Editing IPCC:AR6/WGI/Chapter-11 (section)

== 11.4 Heavy Precipitation ==

<div id="h1-5-siblings" class="h1-siblings"></div>

This section assesses changes in heavy precipitation at global and regional scales. The main focus is on extreme precipitation at a daily scale where literature is most concentrated, though extremes of shorter (sub-daily) and longer (five-day or more) durations are also assessed to the extent the literature allows.

<div id="11.4.1" class="h2-container"></div>

<span id="mechanisms-and-drivers-1"></span>
=== 11.4.1 Mechanisms and Drivers ===

<div id="h2-29-siblings" class="h2-siblings"></div>

The SREX (Chapter 3, [[#Seneviratne--2012|Seneviratne et al., 2012]] ) assessed changes in heavy precipitation in the context of the effects of thermodynamic and dynamic changes. Box 11.1 assesses thermodynamic and dynamic changes in a warming world to aid the understanding of changes in observations and projections in some extremes and the sources of uncertainties (see also [[IPCC:Wg1:Chapter:Chapter-8#8.2.3.2|Section 8.2.3.2]] ). In general, warming increases the atmospheric water-holding capacity following the Clausius–Clapeyron (C-C) relation. This thermodynamic effect results in an increase in extreme precipitation at a similar rate at the global scale. On a regional scale, changes in extreme precipitation are further modulated by dynamic changes (Box 11.1).

Large-scale modes of variability, such as the North Atlantic Oscillation (NAO), El Niño–Southern Oscillation (ENSO), Atlantic Multi-decadal Variability (AMV), and Pacific Decadal Variability (PDV) (Annex IV), modulate precipitation extremes through changes in environmental conditions or embedded storms ( [[IPCC:Wg1:Chapter:Chapter-8#8.3.2|Section 8.3.2]] ). Latent heating can invigorate these storms ( [[#Nie--2018|Nie et al., 2018]] ; Z. [[#Zhang--2019|]] [[#Zhang--2019|]] [[#Zhang--2019|]] [[#Zhang--2019|Zhang et al., 2019]] a); changes in dynamics can increase precipitation intensity above that expected from the C-C scaling rate (Sections 8.2.3.2 and 11.7; Box 11.1). Additionally, the efficiency of converting atmospheric moisture into precipitation can change as a result of cloud microphysical adjustment to warming,resulting in changes in the characteristics of extreme precipitation; but changes in precipitation efficiency in a warming world are highly uncertain ( [[#Sui--2020|Sui et al., 2020]] ).

It is difficult to separate the effect of global warming from internal variability inthe observed changes in the modes of variability ( [[IPCC:Wg1:Chapter:Chapter-2#2.4|Section 2.4]] ). Future projections of modes of variability are highly uncertain [[IPCC:Wg1:Chapter:Chapter-4#4.3.3|Section 4.3.3]] ),resulting in uncertainty in regional projections of extreme precipitation. Future warming may amplify monsoonal extreme precipitation. Changes in extreme storms, including tropical/extratropical cyclones and severe convective storms, result in changes in extreme precipitation ( [[#11.7|Section 11.7]] ). Also, changes in sea surface temperatures (SSTs) alter land–sea contrast, leading to changes in precipitation extremes near coastal regions. For example, the projected larger SST increase near the coasts of East Asia and India can result in heavier rainfall near these coastal areas from tropical cyclones ( [[#Mei--2016|Mei and Xie, 2016]] ) or torrential rains ( [[#Manda--2014|Manda et al., 2014]] ). The warming in the western Indian Ocean is associated with increases in moisture surges on the low-level monsoon westerlies towards the Indian subcontinent, which may lead to an increase in the occurrence of precipitation extremes over central India ( [[#Krishnan--2016|Krishnan et al., 2016]] ; [[#Roxy--2017|Roxy et al., 2017]] ).

Decreases in atmospheric aerosols results in warming and thus an increase in extreme precipitation ( [[#Samset--2018|Samset et al., 2018]] ; [[#Sillmann--2019|Sillmann et al., 2019]] ). Changes in atmospheric aerosols also result in dynamic changes such as in tropical cyclones ( [[#Takahashi--2017|Takahashi et al., 2017]] ; [[#Strong--2018|Strong et al., 2018]] ). Uncertainty in the projections of future aerosol emissions results in additional uncertainty in the heavy precipitation projections of the 21st century ( [[#Lin--2016|Lin et al., 2016]] ).

There has been new evidence of the effect of local land-use and land-cover change on heavy precipitation. There is a growing set of literature linking increases in heavy precipitation in urban centres to urbanization ( [[#Argüeso--2016|Argüeso et al., 2016]] ; Y. [[#Zhang--2019|]] [[#Zhang--2019|]] [[#Zhang--2019|]] [[#Zhang--2019|Zhang et al., 2019]] b). Urbanization intensifies extreme precipitation, especially in the afternoon and early evening, over the urban area and its downwind region ( ''medium confidence'' ) (Box 10.3). There are four possible mechanisms: (i) increases in atmospheric moisture due to horizontal convergence of air associated with the urban heat island effect ( [[#Shastri--2015|Shastri et al., 2015]] ; [[#Argüeso--2016|Argüeso et al., 2016]] ); (ii) increases in condensation due to urban aerosol emissions ( [[#Han--2011|Han et al., 2011]] ; [[#Sarangi--2017|Sarangi et al., 2017]] ); (iii) aerosol pollution that impacts cloud microphysics (Box 8.1; [[#Schmid--2017|Schmid and Niyogi, 2017]] ); and (iv) urban structures that impede atmospheric motion (Shepherd, 2013; [[#Ganeshan--2015|Ganeshan and Murtugudde, 2015]] ; [[#Paul--2018|Paul et al., 2018]] ). Other local forcing, including reservoirs ( [[#Woldemichael--2012|Woldemichael et al., 2012]] ), irrigation ( [[#Devanand--2019|Devanand et al., 2019]] ), or large-scale land-use and land-cover change ( [[#Odoulami--2019|Odoulami et al., 2019]] ), can also affect local extreme precipitation.

In summary, precipitation extremes are controlled by both thermodynamic and dynamic processes. Warming-induced thermodynamic change results in an increase in extreme precipitation, at a rate that closely follows the C-C relationship at the global scale ( ''high confidence'' ). The effects of warming-induced changes in dynamic drivers on extreme precipitation are more complicated, difficult to quantify, and are an uncertain aspect of projections. Precipitation extremes are also affected by forcings other than changes in greenhouse gases, including changes in aerosols, land-use and land-cover change, and urbanization ( ''mediu'' ''m confidence'' ).

<div id="11.4.2" class="h2-container"></div>

<span id="observed-trends-1"></span>
=== 11.4.2 Observed Trends ===

<div id="h2-30-siblings" class="h2-siblings"></div>

Both SREX (Chapter 3, [[#Seneviratne--2012|Seneviratne et al., 2012]] ) and AR5 (IPCC, 2014 Chapter 2) concluded it was ''likely'' that the number of heavy precipitation events over land had increased in more regions than it had decreased, though there were wide regional and seasonal variations, and trends in many locations were not statistically significant. This assessment has been strengthened with multiple studies finding ''robust evidence'' of the intensification of extreme precipitation at global and continental scales, regardless of spatial and temporal coverage of observations and the methods of data processing and analysis.

The average annual maximum precipitation amount in a day (Rx1day) has significantly increased since the mid-20th century over land ( [[#Du--2019|Du et al., 2019]] ; [[#Dunn--2020|Dunn et al., 2020]] ) and in the humid and dry regions of the globe ( [[#Dunn--2020|Dunn et al., 2020]] ). The percentage of observing stations with statistically significant increases in Rx1day is larger than expected by chance, while the percentage of stations with statistically significant decreases is smaller than expected by chance, over the global land as a whole and over North America, Europe, and Asia (Figure 11.13; [[#Sun--2021|Sun et al., 2021]] ) and over global monsoon regions ( [[#Zhang--2019|Zhang and Zhou, 2019]] ) where data coverage is relatively good. The addition of the past decade of observational data shows a more robust increase in Rx1day over the global land region ( [[#Sun--2021|Sun et al., 2021]] ). Light, moderate, and heavy daily precipitation has all intensified in a gridded daily precipitation dataset ( [[#Contractor--2020a|Contractor et al., 2020a]] ). Daily mean precipitation intensities have increased since the mid-20th century in a majority of land regions ( ''high confidence'' ) ( [[IPCC:Wg1:Chapter:Chapter-8#8.3.1.3|Section 8.3.1.3]] ). The probability of precipitation exceeding 50 mm/day increased during 1961–2018 ( [[#Benestad--2019|Benestad et al., 2019]] ). The globally averaged annual fraction of precipitation from days in the top 5% (R95pTOT) has also significantly increased ( [[#Dunn--2020|Dunn et al., 2020]] ). The increase in the magnitude of Rx1day in the 20th century is estimated to be at a rate consistent with C-C scaling with respect to global mean temperature ( [[#Fischer--2016|Fischer and Knutti, 2016]] ; [[#Sun--2021|Sun et al., 2021]] ). Studies on past changes in extreme precipitation of durations longer than a day are more limited, though there are some studies examining long-term trends in annual maximum five-day precipitation (Rx5day). On global and continental scales, long-term changes in Rx5day are similar to those of Rx1day in many aspects (Zhang and Zhou 2019; [[#Sun--2021|Sun et al., 2021]] ). As discussed below, at the regional scale, changes in Rx5day are also similar to those of Rx1day where there are analyses of changes in both Rx1day and Rx5day.

<div id="_idContainer055" class="Basic-Text-Frame"></div>

[[File:c58ed0d3631d679741c575dba07df416 IPCC_AR6_WGI_Figure_11_13.png]]

'''Figure 11.13 |''' '''Signs and significance of the observed trends in annual maximum daily precipitation (Rx1day) during 1950–2018 at 8345 stations with suficient data.''' ''(a)'' Percentage of stations with statistically significant trends in Rx1day; green dots show positive trends and brown dots negative trends. Box and ‘whisker’ plots indicate the expected percentage of stations with significant trends due to chance estimated from 1000 bootstrap realizations under a no-trend null hypothesis. The boxes mark the median, 25th percentile, and 75th percentile. The upper and lower whiskers show the 97.5th and the 2.5th percentiles, respectively. Maps of stations with positive ''(b)'' and negative ''(c)'' trends. The light colour indicates stations with non-significant trends, and the dark colour stations with significant trends. Significance is determined by a two-tailed test conducted at the 5% level. Adapted from [[#Sun--2021|Sun et al. (2021)]] . Figure copyright © American Meteorological Society (used with permission). Further details on data sources and processing are available in the chapter data table (Table 11.SM.9).

Overall, there is a lack of systematic analysis of long-term trends in sub-daily extreme precipitation at the global scale. Often, sub-daily precipitation data have only sporadic spatial coverage and are of limited length. Additionally, the available data records are far shorter than needed for a robust quantification of past changes in sub-daily extreme precipitation ( [[#Li--2019a|]] [[#Li--2019|]] [[#Li--2019|]] [[#Li--2019|C. Li et al., 2019]] a ). Despite these limitations, there are studies in regions of almost all continents that generally indicate intensification of sub-daily extreme precipitation, although there remains ''low'' ''confidence'' in an overall increase at the global scale. Studies include an increase in extreme sub-daily rainfall in summer over South Africa ( [[#Sen%20Roy--2013|Sen Roy and Rouault, 2013]] ), annually in Australia ( [[#Guerreiro--2018b|Guerreiro et al., 2018b]] ), over 23 urban locations in India ( [[#Ali--2018|Ali and Mishra, 2018]] ), in Peninsular Malaysia ( [[#Syafrina--2015|Syafrina et al., 2015]] ), and in eastern China in the summer season during 1971–2013 ( [[#Xiao--2016|Xiao et al., 2016]] ). In some regions in Italy ( [[#Arnone--2013|Arnone et al., 2013]] ; [[#Libertino--2019|Libertino et al., 2019]] ) and in the USA during 1950–2011 ( [[#Barbero--2017|Barbero et al., 2017]] ), there is also an increase. In general, an increase in sub-daily heavy precipitation results in an increase in pluvial floods over smaller watersheds ( [[#Ghausi--2020|Ghausi and Ghosh, 2020]] ).

There is a considerable body of literature examining scaling of sub-daily precipitation extremes, conditional on day-to-day air or dew-point temperatures ( [[#Westra--2014|Westra et al., 2014]] ; [[#Fowler--2021|Fowler et al., 2021]] ). This scaling, also termed ‘apparent scaling’ (Fowler et al., 2021), is robust when different methodologies are used in different regions, ranging between the C-C and two-times the C-C rate (e.g., [[#Formayer--2017|Formayer and Fritz, 2017]] ; [[#Lenderink--2017|Lenderink et al., 2017]] ; [[#Burdanowitz--2019|Burdanowitz et al., 2019]] ). This is confirmed when sub-daily precipitation data collected from multiple continents ( [[#Lewis--2019|Lewis et al., 2019]] ) are analysed in a consistent manner using different methods ( [[#Ali--2021|Ali et al., 2021]] ). It has been hoped that apparent scaling might be used to help understand past and future changes in extreme sub-daily precipitation. However, apparent scaling samples multiple synoptic weather states, mixing thermodynamic and dynamic factors that are not directly relevant for climate change responses ( [[IPCC:Wg1:Chapter:Chapter-8#8.2.3.2|Section 8.2.3.2]] ; [[#Prein--2016b|Prein et al., 2016b]] ; [[#Bao--2017|Bao et al., 2017]] ; X. [[#Zhang--2017|]] [[#Zhang--2017|]] [[#Zhang--2017|Zhang et al., 2017]] ; [[#Drobinski--2018|Drobinski et al., 2018]] ; [[#Sun--2020|Sun et al., 2020]] ). The spatial pattern of apparent scaling is different from those of projected changes over Australia ( [[#Bao--2017|Bao et al., 2017]] ) and North America (Sun et al., 2020) in regional climate model simulations. It thus remains difficult to use the knowledge about apparent scaling to infer past and future changes in extreme sub-daily precipitation according to observed and projected changes in local temperature.

In Africa (Table 11.5), evidence shows an increase in extreme daily precipitation for the late half of the 20th century over the continent where data are available; there is a larger percentage of stations showing significant increases in extreme daily precipitation than decreases ( [[#Sun--2021|Sun et al., 2021]] ). There are increases in different metrics relevant to extreme precipitation in various regions of the continent ( [[#Chaney--2014|Chaney et al., 2014]] ; [[#Harrison--2019|Harrison et al., 2019]] ; [[#Dunn--2020|Dunn et al., 2020]] ; [[#Sun--2021|Sun et al., 2021]] ). There is an increase in extreme precipitation events in Southern Africa (Weldon and Reason, 2014; [[#Kruger--2019|Kruger et al., 2019]] ) and a general increase in heavy precipitation over East Africa, the Greater Horn of Africa ( [[#Omondi--2014|Omondi et al., 2014]] ). Over sub-Saharan Africa, increases in the frequency and intensity of extreme precipitation have been observed over the well-gauged areas during 1950–2013; however, this covers only 15% of the total area of sub-Saharan Africa ( [[#Harrison--2019|Harrison et al., 2019]] ). There is ''medium'' ''confidence'' about the increase in extreme precipitation for some regions where observations are more abundant '','' but for Africa as whole, there is ''low confidence'' because of a general lack of continent-wide systematic analysis, the sporadic nature of available precipitation data over the continent, and spatially non-homogenous trends in places where dataare available (Donat et al., 2014a; [[#Mathbout--2018b|Mathbout et al., 2018b]] ; [[#Alexander--2019|Alexander et al., 2019]] ; [[#Funk--2020|Funk et al., 2020]] ).

In Asia (Table 11.8), there is ''robust evidence'' that extreme precipitation has increased since the 1950s ( ''high confidence'' ), however, this is dominated by high spatial variability. Increases in Rx1day and Rx5day during 1950–2018 are found over two-thirds of stations. The percentage of stations with statistically significant trends is larger than can be expected by chance (Figure 11.13; [[#Sun--2021|Sun et al., 2021]] ). An increase in extreme precipitation has also been observed in various regional studies based on different metrics of extreme precipitation and spatial and temporal coverage of the data. These include an increase in daily precipitation extremes over central Asia ( [[#Hu--2016|Hu et al., 2016]] ), most of South Asia ( [[#Zahid--2012|Zahid and Rasul, 2012]] ; [[#Pai--2015|Pai et al., 2015]] ; [[#Sheikh--2015|Sheikh et al., 2015]] ; [[#Adnan--2016|Adnan et al., 2016]] ; [[#Malik--2016|Malik et al., 2016]] ; [[#Dimri--2017|Dimri et al., 2017]] ; [[#Priya--2017|Priya et al., 2017]] ; [[#Roxy--2017|Roxy et al., 2017]] ; [[#Hunt--2018|Hunt et al., 2018]] ; [[#Kim--2019|Kim et al., 2019]] ; [[#Wester--2019|Wester et al., 2019]] ), the Arabian Peninsula ( [[#Rahimi--2019|Rahimi and Fatemi, 2019]] ; [[#Almazroui--2020|Almazroui and Saeed, 2020]] ; [[#Atif--2020|Atif et al., 2020]] ), South East Asia ( [[#Siswanto--2015|Siswanto et al., 2015]] ; [[#Supari--2017|Supari et al., 2017]] ; [[#Cheong--2018|Cheong et al., 2018]] ); the north-west Himalaya ( [[#Malik--2016|Malik et al., 2016]] ), parts of East Asia (Baeket al., 2017; [[#Nayak--2017|Nayak et al., 2017]] ; [[#Ye--2017|Ye and Li, 2017]] ), the western Himalayas since the 1950s ( [[#Ridley--2013|Ridley et al., 2013]] ; [[#Dimri--2015|Dimri et al., 2015]] ; [[#Madhura--2015|Madhura et al., 2015]] ), West and East Siberia, and Russian Far East ( [[#Donat--2016a|Donat et al., 2016a]] ). A decrease was found over the eastern Himalayas ( [[#Sheikh--2015|Sheikh et al., 2015]] ; [[#Talchabhadel--2018|Talchabhadel et al., 2018]] ). Increases have been observed over Jakarta ( [[#Siswanto--2015|Siswanto et al., 2015]] ), but Rx1day over most parts of the Maritime Continent has decreased ( [[#Villafuerte--2015|Villafuerte and Matsumoto, 2015]] ). Trends in extreme precipitation over China are mixed with increases and decreases (G. [[#Fu--2013|]] [[#Fu--2013|Fu et al., 2013]] ; [[#Jiang--2013|Jiang et al., 2013]] ; [[#Ma--2015|Ma et al., 2015]] ; [[#Yin--2015|Yin et al., 2015]] ; [[#Xiao--2016|Xiao et al., 2016]] ) and are not significant over China as whole ( [[#Jiang--2013|Jiang et al., 2013]] ; [[#Hu--2016|Hu et al., 2016]] ; [[#Ge--2017|Ge et al., 2017]] ; [[#Deng--2018|Deng et al., 2018]] ; [[#He--2018|He and Zhai, 2018]] ; W. [[#Li--2018|]] [[#Li--2018|]] [[#Li--2018|]] [[#Li--2018|Li et al., 2018]] a; [[#Tao--2018|Tao et al., 2018]] ; M. [[#Liu--2019|Liu et al., 2019]] b; [[#Chen--2021|Chen et al., 2021]] ). With few exceptions, most South East Asian countries have experienced an increase in rainfall intensity, but with a reduced number of wet days ( [[#Donat--2016a|Donat et al., 2016a]] ; [[#Cheong--2018|Cheong et al., 2018]] ; [[#Naveendrakumar--2019|Naveendrakumar et al., 2019]] ), though large differences in trends exists if the trends are estimated from different datasets, including gauge-based, remotely sensed, and reanalysis data, over a relatively short period ( [[#Kim--2019|Kim et al., 2019]] ). There is a significant increase in heavy rainfall (&gt;100 mm day <sup>–1</sup> ) and a significant decrease in moderate rainfall (5–100 mm day <sup>–1</sup> ) in central India during the South Asian monsoon season ( [[#Deshpande--2016|Deshpande et al., 2016]] ; [[#Roxy--2017|Roxy et al., 2017]] ).

In Australasia (Table 11.11), available evidence has not shown an increase or a decrease in heavy precipitation over Australasia as a whole ( ''medium confidence'' ), but heavy precipitation tends to increase over Northern Australia (particularly the north-west) and decrease over the eastern and southernregions (e.g., Jakob and Walland, 2016; [[#Guerreiro--2018b|Guerreiro et al., 2018b]] ; [[#Dey--2019b|Dey et al., 2019b]] ; [[#Dunn--2020|Dunn et al., 2020]] ; [[#Sun--2021|Sun et al., 2021]] ). Available studies that used long-term observations since the mid-20th century showed nearly as many stations with an increase as those with a decrease in heavy precipitation ( [[#Jakob--2016|Jakob and Walland, 2016]] ) or slightly more stations with a decrease than with an increase in Rx1day and Rx5day ( [[#Sun--2021|Sun et al., 2021]] ), or strong differences in Rx1day trends with increases over Northern Australia and Central Australia in general, but mostly decreases over Southern Australia and Eastern Australia ( [[#Dunn--2020|Dunn et al., 2020]] ). Over New Zealand, decreases are observed for moderate–heavy precipitation events, but there are no significant trends for very heavy events (more than 64 mm in a day) for the period 1951–2012. The number of stations with an increase in very wet days is similar to that with a decrease during 1960–2019 (MfE and Stats NZ, 2020). Overall, there is ''low confidence'' in trends in the frequency of heavy rain days, with mostly decreases over New Zealand (Harringtonand Renwick, 2014; [[#Caloiero--2015|Caloiero, 2015]] ).

In Central and South America (Table 11.14), evidence shows an increase in extreme precipitation, but in general there is ''low'' ''confidence;'' while continent-wide analyses produced wetting trends are not robust. Rx1day increased at more stations than it decreased in South America between 1950 and 2018 ( [[#Sun--2021|Sun et al., 2021]] ). Over the period 1950–2010, both Rx5day and R99p increased over large regions of South America, including North-Western South America, Northern South America, and South-Eastern South America ( [[#Skansi--2013|Skansi et al., 2013]] ). There are large regional differences. A decrease in daily extreme precipitation is observed in north-eastern Brazil (Skansi et al., 2013; [[#Bezerra--2018|Bezerra et al., 2018]] ; [[#Dereczynski--2020|Dereczynski et al., 2020]] ). Trends in extreme precipitation indices were not statistically significant over the period 1947–2012 within the São Francisco River basin in the Brazilian semi-arid region ( [[#Bezerra--2018|Bezerra et al., 2018]] ). An increase in extreme rainfall is observed in the Amazon with ''medium confidence'' ( [[#Skansi--2013|Skansi et al., 2013]] ) and in South-Eastern South America with ''high confidence'' ( [[#Skansi--2013|Skansi et al., 2013]] ; [[#Valverde--2014|Valverde and Marengo, 2014]] ; [[#Barros--2015|Barros et al., 2015]] ; [[#Ávila--2016|Ávila et al., 2016]] ; [[#Wu--2017|Wu and Polvani, 2017]] ; [[#Lovino--2018|Lovino et al., 2018]] ; [[#Dereczynski--2020|Dereczynski et al., 2020]] ). Among all sub-regions, South-Eastern South America shows the highest rate of increase for rainfall extremes, followed by the Amazon ( [[#Skansi--2013|Skansi et al., 2013]] ). Increases in the intensity of heavy daily rainfall events have been observed in the southern Pacific and in the Titicaca basin ( [[#Skansi--2013|Skansi et al., 2013]] ; Huerta and Lavado‐Casimiro, 2021). In Southern Central America, trends in annual precipitation are generally not significant, although small (but significant) increases are found in Guatemala, El Salvador, and Panama ( [[#Hidalgo--2017|Hidalgo et al., 2017]] ). Small positive trends were found in multiple extreme precipitation indices over the Caribbean region over a short time period (1986–2010) ( [[#Stephenson--2014|Stephenson et al., 2014]] ; [[#McLean--2015|McLean et al., 2015]] ).

In Europe (Table 11.17), there is ''robust evidence'' that the magnitude and intensity of extreme precipitation has ''very likely'' increased since the 1950s. There is a significant increase in Rx1day and Rx5day during 1950–2018 in Europe as a whole ( [[#Sun--2021|Sun et al., 2021]] , also Figure 11.13). The number of stations with increases far exceeds those with decreases in the frequency of daily rainfall exceeding its 90th or 95th percentile in century-long series ( [[#Cioffi--2015|Cioffi et al., 2015]] ). The five-, 10-, and 20-year events of one-day and five-day precipitation during 1951–1960 became more common since the 1950s ( [[#van%20den%20Besselaar--2013|van den Besselaar et al., 2013]] ). There can be large discrepancies among studies and regions and seasons ( [[#Croitoru--2013|Croitoru et al., 2013]] ; [[#Willems--2013|Willems, 2013]] ; [[#Casanueva--2014|Casanueva et al., 2014]] ; [[#Roth--2014|Roth et al., 2014]] ; [[#Fischer--2015|Fischer et al., 2015]] ); evidence for increasing extreme precipitation is more frequently observed for summer and winter, but not in other seasons ( [[#Madsen--2014|Madsen et al., 2014]] ; [[#Helama--2018|Helama et al., 2018]] ) ''.'' An increase is observed in central Europe ( [[#Volosciuk--2016|Volosciuk et al., 2016]] ; [[#Zeder--2020|Zeder and Fischer, 2020]] ), and in Romania ( [[#Croitoru--2016|Croitoru et al., 2016]] ). Trends in the Mediterranean region are in general not spatially consistent ( [[#Reale--2013|Reale and Lionello, 2013]] ), with decreases in the western Mediterranean and some increases in the eastern Mediterranean ( [[#Rajczak--2013|Rajczak et al., 2013]] ; [[#Casanueva--2014|Casanueva et al., 2014]] ; [[#de%20Lima--2015|de Lima et al., 2015]] ; [[#Gajić-Čapka--2015|Gajić-Čapka et al., 2015]] ; [[#Sunyer--2015|Sunyer et al., 2015]] ; [[#Pedron--2017|Pedron et al., 2017]] ; [[#Serrano-Notivoli--2018|Serrano-Notivoli et al., 2018]] ; [[#Ribes--2019|Ribes et al., 2019]] ). In the Netherlands, the total precipitation contributed from extremes higher than the 99th percentile doubles per 1°C increase in warming ( [[#Myhre--2019|Myhre et al., 2019]] ), though extreme rainfall trends in Northern Europe may differ in different seasons ( [[#Irannezhad--2017|Irannezhad et al., 2017]] ).

In North America (Table 11.20), there is ''robust evidence'' that the magnitude and intensity of extreme precipitation has ''very likely'' increased since the 1950s. Both Rx1day and Rx5day have significantly increased in North America during 1950–2018 ( [[#Sun--2021|Sun et al., 2021]] , also Figure 11.13). There is, however, regional diversity. In Canada, there is a lack of detectable trends in observed annual maximum daily (or shorter duration) precipitation ( [[#Shephard--2014|Shephard et al., 2014]] ; [[#Mekis--2015|Mekis et al., 2015]] ; [[#Vincent--2018|Vincent et al., 2018]] ). In the USA, there is an overall increase in one-day heavy precipitation, both in terms of intensity and frequency (Villarini et al.,2012; [[#Donat--2013b|Donat et al., 2013b]] ; [[#Wu--2015|Wu, 2015]] ; Easterling et al., 2017; H. [[#Huang--2017|Huang et al., 2017]] ; [[#Howarth--2019|Howarth et al., 2019]] ; [[#Sun--2021|Sun et al., 2021]] ), except for the southern USA ( [[#Hoerling--2016|Hoerling et al., 2016]] ) where internal variability may have played a substantial role in the lack of observed increases. In Mexico, increases are observed in R10mm and R95p ( [[#Donat--2016a|Donat et al., 2016a]] ), very wet days over the cities ( [[#García-Cueto--2019|García-Cueto et al., 2019]] ) and in total precipitation (PRCPTOT) and Rx1day ( [[#Donat--2016b|Donat et al., 2016b]] ).

In Small Islands, there is a lack of evidence showing changes in heavy precipitation overall. There were increases in extreme precipitation in Tobago from 1985–2015 ( [[#Stephenson--2014|Stephenson et al., 2014]] ; [[#Dookie--2019|Dookie et al., 2019]] ) and decreases in south-western French Polynesia and the southern subtropics ( ''low confidence'' ) (Table 11.5; Atlas.10). Extreme precipitation leading to flooding in the Small Islands has been attributed in part to tropical cyclones, as well as being influenced by ENSO (Box 11.5; [[#Khouakhi--2016|Khouakhi et al., 2016]] ; [[#Hoegh-Guldberg--2018|Hoegh-Guldberg et al., 2018]] ).

In summary, the frequency and intensity of heavy precipitation have ''likely'' increased at the global scale over a majority of land regions with good observational coverage. Since 1950, the annual maximum amount of precipitation falling in a day, or over five consecutive days, has ''likely'' increased over land regions with sufficient observational coverage for assessment, with increases in more regions than there are decreases. Heavy precipitation has ''likely'' increased on the continental scale over three continents (North America, Europe, and Asia) where observational data are more abundant. There is ''very low confidence'' about changes in sub-daily extreme precipitation due to the limited number of studies and available data.

<div id="11.4.3" class="h2-container"></div>

<span id="model-evaluation-1"></span>
=== 11.4.3 Model Evaluation ===

<div id="h2-31-siblings" class="h2-siblings"></div>

Evaluating climate model competence in simulating heavy precipitation extremes is challenging due to a number of factors, including the lack of reliable observations and the spatial scale mismatch between simulated andobserved data ( [[#Avila--2015|Avila et al., 2015]] ; [[#Alexander--2019|Alexander et al., 2019]] ). Simulated precipitation represents areal means, but station-based observations are conducted at point locations and are often sparse. The areal-reduction factor, the ratio between pointwise station estimates of extreme precipitation and extremes of the areal mean, can be as large as 130% at CMIP6 resolutions (about 100 km) ( [[#Gervais--2014|Gervais et al., 2014]] ). Hence, the order in which gridded station based extreme values are constructed (i.e., if the extreme values are extracted at the station first and then gridded, or if the daily station values are gridded and then the extreme values are extracted) represents different spatial scales of extreme precipitation and needs to be taken into account in model evaluation (Wehner et al. 2020). This aspect has been considered in some studies. Reanalysis products are used in place of station observations for their spatial completeness as well as spatial-scale comparability( [[#Sillmann--2013a|Sillmann et al., 2013a]] ; [[#Kim--2020|Kim et al., 2020]] ; [[#Li--2021|Li et al., 2021]] ). However, reanalyses share similar parametrizations to the models themselves, reducing the objectivity of the comparison.

Different generations of CMIP models have improved over time, though quite modestly ( [[#Flato--2013|Flato et al., 2013]] ; [[#Watterson--2014|Watterson et al., 2014]] ). Improvements in the representation of the magnitude of the Expert Team on Climate Change Detection and Indices (ETCCDI) in CMIP5 over CMIP3( [[#Sillmann--2013a|Sillmann et al., 2013a]] ; [[#Chen--2015a|Chen and Sun, 2015a]] ) have been attributed to higher resolution, as higher-resolution models represent smaller areas at individual grid boxes. Additionally, the spatial distribution of extreme rainfall simulated by high-resolution models is generally more comparable to observations ( [[#Sillmann--2013b|Sillmann et al., 2013b]] ; [[#Kusunoki--2017|Kusunoki, 2017]] , 2018b; [[#Scher--2017|Scher et al., 2017]] ) as these models tend to produce more realistic storms compared to coarser models ( [[#11.7.2|Section 11.7.2]] ). Higher horizontal resolution alone improves simulation of extreme precipitation in some models ( [[#Wehner--2014|Wehner et al., 2014]] ; [[#Kusunoki--2017|Kusunoki, 2017]] , 2018b), but this is insufficient in other models ( [[#Bador--2020|Bador et al., 2020]] ) as parametrization also plays a significant role (M. [[#Wu--2020|]] [[#Wu--2020|Wu et al., 2020]] ). A simple comparison of climatology may not fully reflect the improvements of the new models that have more comprehensive process formulations ( [[#Di%20Luca--2015|Di Luca et al., 2015]] ). [[#Dittus--2016|Dittus et al. (2016)]] found that many of the eight CMIP5 models they evaluated reproduced the observed increase in the difference between areas experiencing an extreme high (90%) and an extreme low (10%) proportion of the annual total precipitation from heavy precipitation (R95p/PRCPTOT) for Northern Hemisphere regions. Additionally, CMIP5 models reproduced the relation between changes in extreme and non-extreme precipitation: an increase in extreme precipitation is at the cost of a decrease in non-extreme precipitation ( [[#Thackeray--2018|Thackeray et al., 2018]] ), a characteristic found in the observational record ( [[#Gu--2018|Gu and Adler, 2018]] ).

The CMIP6 models perform reasonably well in capturing large-scale features of precipitation extremes, including intense precipitation extremes in the intertropical convergence zone (ITCZ), and weak precipitation extremes in dry areas in the tropical regions ( [[#Li--2021|Li et al., 2021]] ) but a double-ITCZ bias over the equatorial central and eastern Pacific that appeared in CMIP5 models remains ( [[IPCC:Wg1:Chapter:Chapter-3#3.3.2.3|Section 3.3.2.3]] ). There are also regional biases in the magnitude of precipitation extremes ( [[#Kim--2020|Kim et al., 2020]] ). The models also have difficulties in reproducing detailed regional patterns of extreme precipitation, such as over the north-east USA ( [[#Agel--2020|Agel and Barlow, 2020]] ), though they performed better for summer extremes over the USA ( [[#Akinsanola--2020|Akinsanola et al., 2020]] ). The comparison between climatologies in the observations and in model simulations shows that the CMIP6 and CMIP5 models that have similar horizontal resolutions also have similar model evaluation scores, and their error patterns are highly correlated ( [[#Wehner--2020|Wehner et al., 2020]] ). In general, extreme precipitation in CMIP6 models tends to be somewhat larger than in CMIP5 models ( [[#Li--2021|Li et al., 2021]] ), reflecting smaller spatial scales of extreme precipitation represented by slightly higher-resolution models ( [[#Gervais--2014|Gervais et al., 2014]] ). This is confirmed by [[#Kim--2020|Kim et al. (2020)]] , who showed that Rx1day and Rx5day simulated by CMIP6 models tend to be closer to point estimates of HadEX3 data ( [[#Dunn--2020|Dunn et al., 2020]] ) than those simulated by CMIP5. Figure 11.14 shows the multi-model ensemble bias in mean Rx1day over the period 1979–2014 from 21 available CMIP6 models when compared with observations and reanalyses. Measured by global land root-mean-square error, the model performance is generally consistent across different observed/reanalysis data products for the extreme precipitation metric (Figure 11.14). The magnitude of extreme area mean precipitation simulated by the CMIP6 models is consistently smaller than the point estimates of HadEX3, but the model values are more comparable to those of areal-mean values (Figure 11.14) of the ERA5 reanalysis or REGEN ( [[#Contractor--2020b|Contractor et al., 2020b]] ). Taylor-plot-based performance metrics reveal strong similarities in the patterns of extreme precipitation errors over land regions between CMIP5 and CMIP6 ( [[#Srivastava--2020|Srivastava et al., 2020]] ; [[#Wehner--2020|Wehner et al., 2020]] ) and between annual mean precipitation errors and Rx1day errors for both generations of models ( [[#Wehner--2020|Wehner et al., 2020]] ).

<div id="_idContainer057" class="Basic-Text-Frame"></div>

[[File:aee292997bbe519c407ddaba59181c17 IPCC_AR6_WGI_Figure_11_14.png]]

'''Figure 11.14 |''' '''Multi-model mean bias in annual maximum daily precipitation (Rx1day, %) for the period 1979–2014.''' Calculated as the difference between the Coupled Model Intercomparison Project Phase 6 (CMIP6) multi-model mean and the average of available observational or reanalysis products including ''(a)'' ERA5, ''(b)'' HadEX3, and ''(c)'' REGEN. Bias is expressed as the percent error relative to the long-term mean of the respective observational data products. Brown indicates that models are too dry, while green indicates that they are too wet. Areas without sufficient observational data are shown in grey. Adapted from [[#Wehner--2020|Wehner et al. (2020)]] under the terms of the Creative Commons Attribution licence. Further details on data sources and processing are available in the chapter data table (Table 11.SM.9).

In general, there is ''high confidence'' that historical simulations by CMIP5 and CMIP6 models of similar horizontal resolutions are interchangeable in their performance in simulating the observed climatology of extreme precipitation ''.''

Studies using regional climate models (RCMs), for example, CORDEX ( [[#Giorgi--2009|Giorgi et al., 2009]] ) over Africa ( [[#Dosio--2015|Dosio et al., 2015]] ; [[#Klutse--2016|Klutse et al., 2016]] ; [[#Pinto--2016|Pinto et al., 2016]] ; [[#Gibba--2019|Gibba et al., 2019]] ), Australia, East Asia ( [[#Park--2016|Park et al., 2016]] ), Europe ( [[#Prein--2016a|Prein et al., 2016a]] ; [[#Fantini--2018|Fantini et al., 2018]] ), and parts of North America ( [[#Diaconescu--2018|Diaconescu et al., 2018]] ) suggest that extreme rainfall events are better captured in RCMs compared to their host GCMs due to their ability to address regional characteristics, for example, topography and coastlines. However, CORDEX simulations do not show good skill over South Asia for heavy precipitation, and do not add value with respect to their GCM source of boundary conditions ( [[#Mishra--2014b|Mishra et al., 2014b]] ; S. [[#Singh--2017|]] [[#Singh--2017|Singh et al., 2017]] ). The evaluation of models in simulating regional processes is discussed in detail in [[IPCC:Wg1:Chapter:Chapter-10#10.3.3.4|Section 10.3.3.4]] . The high-resolution simulation of mid-latitude winter extreme precipitation over land is of similar magnitude to point observations. Simulation of summer extreme precipitation has a large bias when compared with observations at the same spatial scale. Simulated extreme precipitation in the tropics also appears to be too large, indicating possible deficiencies in the parametrization of cumulus convection at this resolution. Indeed, precipitation distributions at both daily and sub-daily time scales are much improved with a convection-permitting model ( [[#Belušić--2020|Belušić et al., 2020]] ) over Western Africa ( [[#Berthou--2019b|Berthou et al., 2019b]] ), East Africa ( [[#Finney--2019|Finney et al., 2019]] ), North America and Canada ( [[#Cannon--2019|Cannon and Innocenti, 2019]] ; [[#Innocenti--2019|Innocenti et al., 2019]] ) and over Belgium in Europe ( [[#Vanden%20Broucke--2019|Vanden Broucke et al., 2019]] ).

In summary, there is ''high confidence'' in the ability of models to capture the large-scale spatial distribution of precipitation extremes over land. The magnitude and frequency of extreme precipitation simulated by CMIP6 models are similar to those simulated by CMIP5 models ( ''hig'' ''h confidence'' ).

<div id="11.4.4" class="h2-container"></div>

<span id="detection-and-attribution-event-attribution-1"></span>
=== 11.4.4 Detection and Attribution, Event Attribution ===

<div id="h2-32-siblings" class="h2-siblings"></div>

Both SREX (Chapter 3, [[#Seneviratne--2012|Seneviratne et al., 2012]] ) and AR5 (Chapter 10, IPCC, 2014) concluded with ''medium confidence'' that anthropogenic forcing has contributed to a global-scale intensification of heavy precipitation over the second half of the 20th century. These assessments were based on the evidence of anthropogenic influence on aspects of the global hydrological cycle, in particular, the human contribution to the warming-induced observed increase in atmospheric moisture that leads to an increase in heavy precipitation, and ''limited evidence'' of anthropogenic influence on extreme precipitation of durations of one and five days.

Since AR5 there has been new and ''robust evidence'' and improved understanding of human influence on extreme precipitation. In particular, detection and attribution analyses have provided consistent and ''robust evidence'' of human influence on extreme precipitation of one- and five-day durations at global to continental scales. The observed increases in Rx1day and Rx5day over the Northern Hemisphere land area during 1951–2005 can be attributed to the effect of combined anthropogenic forcing, including greenhouse gases and anthropogenic aerosols, as simulated by CMIP5 models and the rate of intensification with regard to warming is consistent with C-C scaling ( [[#Zhang--2013|Zhang et al., 2013]] ). This is confirmed to be robust when an additional nine years of observational data and the CMIP6 model simulations were used (Cross-Chapter Box 3.2, Figure 1; [[#Paik--2020|Paik et al., 2020]] ). The influence of greenhouse gases is attributed as the dominant contributor to the observed intensification. The global average of Rx1day in the observations is consistent with simulations by both CMIP5 and CMIP6 models under anthropogenic forcing, but not under natural forcing (Cross-Chapter Box 3.2, Figure 1). The observed increase in the fraction of annual total precipitation falling into the top fifth or top first percentiles of daily precipitation can also be attributed to human influence at the global scale ( [[#Dong--2021|Dong et al., 2021]] ). The CMIP5 models were able to capture the fraction of land experiencing a strong intensification of heavy precipitation during 1960–2010 under anthropogenic forcing, but not in unforced simulations ( [[#Fischer--2014|Fischer et al., 2014]] ). But the models underestimated the observed trends ( [[#Borodina--2017a|Borodina et al., 2017a]] ). Human influence also significantly contributed to the historical changes in record-breaking one-day precipitation ( [[#Shiogama--2016|Shiogama et al., 2016]] ). There is also ''limited evidence'' of the influences of natural forcing. Substantial reductions in Rx5day and Simple Daily Intensity Index (SDII) for daily precipitation intensity over the global summer monsoon regions occurred during 1957–2000 after explosive volcanic eruptions ( [[#Paik--2018|Paik and Min, 2018]] ). The reduction in post-volcanic eruption extreme precipitation in the simulations is closely linked to the decrease in mean precipitation, for which both thermodynamic effects (moisture reduction due to surface cooling) and dynamic effects (monsoon circulation weakening) play important roles.

There has been new evidence of human influence on extreme precipitation at continental scales, including the detection of the combined effect of greenhouse gases and aerosol forcing on Rx1day and Rx5day over North America, Eurasia, and mid-latitude land regions ( [[#Zhang--2013|Zhang et al., 2013]] ) and of greenhouse gas forcing in Rx1day and Rx5day in the mid-to-high latitudes, western and eastern Eurasia, and the global dry regions ( [[#Paik--2020|Paik et al., 2020]] ). These findings are corroborated by the detection of human influence in the fraction of extreme precipitation in the total precipitation over Asia, Europe, and North America ( [[#Dong--2021|Dong et al., 2021]] ). Human influence was found to have contributed to the increase in frequency and intensity of regional precipitation extremes in North America during 1961–2010, based on optimal fingerprinting and event attribution approaches ( [[#Kirchmeier-Young--2020|Kirchmeier-Young and Zhang, 2020]] ). [[#Tabari--2020|Tabari et al. (2020)]] found the observed latitudinal increase in extreme precipitation over Europe to be consistent with model-simulated responses to anthropogenic forcing.

Evidence of human influence on extreme precipitation at regional scales is more limited and less robust. In north-west Australia, the increase in extreme rainfall since 1950 can be related to increased monsoonal flow due to increased aerosol emissions, but cannot be attributed to an increase in greenhouse gases ( [[#Dey--2019a|Dey et al., 2019a]] ). Anthropogenic influence on extreme precipitation in China was detected in one study (H. [[#Li--2017|]] [[#Li--2017|]] [[#Li--2017|]] [[#Li--2017|Li et al., 2017]] ), but not in another using different detection and data-processing procedures (W. [[#Li--2018|]] [[#Li--2018|]] [[#Li--2018|]] [[#Li--2018|Li et al., 2018]] a), indicating the lack of robustness in the detection results. A still weak signal-to-noise ratio seems to be the main cause for the lack of robustness, as detection would become robust 20 years in the future (W. [[#Li--2018|]] [[#Li--2018|]] [[#Li--2018|]] [[#Li--2018|Li et al., 2018]] a). [[#Krishnan--2016|Krishnan et al. (2016)]] attributed the observed increase in heavy rain events (intensity &gt;100 mm day <sup>–1</sup> ) in the post-1950s over central India to the combined effects of greenhouse gases, aerosols, land-use and land-cover changes, and rapid warming of the equatorial Indian Ocean SSTs. Roxyet al. (2017) and [[#Devanand--2019|Devanand et al. (2019)]] showed that the increase in widespread extremes over the South Asian Monsoon during 1950–2015 is due to the combined impacts of the warming of the Western Indian Ocean (Arabian Sea) and the intensification of irrigation water management over India.

Anthropogenic influence may have affected the large-scale meteorological processes necessary for extreme precipitation and the localized thermodynamic and dynamic processes, both contributing to changes in extreme precipitation events. Several new methods have been proposed to disentangle these effects by either conditioning on the circulation state or attributing analogues. In particular, the extremely wet winter of 2013–2014 in the UK can be attributed, approximately to the same degree, to both temperature-induced increases in saturation vapour pressure and changes in the large-scale circulation ( [[#Vautard--2016|Vautard et al., 2016]] ; [[#Yiou--2017|Yiou et al., 2017]] ). There are multiple cases indicating that very extreme precipitation may increase at a rate more than the C-C rate (7% per 1°C of warming) ( [[#Pall--2017|Pall et al., 2017]] ; [[#Risser--2017|Risser and Wehner, 2017]] ; [[#van%20der%20Wiel--2017|van der Wiel et al., 2017]] ; [[#van%20Oldenborgh--2017|van Oldenborgh et al., 2017]] ; S.-Y.S. [[#Wang--2018|]] [[#Wang--2018|Wang et al., 2018]] ).

Event attribution studies found an influence of anthropogenic activities on the probability or magnitude of observed extreme precipitation events, including European winters ( [[#Schaller--2016|Schaller et al., 2016]] ; [[#Otto--2018b|Otto et al., 2018b]] ), extreme 2014 precipitation over the northern Mediterranean ( [[#Vautard--2015|Vautard et al., 2015]] ), parts of the USA for individual events ( [[#Knutson--2014a|Knutson et al., 2014a]] ; [[#Szeto--2015|Szeto et al., 2015]] ; [[#Eden--2016|Eden et al., 2016]] ; [[#van%20Oldenborgh--2017|van Oldenborgh et al., 2017]] ), extreme rainfall in 2014 over Northland, New Zealand ( [[#Rosier--2015|Rosier et al., 2015]] ) or China ( [[#Burke--2016|Burke et al., 2016]] ; [[#Sun--2018|Sun and Miao, 2018]] ; [[#Yuan--2018b|Yuan et al., 2018b]] ; [[#Zhou--2018|Zhou et al., 2018]] ). However, for other heavy rainfall events, studies identified a lack of evidence about anthropogenic influences ( [[#Imada--2013|Imada et al., 2013]] ; [[#Schaller--2014|Schaller et al., 2014]] ; [[#Otto--2015c|Otto et al., 2015c]] ; [[#Siswanto--2015|Siswanto et al., 2015]] ). There are also studies where results are inconclusive because of limited reliable simulations ( [[#Christidis--2013b|Christidis et al., 2013b]] ; [[#Angélil--2016|Angélil et al., 2016]] ). Overall, both the spatial and temporal scales on which extreme precipitation events are defined are important for attribution; events defined on larger scales have larger signal-to-noise ratios and thus the signal is more readily detectable. At the current level of global warming, there is a strong enough signal to be detectable for large-scale extreme precipitation events, but the chance of detecting such signals for smaller-scale events decreases ( [[#Kirchmeier-Young--2019|Kirchmeier-Young et al., 2019]] ).

In summary, most of the observed intensification of heavy precipitation over land regions is ''likely'' due to anthropogenic influence, for which greenhouse gases emissions are the main contributor. New and ''robust evidence'' since AR5 includes attribution to human influence of the observed increases in annual maximum one-day and five-day precipitation and in the fraction of annual precipitation falling in heavy events. The evidence since AR5 also includes a larger fraction of land showing enhanced extreme precipitation and a larger probability of record-breaking one-day precipitation than expected by chance, both of which can only be explained when anthropogenic greenhouse gas forcing is considered. Human influence has contributed to the intensification of heavy precipitation in three continents where observational data are more abundant ( ''high confidence'' ) (North America, Europe and Asia). On the spatial scale of AR6 regions, there is ''limited evidence'' of human influence on extreme precipitation, but new evidence is emerging; in particular, studies attributing individual heavy precipitation events found that human influence was a significant driver of the events, particularly in the winter season.

<div id="11.4.5" class="h2-container"></div>

<span id="projections-1"></span>
=== 11.4.5 Projections ===

<div id="h2-33-siblings" class="h2-siblings"></div>

The AR5 concluded it is ''very likely'' that extreme precipitation events will be more frequent and more intense over most of the mid-latitude land masses and wet tropics in a warmer world ( [[#Collins--2013|Collins et al., 2013]] ). Post-AR5 studies provide more and ''robust evidence'' to support the previous assessments. These include an observed increase in extreme precipitation ( [[#11.4.3|Section 11.4.3]] ) and human causes of past changes ( [[#11.4.4|Section 11.4.4]] ), as well as projections based on either GCM and/or RCM simulations. The CMIP5 models project that the rate of increase in Rx1day with warming is independent of the forcing scenario ( [[IPCC:Wg1:Chapter:Chapter-8#8.5.3.1|Section 8.5.3.1]] ; [[#Pendergrass--2015|Pendergrass et al., 2015]] ) or forcing mechanism ( [[#Sillmann--2017a|Sillmann et al., 2017a]] ). This is confirmed in CMIP6 simulations ( [[#Sillmann--2019|Sillmann et al., 2019]] ; [[#Li--2021|Li et al., 2021]] ). In particular, for extreme precipitation that occurs once a year or less frequently, the magnitudes of the rates of change per 1°C change in global mean temperature are similar, regardless of whether the temperature change is caused by increases in carbon dioxide (CO <sub>2</sub> ), methane (CH <sub>4</sub> ), solar forcing, or sulphate (SO <sub>4</sub> ) ( [[#Sillmann--2019|Sillmann et al., 2019]] ). In some models – CESM1 in particular – the extreme precipitation response to warming may follow a quadratic relation ( [[#Pendergrass--2019|Pendergrass et al., 2019]] ). Figure 11.15 shows changes in the 10- and 50-year return values of Rx1day at different warming levels as simulated by the CMIP6 models. The median value of the scaling over land, across all Shared Socio-economic Pathway (SSP) scenarios and all models, is close to 7% per 1°C of warming for the 50-year return value of Rx1day. It is just slightly smaller for the 10- and 50-year return values of Rx5day ( [[#Li--2021|Li et al., 2021]] ). The 90% ranges of the multimodel ensemble changes across all land grid boxes in the 50-year return values for Rx1day and Rx5day do not overlap between 1.5°C and 2°C warming levels ( [[#Li--2021|Li et al., 2021]] ), indicating that a small increment such as 0.5°C in global warming can result in a significant increase in extreme precipitation. Projected long-period Rx1day return value changes are larger than changes in mean Rx1day and with larger relative changes for more rare events ( [[#Pendergrass--2018|Pendergrass, 2018]] ; [[#Mizuta--2020|Mizuta and Endo, 2020]] ; [[#Wehner--2020|Wehner, 2020]] ). The rate of change of moderate extreme precipitation may depend more on the forcing agent, similar to the mean precipitation response to warming ( [[#Lin--2016|Lin et al., 2016]] , 2018). Thus, there is ''high confidence'' that extreme precipitation that occurs once a year or less frequently increases proportionally to the amount of surface warming, and the rate of change in precipitation is not dependent on the underlying forcing agents of warming.

<div id="_idContainer059" class="Basic-Text-Frame"></div>

[[File:16ec961ba91dca7123ead5a0783a5a3d IPCC_AR6_WGI_Figure_11_15.png]]

'''Figure 11.15 |''' '''Projected changes in the intensity of extreme precipitation events under 1°C, 1.5°C, 2°C, 3°C, and 4°C global warming levels relative to the 1850–1900 baseline.''' Extreme precipitation events are defined as the annual maximum daily maximum precipitation (Rx1day) that was exceeded on average once during a 10-year period (10-year event, blue) and once during a 50-year period (50-year event, orange) during the 1850–1900 base period. Results are shown for the global land. For each box plot, the horizontal line and the box represent the median and central 66% uncertainty range, respectively, of the intensity changes across the multi-model median, and the ‘whiskers’ extend to the 90% uncertainty range. The results are based on the multi-model ensemble estimated from simulations of global climate models contributing to the Coupled Model Intercomparison Project Phase 6 (CMIP6) under different Shared Socio-economic Pathway forcing scenarios. Based on [[#Li--2021|Li et al. (2021)]] . Further details on data sources and processing are available in the chapter data table (Table 11.SM.9).

The spatial patterns of the projected changes across different warming levels are quite similar, as shown in Figure 11.16, and confirmed by near-linear scaling between extreme precipitation and global warming levels at regional scales ( [[#Seneviratne--2020|Seneviratne and Hauser, 2020]] ). Internal variability modulates changes in heavy rainfall ( [[#Wood--2020|Wood and Ludwig, 2020]] ), resulting in different changes in different regions ( [[#Seneviratne--2020|Seneviratne and Hauser, 2020]] ). Extreme precipitation nearly always increases across land areas with larger increases at higher global warming levels, except in very few regions, such as Southern Europe around the Mediterranean Basin at low warming levels (Table 11.17). The ''very likely'' ranges of the multi-model ensemble changes across all land grid boxes in the 50-year return values for Rx1day and Rx5day between 1.5°C and 1°C warming levels are above zero for all continents except Europe, with the lower bound of the ''likely'' range above zero over Europe ( [[#Li--2021|Li et al., 2021]] ). Decreases in extreme precipitation are confined mostly to subtropical ocean areas and are highly correlated to decreases in mean precipitation due to storm track shifts. These subtropical decreases can extend to nearby land areas in individual realizations.

<div id="_idContainer061" class="Basic-Text-Frame"></div>

[[File:94210254c8d018b47fc349341792c580 IPCC_AR6_WGI_Figure_11_16.png]]

'''Figure 11.16 |''' '''Projected changes in annual maximum daily precipitation at (a) 1.5°C, (b) 2°C, and (c) 4°C of global warming compared to the 1850–1900 baseline.''' Results are based on simulations from the Coupled Model Intercomparison Project Phase 6 (CMIP6) multi-model ensemble under the Shared Socio-economic Pathway (SSP), SSP1-1.9, SSP1-2.6, SSP2-4.5, SSP3-7.0, and SSP5-8.5 scenarios. The numbers on the top right indicate the number of simulations included. Uncertainty is represented using the simple approach: no overlay indicates regions with high model agreement, where ≥80% of models agree on the sign of change; diagonal lines indicate regions with low model agreement, where &lt;80% of models agree on the sign of change. For more information on the simple approach, please refer to the Cross-Chapter Box ( [[IPCC:Wg1:Chapter:Atlas|Atlas]] 1. For details on the methods see Supplementary Material 11.SM.2. Changes in Rx1day are also displayed in the Interactive Atlas. Further details on data sources and processing are available in the chapter data table (Table 11.SM.9).

Projected increases in the probability of extreme precipitation of fixed magnitudes are nonlinear and show larger increases for more rare events (Figures 11.7 and 11.15; [[#Fischer--2015|Fischer and Knutti, 2015]] ; [[#Kharin--2018|Kharin et al., 2018]] ; [[#Li--2021|Li et al., 2021]] ).The CMIP5 model projected increases in the probability of high (99th and 99.9th) percentile precipitation between 1.5°C and 2°C warming scenarios are consistent with what can be expected based on observed changes ( [[#Fischer--2015|Fischer and Knutti, 2015]] ), providing confidence in the projections. The CMIP5 model simulations show that the frequency for present-day climate 20-year extreme precipitation is projected to increase by 10% at the 1.5°C global warming level, and by 22% at the 2.0°C global warming level, while the increase in the frequency for present-day climate 100-year extreme precipitation is projected to increase by 20% and more than 45% at the 1.5°C and 2.0°C warming levels, respectively ( [[#Kharin--2018|Kharin et al., 2018]] ). CMIP6 simulations with SSP scenarios show that the frequency of 10-year and 50-year events will be approximately doubled and tripled, respectively, at a very high warming level of 4°C (Figure 11.7; [[#Li--2021|Li et al., 2021]] ).

There is a limited number of studies on the projections of extreme hourly precipitation. The ability of GCMs to simulate hourly precipitation extremes is limited ( [[#Morrison--2019|Morrison et al., 2019]] ) and very few modelling centres archive sub-daily and hourly precipitation prior to CMIP6 experiments. RCM simulations project an increase in extreme sub-daily precipitation in North America ( [[#Li--2019b|]] [[#Li--2019|]] [[#Li--2019|]] [[#Li--2019|C. Li et al., 2019]] b ) and Sweden ( [[#Olsson--2013|Olsson and Foster, 2013]] ), but these models still do not explicitly resolve convective processes that are important for properly simulating extreme sub-daily precipitation. Simulations by RCMs that explicitly resolve convective processes (convection-permitting models) are limited in length and only available in a few regions because of high computing costs. Yet, a majority of the available convection-permitting simulations project increases in the intensities of extreme sub-daily precipitation events, with the amount similar to or higher than the C-C scaling rate ( [[#Kendon--2014|Kendon et al., 2014]] , 2019; [[#Ban--2015|Ban et al., 2015]] ; [[#Prein--2016b|Prein et al., 2016b]] ; [[#Helsen--2020|Helsen et al., 2020]] ; [[#Fowler--2021|Fowler et al., 2021]] ). An increase is projected in extreme sub-daily precipitation over Africa ( [[#Kendon--2019|Kendon et al., 2019]] ); East Africa ( [[#Finney--2020|Finney et al., 2020]] ) and Western Africa ( [[#Berthou--2019a|Berthou et al., 2019a]] ; [[#Fitzpatrick--2020|Fitzpatrick et al., 2020]] ), even for areas where parametrized RCMs project a decrease; in Europe (Hodnebrog et al., 2019; [[#Chan--2020|Chan et al., 2020]] ); as well as in the continental USA ( [[#Prein--2016b|Prein et al., 2016b]] ). Overall, while limited, the available evidence points to an increase in extreme sub-daily precipitation in the future. Studies on future changes in extreme precipitation for a month or longer are limited. One study projects an increase in extreme monthly precipitation in Japan under 4°C global warming for around 80% of stations in the summer ( [[#Hatsuzuka--2019|Hatsuzuka and Sato, 2019]] ).

In Africa (Table 11.5), extreme precipitation will ''likely'' increase under warming levels of 2°C or below (compared to pre-industrial values) and ''very likely'' increase at higher warming levels. Simulations by CMIP5, CMIP6 and CORDEX regional models project an increase in daily extreme precipitation between 1.5°C and 2.0°C warming levels. The pattern of change in heavy precipitation under different scenarios or warming levels is similar with larger increases for higher warming levels (e.g., [[#Nikulin--2018|Nikulin et al., 2018]] ; [[#Li--2021|Li et al., 2021]] ). With increases in warming, extreme precipitation is projected to increase in the majority of land regions in Africa ( [[#Mtongori--2016|Mtongori et al., 2016]] ; [[#Pfahl--2017|Pfahl et al., 2017]] ; [[#Diedhiou--2018|Diedhiou et al., 2018]] ; [[#Dunning--2018|Dunning et al., 2018]] ; [[#Akinyemi--2019|Akinyemi and Abiodun, 2019]] ; [[#Giorgi--2019|Giorgi et al., 2019]] ). Over Southern Africa, heavy precipitation will ''likely'' increase by the end of the 21st century under RCP 8.5 ( [[#Dosio--2016|Dosio, 2016]] ; [[#Pinto--2016|Pinto et al., 2016]] ; [[#Abiodun--2017|Abiodun et al., 2017]] ; [[#Dosio--2019|Dosio et al., 2019]] ). However, heavy rainfall amounts are projected to decrease over western South Africa ( [[#Pinto--2018|Pinto et al., 2018]] ) as a result of a projected decrease in the frequency of the prevailing westerly winds south of the continent that translates into fewer cold fronts and closed mid-latitudes cyclones ( [[#Engelbrecht--2009|Engelbrecht et al., 2009]] ; [[#Pinto--2018|Pinto et al., 2018]] ). Heavy precipitation will ''likely'' increase by the end of the century under RCP8.5 in West Africa ( [[#Diallo--2016|Diallo et al., 2016]] ; [[#Dosio--2016|Dosio, 2016]] ; [[#Sylla--2016|Sylla et al., 2016]] ; [[#Abiodun--2017|Abiodun et al., 2017]] ; [[#Akinsanola--2019|Akinsanola and Zhou, 2019]] ; [[#Dosio--2019|Dosio et al., 2019]] ) and is projected to increase ( ''high confidence'' ) in Central Africa ( [[#Fotso-Nguemo--2018|Fotso-Nguemo et al., 2018]] , 2019; [[#Sonkoué--2019|Sonkoué et al., 2019]] ) and eastern Africa ( [[#Thiery--2016|Thiery et al., 2016]] ; [[#Ongoma--2018a|Ongoma et al., 2018a]] ). In north-east and central east Africa, extreme precipitation intensity is projected to increase across CMIP5, CMIP6 and CORDEX-CORE ( ''high confidence'' ) in most areas annually ( [[#Coppola--2021a|Coppola et al., 2021a]] ), but the trends differ from season to season in all future scenarios ( [[#Dosio--2019|Dosio et al., 2019]] ). In northern Africa, there is ''low confidence'' in the projected changes in heavy precipitation, either due to a lack of agreement among studies on the sign of changes ( [[#Sillmann--2013a|Sillmann et al., 2013a]] ; [[#Giorgi--2014|Giorgi et al., 2014]] ) or due to insufficient evidence.

In Asia (Table 11.8), extreme precipitation will ''likely'' increase at global warming levels of 2°C and below, but ''very likely'' increase at higher warming levels for the region as whole. The CMIP6 multi-model median projects an increase in the 10- and 50-year return values of Rx1day and Rx5day over more than 95% of regions, even at the 2°C warming level, with larger increases at higher warming levels, independent of emissions scenarios ( [[#Li--2021|Li et al., 2021]] , also Figure 11.7). The CMIP5 models produced similar projections. Both heavy rainfall and rainfall intensity are projected to increase ( [[#Zhou--2014|Zhou et al., 2014]] ; [[#Guo--2016|Guo et al., 2016]] , 2018; Y. [[#Xu--2016|]] [[#Xu--2016|Xu et al., 2016]] ; [[#Endo--2017|Endo et al., 2017]] ; [[#Han--2018|Han et al., 2018]] ; G. [[#Kim--2018|]] [[#Kim--2018|]] [[#Kim--2018|Kim et al., 2018]] ). A half-degree difference in warming between the 1.5°C and 2.0°C warming levels can result in a detectable increase in extreme precipitation over the region ( [[#Li--2021|Li et al., 2021]] ), in the Asian–Australian monsoon region ( [[#Chevuturi--2018|Chevuturi et al., 2018]] ), and over South Asia and China (D. [[#Lee--2018|]] [[#Lee--2018|]] [[#Lee--2018|Lee et al., 2018]] ; W. [[#Li--2018|]] [[#Li--2018|]] [[#Li--2018|]] [[#Li--2018|Li et al., 2018]] b). While there are regional differences, extreme precipitation is projected to increase in almost all sub-regions, though there can be spatial heterogeneity within sub-regions, such as in India ( [[#Shashikanth--2018|Shashikanth et al., 2018]] ) and South East Asia ( [[#Ohba--2019|Ohba and Sugimoto, 2019]] ). In East and South East Asia, there is ''high confidence'' that extreme precipitation is projected to intensify (Seo et al., 2014; [[#Zhou--2014|Zhou et al., 2014]] ; Y. [[#Xu--2016|]] [[#Xu--2016|Xu et al., 2016]] ; [[#Nayak--2017|Nayak et al., 2017]] ; X. [[#Wang--2017|]] [[#Wang--2017|Wang et al., 2017]] ; Y. [[#Wang--2017|]] [[#Wang--2017|Wang et al., 2017]] ; [[#Guo--2018|Guo et al., 2018]] ; D. [[#Li--2018|]] [[#Li--2018|]] [[#Li--2018|]] [[#Li--2018|Li et al., 2018]] ; [[#Sui--2018|Sui et al., 2018]] ). Extreme daily precipitation is also projected to increase in South Asia (Xu et al., 2017; [[#Han--2018|Han et al., 2018]] ; [[#Shashikanth--2018|Shashikanth et al., 2018]] ). The extreme precipitation indices, including Rx5day, R95p, and days of heavy precipitation (i.e., R10mm), are all projected to increase under the RCP4.5 and RCP8.5 scenarios in central and northern Asia ( [[#Xu--2017|Xu et al., 2017]] ; [[#Han--2018|Han et al., 2018]] ). A general wetting across the whole Tibetan Plateau and the Himalayas is projected, with increases in heavy precipitation in the 21st century (Palazzi et al., 2013; [[#Zhou--2014|Zhou et al., 2014]] ; [[#Rajbhandari--2015|Rajbhandari et al., 2015]] ; R. [[#Zhang--2015|Zhang et al., 2015]] ; [[#Wu--2017|Wu et al., 2017]] ; [[#Gao--2018|Gao et al., 2018]] ; [[#Paltan--2018|Paltan et al., 2018]] ). Agreement in projected changes by different models is low in regions of complex topography such as Hindu-Kush Himalayas ( [[#Roy--2019|Roy et al., 2019]] ), but CMIP5, CMIP6 and CORDEX-CORE simulations consistently project an increase in heavy precipitation in higher latitude areas, such as West and East Siberia, and Russian Far East ( ''high confidence'' ) ( [[#Coppola--2021a|Coppola et al., 2021a]] ).

In Australasia (Table 11.11), most CMIP5 models project an increase in Rx1day under RCP4.5 and RCP8.5 scenarios for the late 21st century (CSIRO and BOM, 2015; [[#Alexander--2017|Alexander and Arblaster, 2017]] ; [[#Grose--2020|Grose et al., 2020]] ) and the CMIP6 multi-model median projects an increase in the 10- and 50-year return values of Rx1day and Rx5day at a rate between 5% and 6% per 1°C of near-surface global mean warming (Figure 11.7; [[#Li--2021|Li et al., 2021]] ). Yet, there is large uncertainty in the increase because projected changes in dynamic processes lead to a decrease in Rx1day that can offset the thermodynamic increase over a large portion of the region (Box 11.1, Figure 1; [[#Pfahl--2017|Pfahl et al., 2017]] ). Projected changes in moderate extreme precipitation (the 99th percentile of daily precipitation) by RCMs under RCP8.5 for 2070–2099 are mixed, with more regions showing decreases than increases ( [[#Evans--2021|Evans et al., 2021]] ). It is ''likely'' that daily rainfall extremes such as Rx1day will increase at the continental scale for global warming levels at or above 3°C. Daily rainfall extremes are projected to increase at the 2.0°C global warming level ( ''medium confidence'' ), and there is ''low confidence'' in changes at the 1.5°C ''.'' Projected changes show important regional differences with ''very likely'' increases over Northern Australia ( [[#Alexander--2017|Alexander and Arblaster, 2017]] ; [[#Herold--2018|Herold et al., 2018]] ; [[#Grose--2020|Grose et al., 2020]] ) and New Zealand ( [[#MfE--2018|MfE, 2018]] ) where projected dynamic contributions are small (Box 11.1 Figure 1; [[#Pfahl--2017|Pfahl et al., 2017]] ) and ''medium confidence'' on increases over central, eastern, and Southern Australia where dynamic contributions are substantial and can affect local phenomena (CSIRO and BOM, 2015; [[#Pepler--2016|Pepler et al., 2016]] ; [[#Bell--2019|Bell et al., 2019]] ; [[#Dowdy--2019|Dowdy et al., 2019]] ).

In Central and South America (Table 11.14), extreme precipitation will ''likely'' increase at global warming levels of 2°C and below, but ''very likely'' increase at higher warming levels for the region as whole. A larger increase in global surface temperature leads to a larger increase in extreme precipitation, independent of emissions scenarios ( [[#Li--2021|Li et al., 2021]] ). But there are regional differences in the projection, and projected changes for more moderate extreme precipitation are also more uncertain. Extreme precipitation, represented by the number of days with daily precipitation exceeding 50 mm and the annual fraction of precipitation falling during days with the top 10% daily precipitation amount, is projected to increase on the eastern coast of Southern Central America, but to decrease along the Pacific coasts of El Salvador and Guatemala ( [[#Imbach--2018|Imbach et al., 2018]] ). Chouet al. (2014b) and [[#Giorgi--2014|Giorgi et al. (2014)]] projected an increase in extreme precipitation over South-Eastern South America and the Amazon. Projected changes in moderate extreme precipitation represented by the 99th percentile of daily precipitation by different models under different emissions scenarios, even at high warming levels, are mixed: increases are projected for all regions by the CORDEX-CORE and CMIP5 simulations, while increases for some regions and decreases for other regions are projected by CMIP6 simulations ( [[#Coppola--2021a|Coppola et al., 2021a]] ). Extreme precipitation is projected to increase in the La Plata basin ( [[#Cavalcanti--2015|Cavalcanti et al., 2015]] ; [[#Carril--2016|Carril et al., 2016]] ). [[#Taylor--2018|Taylor et al. (2018)]] projected a decrease in days with intense rainfall in the Caribbean under 2°C global warming by the 2050s under RCP4.5 relative to 1971–2000.

In Europe (Table 11.17), extreme precipitation will ''likely'' increase at global warming levels of 2°C and below, but ''very likely'' increase for higher warming levels for the region as whole. The CMIP6 multi-model median projects an increase in the 10- and 50-year return values of Rx1day and Rx5day over a majority of the region at the 2°C global warming level, with more than 95% of the region showing an increase at higher warming levels (Figure 11.7; [[#Li--2021|C. Li et al., 2021]] ). The most intense precipitation events observed today in Europe are projected to almost double in occurrence for each 1°C of further global warming ( [[#Myhre--2019|Myhre et al., 2019]] ). Extreme precipitation is projected to increase in both boreal winter and summer over Europe ( [[#Madsen--2014|Madsen et al., 2014]] ; [[#Christensen--2015|Christensen et al., 2015]] ; [[#Nissen--2017|Nissen and Ulbrich, 2017]] ). There are regional differences, with decreases or no change for the southern part of Europe, such as the southern Mediterranean (Tramblay and Somot, 2018; [[#Lionello--2020|Lionello and Scarascia, 2020]] ; [[#Coppola--2021a|Coppola et al., 2021a]] ), uncertain changes over central Europe ( [[#Argüeso--2012|Argüeso et al., 2012]] ; [[#Croitoru--2013|Croitoru et al., 2013]] ; [[#Rajczak--2013|Rajczak et al., 2013]] ; [[#Casanueva--2014|Casanueva et al., 2014]] ; [[#Patarčić--2014|Patarčić et al., 2014]] ; [[#Paxian--2014|Paxian et al., 2014]] ; [[#Roth--2014|Roth et al., 2014]] ; [[#Fischer--2015|Fischer and Knutti, 2015]] ; [[#Monjo--2016|Monjo et al., 2016]] ) and a strong increase in the remaining parts, including the Alps region ( [[#Gobiet--2014|Gobiet et al., 2014]] ; [[#Donnelly--2017|Donnelly et al., 2017]] ), particularly in winter ( [[#Fischer--2015|Fischer et al., 2015]] ), and in northern Europe. In a 3°C warmer world, there will be a robust increase in extreme rainfall over 80% of land areas in northern Europe ( [[#Madsen--2014|Madsen et al., 2014]] ; [[#Donnelly--2017|Donnelly et al., 2017]] ; [[#Cardell--2020|Cardell et al., 2020]] ).

In North America (Table 11.20), the intensity and frequency of extreme precipitation will ''likely'' increase at the global warming levels of 2°C and below, and ''very likely'' increase at higher warming levels. An increase is projected by CMIP6 model simulations ( [[#Li--2021|Li et al., 2021]] ) and by previous model generations (Wu,2015; Easterling et al., 2017; [[#Innocenti--2019|Innocenti et al., 2019]] ), as well as by RCMs (Coppola et al., 2021a). Projections of extreme precipitation over the southern portion of the continent and over Mexico are more uncertain, with decreases possible ( [[#Sillmann--2013b|Sillmann et al., 2013b]] ; [[#Alexandru--2018|Alexandru, 2018]] ; [[#Coppola--2021a|Coppola et al., 2021a]] ).

In summary, heavy precipitation will generally become more frequent and more intense with additional global warming. At global warming levels of 4°C relative to the pre-industrial, very rare (e.g., one in 10 or more years) heavy precipitation events would become more frequent and more intense than in the recent past, on the global scale ( ''virtually certain'' ), and in all continents and AR6 regions: The increase in frequency and intensity is ''extremely likely'' for most continents and ''very likely'' for most AR6 regions. The likelihood is lower at lower global warming levels and for less-rare heavy precipitation events. At the global scale, the intensification of heavy precipitation will follow the rate of increase in the maximum amount of moisture that the atmosphere can hold as it warms ( ''high confidence'' ), of about 7% per 1°C of global warming. The increase in the frequency of heavy precipitation events will be non-linear with more warming and will be higher for rarer events ( ''high confidence'' ), with 10- and 50-year events to be approximately double and triple, respectively, at the 4°C warming level. Increases in the intensity of extreme precipitation events at regional scales will depend on the amount of regional warming as well as changes in atmospheric circulation and storm dynamics leading to regional differences in the rate of heavy precipitation changes ( ''hi'' ''gh confidence'' ).

<div id="11.5" class="h1-container"></div>

<span id="floods"></span>