbreaks including the 2018 West Nile fever outbreak ( ''medium confidence'' , [[#Medlock--2013|Medlock et al., 2013]] ; [[#Paz--2013|Paz et al., 2013]] ; [[#Roiz--2015|Roiz et al., 2015]] ; ECDC, 2018, see ‘Other societal impacts—Vector-borne diseases’, Table SM16.22).

<div id="16.2.3.8" class="h3-container"></div>

<span id="economic-impacts"></span>
==== 16.2.3.8 Economic Impacts ====

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

Since the AR5, there has been significant progress regarding the identification of economic responses to weather fluctuations: evidence has increased that ''extreme weather events'' such as tropical cyclones, droughts, and severe fluvial floods have not only caused substantial immediate direct economic damage ( ''high confidence'' , see ‘Coastal Systems—Damages, Table SM16.23, ‘Water distribution—Reductions in water availability + induced damages and fatalities’, Table SM16.23, and ‘Water distribution—Flood-induced economic damages’, Table SM16.22) but have also reduced economic growth in the short term (year of, and year after event) ( [[#Strobl--2011|Strobl, 2011]] ; [[#Strobl--2012|Strobl, 2012]] ; [[#Fomby--2013|Fomby et al., 2013]] ; [[#Felbermayr--2014|Felbermayr and Gröschl, 2014]] , Loyaza et al. 2012) ( ''high confidence'' ) as well as in the long term (up to 10–15 years after event) ( ''medium confidence'' ) ( [[#Hsiang--2014|Hsiang and Jina, 2014]] ; [[#Berlemann--2016|Berlemann and Wenzel, 2016]] ; [[#Berlemann--2018|Berlemann and Wenzel, 2018]] ; [[#Krichene--2020|Krichene et al., 2020]] ; [[#Tanoue--2020|Tanoue et al., 2020]] , see ‘Other societal impacts—Macroeconomic output’, Table SM16.23). Short- and long-term reductions of economic growth by ''extreme weather events'' affect both developing and industrialised countries, but have been shown to be more severe in developing than in industrialised economies, thereby increasing inequality between countries ( ''high confidence'' , see ‘Other societal impacts—Between-country inequality’, Table SM16.23). Further, ''extreme weather events'' have increased within-country inequality since poorer people are more exposed and suffer relatively higher well-being losses than richer parts of the population ( ''medium confidence'' , see ‘Other societal impacts—Within-country inequality’, Table SM16.23). Going beyond ''extreme weather events'' , economic production depends nonlinearly on temperature fluctuations: below a certain threshold temperature, economic production increases with temperature, whereas it decreases above a certain threshold temperature ( ''high confidence'' ) ( [[#Burke--2015|Burke et al., 2015]] ; [[#Pretis--2018|Pretis et al., 2018]] ; [[#Kalkuhl--2020|Kalkuhl and Wenz, 2020]] ; [[#Kotz--2021|Kotz et al., 2021]] ).

So far, there are few individual studies attributing observed economic damages to long-term climate change except for damages induced by river flooding, droughts and tropical cyclones (see ‘Coastal systems—Damages’, ‘Water distribution—Flood-induced damages’, and ‘Water distribution—Reduction in water availability + induced damages and fatalities’, Table SM16.22). In addition, the empirical findings on the sensitivity of macroeconomic development to weather fluctuations and ''extreme weather events'' have been used to estimate the cumulative effect of historical warming on long-term economic development (see ‘Other societal impacts—Macroeconomic output’, Table SM16.22): anthropogenic climate change is estimated to have reduced gross domestic product (GDP) growth over the last 50 years, with substantially larger negative effects on developing countries and in some cases positive effects on colder industrialised countries ( ''low confidence'' ) ( [[#Diffenbaugh--2019|Diffenbaugh and Burke, 2019]] ). Globally, between-country inequality has decreased over the last 50 years. Climate change is estimated to have substantially slowed down this trend, that is, increased inequality compared with a counterfactual no-climate-change baseline ( ''low confidence'' ) ( [[#Diffenbaugh--2019|Diffenbaugh and Burke, 2019]] ). On a regional level, decreasing rainfall trends in Sub-Saharan Africa may have increased the GDP per capita gap between Sub-Saharan Africa and other developing countries ( ''low confidence'' ) ( [[#Barrios--2010|Barrios et al., 2010]] ). Overall, more research is needed on the impact channels through which ''extreme weather events'' and weather variability can hinder economic development, especially in the long term.

<div id="16.2.3.9" class="h3-container"></div>

<span id="social-conflict"></span>
==== 16.2.3.9 Social Conflict ====

<div id="h3-16-siblings" class="h3-siblings"></div>

There are few studies directly attributing changes in conflict risk to climate change in the modern era ( [[#van%20Weezel--2020|van Weezel, 2020]] ), preventing a confident assessment of the effect of long-term changes in the climate-related systems on armed conflict (see ‘Other societal impacts—Social conflict’, Table SM16.22). However, a sizeable literature links the prevalence of armed conflict within countries to within- and between-year variations in rainfall, temperature or drought exposure, often via reduced-form econometric analysis or statistical models that control for important non-climatic factors, such as agricultural dependence, level of economic development, state capacity and ethnopolitical marginalisation (see ‘Other societal impacts—Social conflict’, Table SM16.23). Overall, there is more consistent evidence that climate variability has influenced low-intensity organised violence than major civil wars ( [[#Detges--2017|Detges, 2017]] ; [[#Nordkvelle--2017|Nordkvelle et al., 2017]] ; [[#Linke--2018|Linke et al., 2018]] ). Likewise, there is more consistent evidence that climate variability has affected dynamics of conflict, such as continuation, severity and frequency of violent conflict events, than the likelihood of initial conflict outbreak ( [[#Yeeles--2015|Yeeles, 2015]] ; [[#Eastin--2016|Eastin, 2016]] ; [[#Von%20Uexkull--2016|Von Uexkull et al., 2016]] , [[IPCC:Wg2:Chapter:Chapter-7#7.2.7|Section 7.2.7]] ). Moreover, research suggests with ''medium confidence'' ( ''medium evidence'' , ''medium agreement'' ) that weather effects on armed conflict have been most prominent in contexts marked by a large population, low socioeconomic development, high political marginalisation and high agricultural dependence ( [[#Theisen--2017|Theisen, 2017]] ; [[#Koubi--2019|Koubi, 2019]] ; [[#Buhaug--2020|Buhaug et al., 2020]] ; [[#Ide--2020|Ide et al., 2020]] ).

Some studies also seek to evaluate potential indirect links between climate and weather anomalies and prevalence of armed conflict via food price shocks or forced migration. While there is ''robust evidence'' that the likelihood of social unrest in the developing world generally increases in response to rapid growth in food prices ( [[#Bellemare--2015|Bellemare, 2015]] ; [[#Rudolfsen--2018|Rudolfsen, 2018]] ), the magnitude of the climate effect on unrest via food prices is less well established ( [[#Martin-Shields--2019|Martin-Shields and Stojetz, 2019]] ). Similarly, research shows with ''high confidence'' that climate variability and extremes have affected human mobility (see ‘Other societal impacts—Displacement and migration’, Table SM16.23), but there is ''low agreement'' and ''limited evidence'' that weather-induced migration has increased the likelihood of armed conflict ( [[IPCC:Wg2:Chapter:Chapter-7#7.2.7|Section 7.2.7]] , [[#Brzoska--2016|Brzoska and]] [[#Fröhlich--2016|Fröhlich, 2016]] ; [[#Kelley--2017|Kelley et al., 2017]] ; [[#Selby--2017|Selby et al., 2017]] ; [[#Abel--2019|Abel, 2019]] ). Research on weather-related effects on interstate security generally concludes that periods of transboundary water scarcity are more likely to facilitate increased international cooperation than conflict ( [[#Bernauer--2020|Bernauer and Böhmelt, 2020]] ).

In general, the historical influence of climate on conflict is judged to be small when compared with dominant conflict drivers ( [[#Mach--2019|Mach et al., 2019]] ). Much of this research is limited to (parts of) Sub-Saharan Africa, which raises some concerns about selection bias and generalisability of results ( [[#Adams--2018|Adams et al., 2018]] ).

<div id="16.2.3.10" class="h3-container"></div>

<span id="displacement-and-migration"></span>
==== 16.2.3.10 Displacement and Migration ====

<div id="h3-17-siblings" class="h3-siblings"></div>

Given the complexity of human migration processes and decisions (e.g., [[#Boas--2019|Boas et al., 2019]] , [[#Cattaneo--2019|Cattaneo et al., 2019]] ) and the paucity of long-term, reliable and internally consistent observational data on displacement ( [[#IDMC--2019|IDMC, 2019]] ; [[#IDMC--2020|IDMC, 2020]] ) and migration ( [[#Laczko--2016|Laczko, 2016]] ), the contribution of long-term changes in climate-related systems to observed human displacement or migration patterns has not been quantified so far, except for individual examples of displacement induced by inland flooding where the heavy precipitation has been attributed to anthropogenic climate forcing and coastal flooding (see ‘Other societal impacts—Displacement and migration’, Table SM16.22; Section CCP2).

However, new evidence has emerged since the AR5 that further documents widespread effects of weather fluctuations and extreme events on migration (see ‘Other societal impacts—Displacement and migration’, Table SM16.23). Numerous studies find significant links between temperature or precipitation anomalies, or ''extreme weather events'' such as storms or floods, and internal as well as international migration ( [[#Coniglio--2015|Coniglio and Pesce, 2015]] ; [[#Cattaneo--2016|Cattaneo and Peri, 2016]] ; [[#Nawrotzki--2016|Nawrotzki and DeWaard, 2016]] ; [[#Beine--2017|Beine and Parsons, 2017]] , for international migration; and [[#IDMC--2019|IDMC, 2019]] , for internal displacement). Internal displacement of millions of people every year is triggered by natural hazards, mainly floods and storms ( [[#IDMC--2019|IDMC, 2019]] ). The effects of weather fluctuations and extremes on migration are considered more important for temporary mobility and displacement than permanent migration, and more influential on short-distance movement, including urbanisation, than international migration ( [[#McLeman--2014|McLeman, 2014]] ; [[#Hauer--2020|Hauer et al., 2020]] ; [[#Hoffmann--2020|Hoffmann et al., 2020]] , [[IPCC:Wg2:Chapter:Chapter-7#7.2.6|Section 7.2.6]] ). Importantly, these links are conditional on the socioeconomic situation in the origin; for example, poor populations may be ‘trapped’ and not able to migrate in the face of adverse climate or weather conditions ( [[#Black--2013|Black et al., 2013]] ; [[#Adams--2016|Adams, 2016]] ). Many studies have also explored the channels through which climate or weather influence migration, and have identified incomes in the agricultural sector as one of the main channels ( [[#Nawrotzki--2015|Nawrotzki et al., 2015]] ; [[#Viswanathan--2015|Viswanathan and Kavi Kumar, 2015]] ; [[#Cai--2016a|Cai et al., 2016a]] ). In particular, declines in agricultural incomes and employment due to changed weather variability may foster increased rural–urban movement, and the resulting pressures on urban wages in turn foster international migration ( [[#Marchiori--2012|Marchiori et al., 2012]] ; [[#Maurel--2016|Maurel and Tuccio, 2016]] ). Another possible but controversial channel is violent conflict, which may be fostered (though not exclusively caused) by adverse climate conditions such as drought, and in turn lead to people seeking refugee status, although evidence of such an indirect effect is weak ( [[#Brzoska--2016|Brzoska and]] [[#Fröhlich--2016|Fröhlich, 2016]] ; [[#Abel--2019|Abel et al., 2019]] ; [[#Schutte--2021|Schutte et al., 2021]] ).

<div id="16.3" class="h1-container"></div>

<span id="synthesis-of-observed-adaptation-related-responses"></span>