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

== 11.6 Droughts ==

<div id="h1-7-siblings" class="h1-siblings"></div>

Droughts refer to periods of time with substantially below-average moisture conditions, usually covering large areas, during which limitations in water availability result in negative impacts for various components of natural systems and economic sectors ( [[#Wilhite--2017|Wilhite and Pulwarty, 2017]] ; [[#Ault--2020|Ault, 2020]] ). Depending on the variables used to characterize it and the systems or sectors being impacted, drought may be classified in different types (Figure 8.6 and Appendix Table 11.A.1) such as meteorological (precipitation deficits), agricultural (e.g., crop yield reductions or failure, often related to soil moisture deficits), ecological (related to plant water stress that causes e.g., tree mortality), or hydrological droughts (e.g., water shortage in streams or storages such as reservoirs, lakes, lagoons, and groundwater; see Glossary). The distinction of drought types is not absolute, as drought can affect different sub-domains of the Earth system concomitantly, but sometimes also asynchronously, including propagation from one drought type to another ( [[#Brunner--2019|Brunner and Tallaksen, 2019]] ). Because of this, drought cannot be characterized using a single universal definition ( [[#Lloyd-Hughes--2014|Lloyd-Hughes, 2014]] ) or directly measured based on a single variable (SREX Chapter 3; [[#Wilhite--2017|Wilhite and Pulwarty, 2017]] ). Drought can happen on a wide range of timescales – from ‘flash droughts’ on a scale of weeks, and characterized by a sudden onset and rapid intensification of drought conditions ( [[#Hunt--2014|Hunt et al., 2014]] ; [[#Otkin--2018|Otkin et al., 2018]] ; [[#Pendergrass--2020|Pendergrass et al., 2020]] ) to multi-year or decadal rainfall deficits – sometimes termed ‘megadroughts’ (see Glossary; [[#Ault--2014|Ault et al., 2014]] ; [[#Cook--2016b|Cook et al., 2016b]] ; [[#Garreaud--2017|Garreaud et al., 2017]] ). Droughts are often analysed using indices that are measures of drought severity, duration and frequency (Sections 8.3.1.6, 8.4.1.6, 12.3.2.6 and 12.3.2.7, and Table 11.A.1). There are many drought indices published in the scientific literature, as also highlighted in SREX (SREX Chapter 3). These can range from anomalies in single variables (e.g., precipitation, soil moisture, runoff, evapotranspiration) to indices combining different atmospheric variables.

This assessment is focused on changes in physical conditions and metrics of direct relevance to droughts: (i) precipitation deficits; (ii) excess of atmospheric evaporative demand (AED); (iii) soil moisture deficits; (iv) hydrological deficits; and e) atmospheric-based indices combining precipitation and AED (Table 11.A.1). In the regional tables ( [[#11.9|Section 11.9]] ), the assessment is structured by drought types, addressing: (i) meteorological, (ii) agricultural and ecological, and (iii) hydrological droughts. Note that the latter two assessments directly inform the [[IPCC:Wg1:Chapter:Chapter-12|Chapter 12]] assessment on projected regional changes in these climatic impact-drivers ( [[IPCC:Wg1:Chapter:Chapter-12#12.4|Section 12.4]] ). The text refers to AR6 region acronyms ( [[#11.9|Section 11.9]] , and see [[IPCC:Wg1:Chapter:Chapter-1#1.4.5|Section 1.4.5]] ).

<div id="11.6.1" class="h2-container"></div>

<span id="mechanisms-and-drivers-3"></span>
=== 11.6.1 Mechanisms and Drivers ===

<div id="h2-39-siblings" class="h2-siblings"></div>

Similar to many other extreme events, droughts occur as a combination of thermodynamic and dynamic processes (Box 11.1). Thermodynamic processes contributing to drought, which are modified by greenhouse gas forcing both at global and regional scales, are mostly related to heat and moisture exchanges, and are also partly modulated by plant coverage and physiology. They affect, for instance, atmospheric humidity, temperature, and radiation, which in turn affect precipitation and/or evapotranspiration in some regions and time frames. However, dynamic processes are particularly important to explain drought variability on different time scales, from a few weeks (flash droughts) to multiannual (megadroughts). There is ''low confidence'' in the effects of greenhouse gas forcing on changes in atmospheric dynamic ( [[IPCC:Wg1:Chapter:Chapter-2#2.4|Section 2.4]] ; [[IPCC:Wg1:Chapter:Chapter-4#4.3.3|Section 4.3.3]] ), and on associated changes in drought occurrence. Thermodynamic processes are thus the main driver of drought changes in a warming climate ( ''hig'' ''h confidence'' ).

<div id="11.6.1.1" class="h3-container"></div>

<span id="precipitation-deficits"></span>
==== 11.6.1.1 Precipitation Deficits ====

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

Lack of precipitation is generally the main factor controlling drought onset. There is ''high confidence'' that atmospheric dynamics, which vary on interannual, decadal and longer time scales, is the dominant contributor to variations in precipitation deficits in the majority of world regions ( [[#Dai--2013|Dai, 2013]] ; [[#Miralles--2014b|Miralles et al., 2014b]] ; [[#Seager--2014|Seager and Hoerling, 2014]] ; [[#Burgman--2015|Burgman and Jang, 2015]] ; [[#Dong--2015|Dong and Dai, 2015]] ; [[#Schubert--2016|Schubert et al., 2016]] ; [[#Raymond--2018|Raymond et al., 2018]] ; [[#Baek--2019|Baek et al., 2019]] ; [[#Drumond--2019|Drumond et al., 2019]] ; [[#Herrera-Estrada--2019|Herrera-Estrada et al., 2019]] ; [[#Gimeno--2020|Gimeno et al., 2020]] ; [[#Mishra--2020|Mishra, 2020]] ). Precipitation deficits are driven by dynamic mechanisms taking place on different spatial scales, including synoptic processes – atmospheric rivers and extratropical cyclones, blocking and ridges ( [[#11.7|Section 11.7]] ; [[#Sousa--2017|Sousa et al., 2017]] ), dominant large-scale circulation patterns ( [[#Kingston--2015|Kingston et al., 2015]] ), and global ocean–atmosphere coupled patterns such as inter-decadal Pacific Oscillation (IPO), Atlantic Multi-decadal Oscillation (AMO) and El Niño–Southern Oscillation (ENSO; [[#Dai--2017|Dai and Zhao, 2017]] ). These various mechanisms occur on different scales, are not independent, and substantially interact with one another. Also regional moisture recycling and land–atmosphere feedbacks play an important role for some precipitation anomalies (see below).

There is ''high confidence'' that land–atmosphere feedbacks play a substantial or dominant role in affecting precipitation deficits in someregions (SREX, Chapter 3; [[#Koster--2011|Koster et al., 2011]] ; [[#Gimeno--2012|Gimeno et al., 2012]] ; [[#Taylor--2012|Taylor et al., 2012]] ; [[#Guillod--2015|Guillod et al., 2015]] ; [[#Tuttle--2016|Tuttle and Salvucci, 2016]] ; [[#Santanello%20Jr.--2018|Santanello Jr. et al., 2018]] ; [[#Haslinger--2019|Haslinger et al., 2019]] ; [[#Herrera-Estrada--2019|Herrera-Estrada et al., 2019]] ). The sign of the feedbacks can be either positive or negative, as well as local or non-local ( [[#Taylor--2012|Taylor et al., 2012]] ; [[#Guillod--2015|Guillod et al., 2015]] ; [[#Tuttle--2016|Tuttle and Salvucci, 2016]] ). Earth system models (ESMs) tend to underestimate non-local negative soil-moisture–precipitation feedbacks ( [[#Taylor--2012|Taylor et al., 2012]] ) and also show high variations in their representation in some regions ( [[#Berg--2017b|Berg et al., 2017b]] ). Soil-moisture–precipitation feedbacks contribute to changes in precipitation in climate model projections in some regions, but ESMs display substantial uncertainties in their representation, and there is thus only ''low confidence'' in these contributions ( [[#Berg--2017b|Berg et al., 2017b]] ; [[#Vogel--2017|Vogel et al., 2017]] , 2018).

<div id="11.6.1.2" class="h3-container"></div>

<span id="atmospheric-evaporative-demand"></span>
==== 11.6.1.2 Atmospheric Evaporative Demand ====

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

Atmospheric evaporative demand (AED) quantifies the maximum amount of actual evapotranspiration (ET) that can happen from land surfaces if they are not limited by water availability (Table 11.A.1). AED is affected by radiative and aerodynamic components. For this reason, the atmospheric dryness, often quantified with the relative humidity or the vapour pressure deficit (VPD), is not equivalent to the AED, as other variables are also highly relevant, including solar radiation and wind speed ( [[#Hobbins--2012|Hobbins et al., 2012]] ; [[#McVicar--2012a|McVicar et al., 2012a]] ; [[#Sheffield--2012|Sheffield et al., 2012]] ). AED can be estimated using different methods ( [[#McMahon--2013|McMahon et al., 2013]] ), and those solely based on air temperature (e.g., Hargreaves, Thornthwaite) usually overestimate it in terms of magnitude and temporal trends ( [[#Sheffield--2012|Sheffield et al., 2012]] ), in particular, in the context of substantial background warming. Physically-based combination methods such as the Penman-Monteith equation are more adequate and recommended since 1998 by the United Nations Food and Agriculture Oganization ( [[#Pereira--2015|Pereira et al., 2015]] ). For this reason, the assessment of this Chapter, when considering atmospheric-based drought indices, only includes AED estimates using the latter (see also [[#11.9|Section 11.9]] ). AED is generally higher than ET, since AED represents an upper bound for ET. Hence, an AED increase does not necessarily lead to increased ET ( [[#Milly--2016|Milly and Dunne, 2016]] ), in particular under drought conditions given soil moisture limitation ( [[#Bonan--2014|Bonan et al., 2014]] ; [[#Berg--2016|Berg et al., 2016]] ; [[#Konings--2017|Konings et al., 2017]] ; [[#Stocker--2018|Stocker et al., 2018]] ). In general, AED is highest in regions where ET is lowest (e.g., desert areas), further illustrating the decoupling between the two variables under limited soil moisture.

The influence of AED on drought depends on the drought type, background climate, the environmental conditions and the moisture availability ( [[#Hobbins--2016|Hobbins et al., 2016]] , 2017; [[#Vicente-Serrano--2020a|Vicente-Serrano et al., 2020a]] ). This influence also includes effects not related to increased ET. Under low soil moisture conditions, increased AED increases plant stress, enhancing the severity of agricultural and ecological droughts ( [[#Williams--2013|Williams et al., 2013]] ; [[#Allen--2015|Allen et al., 2015]] ; [[#McDowell--2016|McDowell et al., 2016]] ; [[#Grossiord--2020|Grossiord et al., 2020]] ). Moreover, high VPD impacts overall plant physiology; it affects the leaf and xylem safety margins, and decreases the sap velocity and plant hydraulic conductance ( [[#Fontes--2018|Fontes et al., 2018]] ). VPD also affects the plant metabolism of carbon and, if prolonged, it may cause plant mortality via carbon starvation ( [[#Breshears--2013|Breshears et al., 2013]] ; [[#Hartmann--2015|Hartmann, 2015]] ). Drought projections based exclusively on AED metrics overestimate changes in soil moisture and runoff deficits. Nevertheless, AED also directly impacts hydrological drought, as ET from surface waters is not limited ( [[#Wurbs--2014|Wurbs and Ayala, 2014]] ; [[#Friedrich--2018|Friedrich et al., 2018]] ; [[#Hogeboom--2018|Hogeboom et al., 2018]] ; K. [[#Xiao--2018|]] [[#Xiao--2018|Xiao et al., 2018]] ), and this effect increases under climate change projections (W. [[#Wang--2018|]] [[#Wang--2018|Wang et al., 2018]] ; [[#Althoff--2020|Althoff et al., 2020]] ). In addition, high AED increases crop water consumptions in irrigated lands ( [[#García-Garizábal--2014|García-Garizábal et al., 2014]] ), contributing to intensifying hydrological droughts downstream ( [[#Fazel--2017|Fazel et al., 2017]] ; [[#Vicente-Serrano--2017|Vicente-Serrano et al., 2017]] ).

On subseasonal to decadal scales, temporal variations in AED are strongly controlled by circulation variability ( [[#Williams--2014|Williams et al., 2014]] ; [[#Chai--2018|Chai et al., 2018]] ; [[#Martens--2018|Martens et al., 2018]] ), but thermodynamic processes also play a fundamental role and, under human-induced climate change, dominate the changes in AED. Atmospheric warming due to increased atmospheric CO <sub>2</sub> concentrations increases AED by means of enhanced VPD in the absence of other influences ( [[#Scheff--2015|Scheff and Frierson, 2015]] ). Because of the greater warming over land than over oceans (Sections 2.3.1.1 and 11.3), the saturation pressure of water vapour increases more over land than over oceans; oceanic air masses advected over land thus contain insufficient water vapour to keep pace with the greater increase in saturation vapour pressure over land ( [[#Sherwood--2014|Sherwood and Fu, 2014]] ; [[#Byrne--2018|Byrne and O’Gorman, 2018]] ; [[#Findell--2019|Findell et al., 2019]] ). Land–atmosphere feedbacks are also important in affecting atmospheric moisture content and temperature, with resulting effects on relative humidity and VPD (Box 11.1; [[#Berg--2016|Berg et al., 2016]] ; [[#Haslinger--2019|Haslinger et al., 2019]] ; S. [[#Zhou--2019|]] [[#Zhou--2019|Zhou et al., 2019]] ).

<div id="11.6.1.3" class="h3-container"></div>

<span id="soil-moisture-deficits"></span>
==== 11.6.1.3 Soil Moisture Deficits ====

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

Soil moisture shows an important correlation with precipitation variability ( [[#Khong--2015|Khong et al., 2015]] ; [[#Seager--2019|Seager et al., 2019]] ), but ET also plays a substantial role in further depleting moisture from soils, in particular in humid regions during periods of precipitation deficits ( [[#Teuling--2013|Teuling et al., 2013]] ; [[#Padrón--2020|Padrón et al., 2020]] ). In addition, soil moisture plays a role in drought self-intensification under dry conditions in which ET is decreased and leads to higher AED ( [[#Miralles--2019|Miralles et al., 2019]] ), an effect that can also contribute to triggering flash droughts ( [[#Otkin--2016|Otkin et al., 2016]] , 2018; [[#DeAngelis--2020|DeAngelis et al., 2020]] ; [[#Pendergrass--2020|Pendergrass et al., 2020]] ). If soil moisture becomes limited, ET is reduced, which may decrease the rate of soil drying, but can also lead to further atmospheric dryness through various feedback loops ( [[#Seneviratne--2010|Seneviratne et al., 2010]] ; [[#Miralles--2014a|Miralles et al., 2014a]] , 2019; [[#Teuling--2018|Teuling, 2018]] ; [[#Vogel--2018|Vogel et al., 2018]] ; S. [[#Zhou--2019|]] [[#Zhou--2019|Zhou et al., 2019]] ; [[#Liu--2020|Liu et al., 2020]] ). The process is complex since vegetation cover plays a role in modulating albedo and in providing access to deeper stores of water (both in the soil and groundwater). Also, changes in land cover and in plant phenology may alter ET ( [[#Sterling--2013|Sterling et al., 2013]] ; [[#Woodward--2014|Woodward et al., 2014]] ; [[#Frank--2015|Frank et al., 2015]] ; [[#Döll--2016|Döll et al., 2016]] ; [[#Ukkola--2016|Ukkola et al., 2016]] ; [[#Trancoso--2017|Trancoso et al., 2017]] ; [[#Hao--2019|Hao et al., 2019]] ; [[#Lian--2020|Lian et al., 2020]] ). Snow depth has strong and direct impacts on soil moisture in many systems ( [[#Gergel--2017|Gergel et al., 2017]] ; [[#Williams--2020|Williams et al., 2020]] ).

Soil moisture directly affects plant water stress and ET. Soil moisture is the primary factor that controls xylem hydraulic conductance – that is, water uptake in plants ( [[#Sperry--2016|Sperry et al., 2016]] ; [[#Hayat--2019|Hayat et al., 2019]] ; X. [[#Chen--2020|]] [[#Chen--2020|Chen et al., 2020]] ). For this reason, soil moisture deficits are the main driver of xylem embolism, the primary cause of plant mortality ( [[#Anderegg--2012|Anderegg et al., 2012]] , 2016; [[#Rowland--2015|Rowland et al., 2015]] ). Also carbon assimilation by plants strongly depends on soil moisture ( [[#Hartzell--2017|Hartzell et al., 2017]] ), with implications for carbon starvation and plant dying if soil moisture deficits are prolonged ( [[#Sevanto--2014|Sevanto et al., 2014]] ). These mechanisms explain that soil moisture deficits are usually more relevant than AED excess to explain gross primary production anomalies and vegetation stress, mostly in sub-humid and semi-arid regions ( [[#Stocker--2018|Stocker et al., 2018]] ; [[#Liu--2020|Liu et al., 2020]] ). High CO <sub>2</sub> concentrations are shown to potentially decrease plant ET and increase plant water-use efficiency, affecting soil moisture levels, but this effect interacts with other CO <sub>2</sub> physiological and radiative effects ( [[#11.6.5.2|Section 11.6.5.2]] and Cross-Chapter Box 5.1), and has less relevance under low soil moisture ( [[#Morgan--2011|Morgan et al., 2011]] ; Z. [[#Xu--2016|]] [[#Xu--2016|Xu et al., 2016]] ; [[#Nackley--2018|Nackley et al., 2018]] ; [[#Dikšaitytė--2019|Dikšaitytė et al., 2019]] ). ESMs represent both surface (around 10cm) and total column soil moisture, whereby total soil moisture is of more direct relevance for root water uptake, in particular by trees. There is evidence that surface soil moisture projections are substantially drier than total soil moisture projections, and may overestimate drying of relevance for most vegetation ( [[#Berg--2017a|Berg et al., 2017a]] ).

<div id="11.6.1.4" class="h3-container"></div>

<span id="hydrological-deficits"></span>
==== 11.6.1.4 Hydrological Deficits ====

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

Drivers of streamflow and surface water deficits are complex and strongly depend on the hydrological system analysed (e.g., streamflows in the headwaters, medium course of the rivers, groundwater, highly regulated hydrological basins). Soil hydrological processes, which control the propagation of meteorological droughts throughout different parts of the hydrological cycle ( [[#Van%20Loon--2012|Van Loon and Van Lanen, 2012]] ), are spatially and temporally complex ( [[#Herrera-Estrada--2017|Herrera-Estrada et al., 2017]] ; S. [[#Huang--2017|Huang et al., 2017]] b) and difficult to quantify ( [[#Van%20Lanen--2016|Van Lanen et al., 2016]] ; [[#Apurv--2017|Apurv et al., 2017]] ; [[#Caillouet--2017|Caillouet et al., 2017]] ; [[#Konapala--2017|Konapala and Mishra, 2017]] ; [[#Hasan--2019|Hasan et al., 2019]] ). The physiographic characteristics of the basins also affect how droughts propagate throughout the hydrological cycle ( [[#Van%20Loon--2012|Van Loon and Van Lanen, 2012]] ; [[#Van%20Lanen--2013|Van Lanen et al., 2013]] ; [[#Van%20Loon--2015|Van Loon, 2015]] ; [[#Konapala--2020|Konapala and]] [[#Mishra--2020|Mishra, 2020]] ; Veettil and [[#Mishra--2020|Mishra, 2020]] ). In addition, the assessment of groundwater deficits is very difficult given the complexity of processes that involve natural and human-driven feedbacks and interactions with the climate system ( [[#Taylor--2013|Taylor et al., 2013]] ). Streamflow and surface water deficits are affected by land cover, groundwater and soil characteristics ( [[#Van%20Lanen--2013|Van Lanen et al., 2013]] ; [[#Van%20Loon--2015|Van Loon and Laaha, 2015]] ; [[#Barker--2016|Barker et al., 2016]] ; [[#Tijdeman--2018|Tijdeman et al., 2018]] ), as well as human activities (water management and demand, damming) and land-use changes ( [[#11.6.4.3|Section 11.6.4.3]] ; [[#Van%20Loon--2016|Van Loon et al., 2016]] ; [[#He--2017|He et al., 2017]] ; [[#Veldkamp--2017|Veldkamp et al., 2017]] ; J. [[#Wu--2018|]] [[#Wu--2018|Wu et al., 2018]] ; Y. [[#Xu--2019|]] [[#Xu--2019|Xu et al., 2019]] ; [[#Jehanzaib--2020|Jehanzaib et al., 2020]] ). Finally, snow and glaciers are relevant for water resources in some regions. For instance, warming affects snowpack levels ( [[#Dierauer--2019|Dierauer et al., 2019]] ; [[#Huning--2020|Huning and AghaKouchak, 2020]] ), as well as the timing of snow melt, thus potentially affecting the seasonality and magnitude of low flows ( [[#Barnhart--2016|Barnhart et al., 2016]] ).

<div id="11.6.1.5" class="h3-container"></div>

<span id="atmospheric-based-drought-indices"></span>
==== 11.6.1.5 Atmospheric-based Drought Indices ====

<div id="h3-5-siblings" class="h3-siblings"></div>

Given the difficulties of drought quantification and data constraints, atmospheric-based drought indices combining both precipitation and AED have been developed, as they can be derived from meteorological data that is available in most regions (with few exceptions). These demand/supply indices are not intended to be metrics of soil moisture, streamflow or vegetation water stress. Because of their reliance on precipitation and AED, they are mostly related to the actual water balance in humid regions, in which ET is not limited by soil moisture and tends towards AED. In water-limited regions and in dry periods everywhere, they constitute an upper bound for overall water-balance deficits (e.g., of surface waters) but are also related to conditions conducive to vegetation stress, particularly under soil moisture limitation ( [[#11.6.1.2|Section 11.6.1.2]] ).

Although there are many atmospheric-based drought indices, two are assessed in this chapter: the Palmer Drought Severity Index (PDSI) and the Standardized Precipitation Evapotranspiration Index (SPEI). The PDSI has been widely used to monitor and quantify drought severity ( [[#Dai--2018|Dai et al., 2018]] ), but is affected by some constraints (SREX Chapter 3; [[#Mukherjee--2018a|Mukherjee et al., 2018a]] ). Although the calculation of the PDSI is based on a soil water budget, the PDSI is essentially a climate drought index that mostly responds to the precipitation and the AED ( [[#van%20der%20Schrier--2013|van der Schrier et al., 2013]] ; [[#Vicente-Serrano--2015|Vicente-Serrano et al., 2015]] ; [[#Dai--2018|Dai et al., 2018]] ). The SPEI also combines precipitation and AED, being equally sensitive to these two variables ( [[#Vicente-Serrano--2015|Vicente-Serrano et al., 2015]] ). The SPEI is more sensitive to AED than the PDSI ( [[#Cook--2014a|Cook et al., 2014a]] ; [[#Vicente-Serrano--2015|Vicente-Serrano et al., 2015]] ), although under humid and normal precipitation conditions, the effects of AED on the SPEI are small ( [[#Tomas-Burguera--2020|Tomas-Burguera et al., 2020]] ). Given the limitations associated with temperature-based AED estimates ( [[#11.6.1.2|Section 11.6.1.2]] ), only studies using the Penman-Monteith-based SPEI and PDSI (hereafter SPEI-PM and PDSI-PM) are considered in this assessment and in the regional tables in [[#11.9|Section 11.9]] .

<div id="11.6.1.6" class="h3-container"></div>

<span id="relation-of-assessed-variables-and-metrics-for-changes-in-different-drought-types"></span>
==== 11.6.1.6 Relation of Assessed Variables and Metrics for Changes in Different Drought Types ====

<div id="h3-6-siblings" class="h3-siblings"></div>

This Chapter assesses changes in meteorological drought, agricultural and ecological droughts, and hydrological droughts. Precipitation-based indices are used for the estimation of changes in meteorological droughts, such as the Standardized Precipitation Index (SPI) and the number of consecutive dry days (CDD). Changes in total soil moisture and soil moisture-based drought events are used for the estimation of changes in agricultural and ecological droughts, complemented by changes in surface soil moisture, water-balance estimates (precipitation minus ET), and SPEI-PM and PDSI-PM. For hydrological droughts, changes in low flows are assessed, sometimes complemented by changes in mean streamflow.

In summary, different drought types exist and they are associated with different impacts and respond differently to increasing greenhouse gas concentrations. Precipitation deficits and changes in evapotranspiration govern net water availability. A lack of sufficient soil moisture, sometimes amplified by increased atmospheric evaporative demand, result in agricultural and ecological drought. Lack of runoff and surface water result in hydrological drought. Drought events are the result of dynamic and/or thermodynamic processes, with thermodynamic processes being the main driver of drought changes under human-induced climate change ( ''hig'' ''h confidence'' ).

<div id="11.6.2" class="h2-container"></div>

<span id="observed-trends-3"></span>
=== 11.6.2 Observed Trends ===

<div id="h2-40-siblings" class="h2-siblings"></div>

Evidence on observed drought trends was limited at the time of SREX (Chapter 3) and AR5 (Chapter 2). The SREX concluded: ‘There is ''medium confidence'' that since the 1950s some regions of the world have experienced a trend to more intense and longer droughts, in particular in southern Europe and west Africa, but in some regions droughts have become less frequent, less intense, or shorter, for example, in Central North America and north-western Australia.’ The assessment at the time did not distinguish between different drought types. This Chapter includes numerous updates on observed drought trends, associated with extensive new literature and longer datasets since AR5.

<div id="11.6.2.1" class="h3-container"></div>

<span id="precipitation-deficits-1"></span>
==== 11.6.2.1 Precipitation Deficits ====

<div id="h3-7-siblings" class="h3-siblings"></div>

Strong precipitation deficits have been recorded in recent decades in the Amazon (2005, 2010), south-western China (2009–2010), south-western North America (2011–2014), Australia (1997–2009), California (2014), the middle East (2012–2016), Chile (2010–2015), the Great Horn of Africa (2011), among others ( [[#van%20Dijk--2013|van Dijk et al., 2013]] ; [[#Mann--2015|Mann and Gleick, 2015]] ; [[#Rowell--2015|Rowell et al., 2015]] ; [[#Marengo--2016|Marengo and Espinoza, 2016]] ; [[#Dai--2017|Dai and Zhao, 2017]] ; [[#Garreaud--2017|Garreaud et al., 2017]] , 2020; [[#Marengo--2017|Marengo et al., 2017]] ; [[#Brito--2018|Brito et al., 2018]] ; [[#Cook--2018|Cook et al., 2018]] ). Global studies generally show no significant trends in SPI time series ( [[#Orlowsky--2013|Orlowsky and Seneviratne, 2013]] ; [[#Spinoni--2014|Spinoni et al., 2014]] ), and in derived drought frequency and severity data ( [[#Spinoni--2019|Spinoni et al., 2019]] ), with very few regional exceptions ( [[#11.9|Section 11.9]] and Figure 11.17). Long-term decreases in precipitation are found in some AR6 regions in Africa (Central Africa and East Southern Africa), and several regions in South America (North-Eastern South America, South American Monsoon, South-Western South America, and Southern South America) ( [[#11.9|Section 11.9]] ). Evidence of precipitation-based drying trends is also found in Western Africa, consistent with studies based on CDD trends (Figure 11.17; [[#Chaney--2014|Chaney et al., 2014]] ; [[#Donat--2014b|Donat et al., 2014b]] ; [[#Barry--2018|Barry et al., 2018]] ; [[#Dunn--2020|Dunn et al., 2020]] ), however, there is a partial recovery of the rainfall trends since the 1980s in this region ( [[IPCC:Wg1:Chapter:Chapter-10#10.4.2.1|Section 10.4.2.1]] ). Some AR6 regions show a decrease in meteorological drought, including Northern Australia, Central Australia, Northern Europe and Central North America ( [[#11.9|Section 11.9]] ). Other regions either do not show substantial trends in long-term meteorological drought, or they display mixed signals depending on the considered time frame and sub-regions, such as in Southern Australia ( [[#Gallant--2013|Gallant et al., 2013]] ; [[#Delworth--2014|Delworth and Zeng, 2014]] ; [[#Alexander--2017|Alexander and Arblaster, 2017]] ; [[#Spinoni--2019|Spinoni et al., 2019]] ; [[#Dunn--2020|Dunn et al., 2020]] ; [[#Rauniyar--2020|Rauniyar and Power, 2020]] ) and the Mediterranean ( [[#Camuffo--2013|Camuffo et al., 2013]] ; [[#Gudmundsson--2016|Gudmundsson and Seneviratne, 2016]] ; [[#Spinoni--2017|Spinoni et al., 2017]] ; [[#Stagge--2017|Stagge et al., 2017]] ; [[#Caloiero--2018|Caloiero et al., 2018]] ; [[#Peña-Angulo--2020b|Peña-Angulo et al., 2020b]] ; see also [[#11.9|Section 11.9]] and Atlas.8.2).

<div id="_idContainer063" class="Basic-Text-Frame"></div>

[[File:a97af5c3786b1cd60461310b16197a9a IPCC_AR6_WGI_Figure_11_17.png]]

'''Figure 11.17 |''' '''Observed linear trend for (a) consecutive dry days (CDD) during 1960–2018, (b) standardized precipitation index (SPI) and (c) standardized precipitation-evapotranspiration index (SPEI) dur''' ing 1951–2016. CDD data are from the HadEx3 dataset ( [[#Dunn--2020|Dunn et al., 2020]] ), trend calculation of CDD as in Figure 11.9. Drought severity is estimated using 12-month SPI (SPI-12) and 12-month SPEI (SPEI-12). SPI and SPEI datasets are from [[#Spinoni--2019|Spinoni et al. (2019)]] . The threshold to identify drought episodes was set at -1 SPI/SPEI units. Areas without sufficient data are shown in grey. No overlay indicates regions where the trends are significant at the p = 0.1 level. Crosses indicate regions where trends are not significant. For details on the methods see Supplementary Material 11.SM.2. Further details on data sources and processing are available in the chapter data table (Table 11.SM.9).

<div id="11.6.2.2" class="h3-container"></div>

<span id="atmospheric-evaporative-demand-1"></span>
==== 11.6.2.2 Atmospheric Evaporative Demand ====

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

In several regions, AED increases have intensified recent drought events ( [[#Williams--2014|Williams et al., 2014]] , 2020; [[#Seager--2015b|Seager et al., 2015b]] ; [[#Basara--2019|Basara et al., 2019]] ; [[#García-Herrera--2019|García-Herrera et al., 2019]] ), enhanced vegetation stress ( [[#Allen--2015|Allen et al., 2015]] ; [[#Sanginés%20de%20Cárcer--2018|Sanginés de Cárcer et al., 2018]] ; [[#Yuan--2019|Yuan et al., 2019]] ), or contributed to the depletion of soil moisture or runoff through enhanced ET ( ''high confidence'' ) ( [[#Teuling--2013|Teuling et al., 2013]] ; [[#Padrón--2020|Padrón et al., 2020]] ). Trends in pan evaporation measurements and Penman-Monteith AED estimates provide an indication of possible trends in the influence of AED on drought. Given the observed global temperature increases (Sections 2.3.1.1 and 11.3) and dominant decrease in relative humidity over land areas ( [[#Simmons--2010|Simmons et al., 2010]] ; [[#Willett--2014|Willett et al., 2014]] ), VPD has increased globally ( [[#Barkhordarian--2019|Barkhordarian et al., 2019]] ; [[#Yuan--2019|Yuan et al., 2019]] ). Pan evaporation has increased as a consequence of VPD changes in several AR6 regions, such as East Asia ( [[#Li--2013|Li et al., 2013]] ; Z. [[#Sun--2018|Sun et al., 2018]] ; M.-Z. [[#Yang--2018|]] [[#Yang--2018|]] [[#Yang--2018|]] [[#Yang--2018|]] [[#Yang--2018|Yang et al., 2018]] ), Western and Central Europe ( [[#Mozny--2020|Mozny et al., 2020]] ), the Mediterranean, ( [[#Azorin-Molina--2015|Azorin-Molina et al., 2015]] ) and Central and Southern Australia ( [[#Stephens--2018|Stephens et al., 2018]] ). Nevertheless, there is an important regional variability in observed trends, and in other AR6 regions pan evaporation has decreased – for example, in North Central America ( [[#Breña-Naranjo--2017|Breña-Naranjo et al., 2017]] ) and in the Tibetan Plateau ( [[#Zhang--2018|]] [[#Zhang--2018|]] [[#Zhang--2018|C. Zhang et al., 2018]] )). Physical models also show an important regional diversity, with an increase in New Zealand ( [[#Salinger--2014|Salinger and Porteous, 2014]] ) and the Mediterranean ( [[#Gocic--2014|Gocic and Trajkovic, 2014]] ; [[#Azorin-Molina--2015|Azorin-Molina et al., 2015]] ; [[#Piticar--2016|Piticar et al., 2016]] ), a decrease in South Asia ( [[#Jhajharia--2015|Jhajharia et al., 2015]] ), and strong spatial variability in North America ( [[#Seager--2015b|Seager et al., 2015b]] ). This variability is driven by the role of other meteorological variables affecting AED. Changes in solar radiation as a consequence of solar dimming and brightening may affect trends ( [[IPCC:Wg1:Chapter:Chapter-7#7.2.2.2|Section 7.2.2.2]] ; [[#Kambezidis--2012|Kambezidis et al., 2012]] ; [[#Wang--2014|Wang and Yang, 2014]] ; [[#Sanchez-Lorenzo--2015|Sanchez-Lorenzo et al., 2015]] ). Wind speed is also relevant ( [[#McVicar--2012b|McVicar et al., 2012b]] ), and studies suggest a reduction of the wind speed in some regions (Z. [[#Zhang--2019|]] [[#Zhang--2019|]] [[#Zhang--2019|]] [[#Zhang--2019|Zhang et al., 2019]] b) that could compensate the role of the VPD increase. Nevertheless, the VPD trend seems to dominate the overall AED trends, compared to the effects of trends in wind speed and solar radiation ( [[#Wang--2012|Wang et al., 2012]] ; [[#Park%20Williams--2017|Park Williams et al., 2017]] ; [[#Vicente-Serrano--2020a|Vicente-Serrano et al., 2020a]] ).

<div id="11.6.2.3" class="h3-container"></div>

<span id="soil-moisture-deficits-1"></span>
==== 11.6.2.3 Soil Moisture Deficits ====

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

There are limited long-term measurements of soil moisture from ground observations ( [[#Dorigo--2011|Dorigo et al., 2011]] ; [[#Qiu--2016|Qiu et al., 2016]] ; [[#Quiring--2016|Quiring et al., 2016]] ), which impedes their use in the analysis of trends. Among the few existing observational studies covering at least two decades, several studies have investigated trends in ground soil moisture in East Asia ( [[#11.9|Section 11.9]] ; [[#Chen--2015b|Chen and Sun, 2015b]] ; [[#Liu--2015|Liu et al., 2015]] ; [[#Qiu--2016|Qiu et al., 2016]] ). Alternatively, microwave-based satellite measurements of surface soil moisture have also been used to analyse trends ( [[#Dorigo--2012|Dorigo et al., 2012]] ; [[#Jia--2018|Jia et al., 2018]] ). Although there is regional evidence that microwave-based soil moisture estimates can capture well drying trends in comparison with ground soil moisture observations ( [[#Jia--2018|Jia et al., 2018]] ), there is only ''medium confidence'' in the derived trends, since satellite soil moisture data are affected by inhomogeneities ( [[#Dorigo--2015|Dorigo et al., 2015]] ; [[#Rodell--2018|Rodell et al., 2018]] ; [[#Preimesberger--2021|Preimesberger et al., 2021]] ). Furthermore, microwave-based satellites only sense surface soil moisture, which differs from root-zone soil moisture ( [[#Berg--2017a|Berg et al., 2017a]] ), although relationships can be derived between the two ( [[#Brocca--2011|Brocca et al., 2011]] ). Several studies have also analysed long-term soil moisture time series from observation-driven land-surface or hydrological models, including land-based reanalysis products ( [[#Albergel--2013|Albergel et al., 2013]] ; [[#Jia--2018|Jia et al., 2018]] ; [[#Gu--2019b|Gu et al., 2019b]] ; [[#Markonis--2021|Markonis et al., 2021]] ). Such models have also been used to assess changes in land water availability, estimated as precipitation minus ET, which is equal to the sum of soil moisture and runoff ( [[#Greve--2014|Greve et al., 2014]] ; [[#Padrón--2020|Padrón et al., 2020]] ).

Overall, evidence from global studies suggests that several land regions have been affected by increased soil moisture drying or water balance drying in past decades, despite some spread among products ( [[#Albergel--2013|Albergel et al., 2013]] ; [[#Greve--2014|Greve et al., 2014]] ; [[#Gu--2019b|Gu et al., 2019b]] ; [[#Padrón--2020|Padrón et al., 2020]] ). Drying has not only occurred in dry regions but also in humid regions ( [[#Greve--2014|Greve et al., 2014]] ). Some studies have specifically addressed changes in soil moisture at regional scale ( [[#11.9|Section 11.9]] ). For AR6 regions, several studies suggest an increase in the frequency and areal extent of soil moisture deficits, with examples in East Asia ( [[#Cheng--2015|Cheng et al., 2015]] ; Y. [[#Qin--2015|]] [[#Qin--2015|Qin et al., 2015]] ; [[#Jia--2018|Jia et al., 2018]] ), Western and Central Europe ( [[#Trnka--2015b|Trnka et al., 2015b]] ), and the Mediterranean ( [[#Hanel--2018|Hanel et al., 2018]] ; [[#Moravec--2019|Moravec et al., 2019]] ; [[#Markonis--2021|Markonis et al., 2021]] ). Nonetheless, some analyses also show no long-term trends in soil drying in some AR6 regions – for example, in Eastern North America ( [[#Park%20Williams--2017|Park Williams et al., 2017]] ) and Central North America ( [[#Seager--2019|Seager et al., 2019]] ), as well as in North Eastern Africa ( [[#Kew--2021|Kew et al., 2021]] ). The soil moisture drying trends identified in both global and regional studies are generally related to increases in ET (associated with higher AED) rather than decreases in precipitation, as identified on global land for trends in water balance in the dry season ( [[#Padrón--2020|Padrón et al., 2020]] ), as well as for some regions ( [[#Teuling--2013|Teuling et al., 2013]] ; [[#Cheng--2015|Cheng et al., 2015]] ; [[#Trnka--2015a|Trnka et al., 2015a]] ; [[#van%20Der%20Linden--2019|van Der Linden et al., 2019]] ; X. [[#Li--2020|]] [[#Li--2020|]] [[#Li--2020|Li et al., 2020]] ).

Evidence from observed or observations-derived trends in soil moisture and precipitation minus ET, are combined with evidence from SPEI and PDSI-PM studies to derive regional assessments of changes in agricultural and ecological droughts ( [[#11.9|Section 11.9]] ). This assessment is summarized in [[#11.6.2.6|Section 11.6.2.6]] .

<div id="11.6.2.4" class="h3-container"></div>

<span id="hydrological-deficits-1"></span>
==== 11.6.2.4 Hydrological Deficits ====

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

There is evidence based on streamflow records of increased hydrological droughts in East Asia (D. [[#Zhang--2018|]] [[#Zhang--2018|]] [[#Zhang--2018|Zhang et al., 2018]] ) and southern Africa ( [[#Gudmundsson--2019|Gudmundsson et al., 2019]] ). In areas of Western and Central Europe and Northern Europe, there is no evidence of changes in the severity of hydrological droughts since 1950 based on flow reconstructions ( [[#Caillouet--2017|Caillouet et al., 2017]] ; [[#Barker--2019|Barker et al., 2019]] ) and observations ( [[#Vicente-Serrano--2019|Vicente-Serrano et al., 2019]] ). In the Mediterranean region, there is ''high confidence'' in hydrological drought intensification ( [[#11.9|Section 11.9]] ; [[#Giuntoli--2013|Giuntoli et al., 2013]] ; [[#Lorenzo-Lacruz--2013|Lorenzo-Lacruz et al., 2013]] ; [[#Gudmundsson--2019|Gudmundsson et al., 2019]] ). In south-eastern South America there is a decrease in the severity of hydrological droughts ( [[#Rivera--2018|Rivera and Penalba, 2018]] ). In North America, depending on the methods, datasets and study periods, there are differences between studies that suggest an increase ( [[#Shukla--2015|Shukla et al., 2015]] ; [[#Udall--2017|Udall and Overpeck, 2017]] ) versus a decrease in hydrological drought frequency ( [[#Mo--2018|Mo and Lettenmaier, 2018]] ), but in general there is strong spatial variability ( [[#Poshtiri--2016|Poshtiri and Pal, 2016]] ). Streamflow observation reference networks of near-natural catchments have also been used to isolate the effect of climate trends on hydrological drought trends in a few regions, but these show limited trends in Northern Europe and Western and Central Europe ( [[#Stahl--2010|Stahl et al., 2010]] ; [[#Bard--2015|Bard et al., 2015]] ; [[#Harrigan--2018|Harrigan et al., 2018]] ), North America ( [[#Dudley--2020|Dudley et al., 2020]] ) and most of Australia, with the exception of Eastern and Southern Australia (X.S. [[#Zhang--2016|Zhang et al., 2016]] ). Given the low availability of observations, there are few studies analysing trends of drought severity in the groundwater. Nevertheless, some studies suggest a noticeable response of groundwater droughts to climate variability ( [[#Lorenzo-Lacruz--2017|Lorenzo-Lacruz et al., 2017]] ) and increased drought frequency and severity associated with warming, probably as a consequence of enhanced ET induced by higher AED ( [[#Maxwell--2016|Maxwell and Condon, 2016]] ). This is supported by studies in Northern Europe ( [[#Bloomfield--2019|Bloomfield et al., 2019]] ) and North America ( [[#Condon--2020|Condon et al., 2020]] ).

<div id="11.6.2.5" class="h3-container"></div>

<span id="atmospheric-based-drought-indices-1"></span>
==== 11.6.2.5 Atmospheric-based Drought Indices ====

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

Globally, trends in SPEI-PM and PDSI-PM suggest slightly higher increases of drought frequency and severity in regions affected by drying over the last decades in comparison to the SPI ( [[#Dai--2017|Dai and Zhao, 2017]] ; [[#Spinoni--2019|Spinoni et al., 2019]] ; [[#Song--2020|Song et al., 2020]] ), mainly in regions of Western and Southern Africa, the Mediterranean and East Asia (Figure 11.17), which is consistent with observed soil moisture trends ( [[#11.6.2.3|Section 11.6.2.3]] ). These indices suggest that AED has contributed to increase the severity of agricultural and ecological droughts compared to meteorological droughts ( [[#García-Herrera--2019|García-Herrera et al., 2019]] ; [[#Williams--2020|Williams et al., 2020]] ), reduce soil moisture during the dry season ( [[#Padrón--2020|Padrón et al., 2020]] ), increase plant water stress ( [[#Allen--2015|Allen et al., 2015]] ; [[#Grossiord--2020|Grossiord et al., 2020]] ; [[#Solander--2020|Solander et al., 2020]] ) and trigger more severe forest fires ( [[#Abatzoglou--2016|Abatzoglou and Williams, 2016]] ; [[#Turco--2019|Turco et al., 2019]] ; [[#Nolan--2020|Nolan et al., 2020]] ). A number of regional studies based on these drought indices have also shown stronger drying trends in comparison to trends in precipitation-based indices in the following AR6 regions (see also [[#11.9|Section 11.9]] ): NSA (R. [[#Fu--2013|]] [[#Fu--2013|Fu et al., 2013]] ; [[#Marengo--2016|Marengo and Espinoza, 2016]] ), SCA ( [[#Hidalgo--2017|Hidalgo et al., 2017]] ), WCA ( [[#Tabari--2013|Tabari and Aghajanloo, 2013]] ; [[#Sharafati--2020|Sharafati et al., 2020]] ), SAS ( [[#Niranjan%20Kumar--2013|Niranjan Kumar et al., 2013]] ), NEAF ( [[#Zeleke--2017|Zeleke et al., 2017]] ), WSAF ( [[#Edossa--2016|Edossa et al., 2016]] ), NWN and NEN ( [[#Bonsal--2013|Bonsal et al., 2013]] ), EAS ( [[#Yu--2014|Yu et al., 2014]] ; [[#Chen--2015b|Chen and Sun, 2015b]] ; L. [[#Li--2020|]] [[#Li--2020|]] [[#Li--2020|Li et al., 2020]] ; [[#Liang--2020|Liang et al., 2020]] ; Z. [[#Wu--2020|]] [[#Wu--2020|Wu et al., 2020]] ) and MED ( [[#Kelley--2015|Kelley et al., 2015]] ; [[#Stagge--2017|Stagge et al., 2017]] ; [[#González-Hidalgo--2018|González-Hidalgo et al., 2018]] ; [[#Mathbout--2018a|Mathbout et al., 2018a]] ).

<div id="11.6.2.6" class="h3-container"></div>

<span id="synthesis-for-different-drought-types"></span>
==== 11.6.2.6 Synthesis for Different Drought Types ====

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

Few AR6 regions show observed increases in meteorological drought ( [[#11.9|Section 11.9]] ), mostly in Africa and South America (NES: ''high confidence'' ; WAF, CAF, ESAF, SAM, SWS, SSA, SAS: ''medium confidence'' ); a few others show a decrease (WSB, ESB, NAU, CAU, NEU, CNA: ''medium confidence'' ). There are stronger signals indicating observed increases in agricultural and ecological drought ( [[#11.9|Section 11.9]] ), which highlights the role of increased ET, driven by increased AED, for these trends (Sections 11.6.2.3 and11.6.2.5). Past increases in agricultural and ecological droughts are found on all continents and several regions (WAF, CAF, WSAF, ESAF, WCA, ECA, EAS, SAU, MED, WCE, NES: ''medium confidence'' ), while decreases are found only in one AR6 region (NAU: ''medium confidence'' ). The more limited availability of datasets makes it more difficult to assess historical trends in hydrological drought at regional scale ( [[#11.9|Section 11.9]] ). Increasing (MED: ''high confidence'' ; WAF, EAS, SAU: ''medium confidence'' ) and decreasing (NEU, SES: ''medium confidence'' ) trends in hydrological droughts have only been observed in a few regions.

In summary, there is ''high confidence'' that AED has increased on average on continents, contributing to increased ET and resulting water stress during periods with precipitation deficits, in particular during dry seasons. There is ''medium confidence'' in increases in precipitation deficits in a few regions of Africa and South America. Based on multiple evidence, there is ''medium confidence'' that agricultural and ecological droughts have increased in several regions on all continents (WAF, CAF, WSAF, ESAF, WCA, ECA, EAS, SAU, MED, WCE, NES: ''medium confidence'' ), while there is only ''medium confidence'' in decreases in one AR6 region (NAU). More severe hydrological droughts are found in fewer regions (MED: ''high confidence'' ; WAF, EAS, SAU: ''mediu'' ''m confidence'' ).

<div id="11.6.3" class="h2-container"></div>

<span id="model-evaluation-3"></span>
=== 11.6.3 Model Evaluation ===

<div id="h2-41-siblings" class="h2-siblings"></div>

<div id="11.6.3.1" class="h3-container"></div>

<span id="precipitation-deficits-2"></span>
==== 11.6.3.1 Precipitation Deficits ====

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

ESMs generally show limited performance and large spread in identifying precipitation deficits and associated long-term trends in comparison with observations ( [[#Nasrollahi--2015|Nasrollahi et al., 2015]] ). Meteorological drought trends in the CMIP5 ensemble showed substantial disagreements compared with observations ( [[#Orlowsky--2013|Orlowsky and Seneviratne, 2013]] ; [[#Knutson--2018|Knutson and Zeng, 2018]] ) including a tendency to overestimate drying, in particular in mid- to high latitudes ( [[#Knutson--2018|Knutson and Zeng, 2018]] ). The CMIP6 models display a better performance in reproducing long-term precipitation trends or seasonal dynamics in some studies in Southern South America ( [[#Rivera--2020|Rivera and Arnould, 2020]] ), East Asia ( [[#Xin--2020|Xin et al., 2020]] ), southern Asia ( [[#Gusain--2020|Gusain et al., 2020]] ), and south-western Europe ( [[#Peña-Angulo--2020b|Peña-Angulo et al., 2020b]] ), but there is still too ''limited evidence'' to allow for an assessment of possible differences in performance between CMIP5 and CMIP6. Furthermore, ESMs are generally found to underestimate the severity of precipitation deficits and the dry day frequencies in comparison to observations ( [[#Fantini--2018|Fantini et al., 2018]] ; [[#Ukkola--2018|Ukkola et al., 2018]] ). This is probably related to shortcomings in the simulation of persistent weather events in the mid-latitudes ( [[IPCC:Wg1:Chapter:Chapter-10#10.3.3.3|Section 10.3.3.3]] ). ESMs also show a tendency to underestimate precipitation-based drought persistence at monthly to decadal time scales ( [[#Ault--2014|Ault et al., 2014]] ; [[#Moon--2018|Moon et al., 2018]] ). The overall inter-model spread in the projected frequency of precipitation deficits is also substantial ( [[#Touma--2015|Touma et al., 2015]] ; [[#Zhao--2016|Zhao et al., 2016]] ; [[#Engström--2018|Engström and Keellings, 2018]] ). Moreover, there are spatial differences in the spread, which is higher in the regions where enhanced drought conditions are projected and under high-emissions scenarios ( [[#Orlowsky--2013|Orlowsky and Seneviratne, 2013]] ). Nonetheless, some event attribution studies have concluded that droughts at regional scales can be adequately simulated by some climate models ( [[#Schaller--2016|Schaller et al., 2016]] ; [[#Otto--2018c|Otto et al., 2018c]] ).

<div id="11.6.3.2" class="h3-container"></div>

<span id="atmospheric-evaporative-demand-2"></span>
==== 11.6.3.2 Atmospheric Evaporative Demand ====

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

There is only ''limited evidence'' on the evaluation of AED in state-of-the-art ESMs, which is performed on externally computed AED, based on model output ( [[#Scheff--2015|Scheff and Frierson, 2015]] ; [[#Liu--2016|Liu and Sun, 2016]] , 2017). An evaluation of average AED in 17 CMIP5 ESMs for 1981–1999 based on potential evaporation show that the models’ spatial patterns resemble the observations, but the magnitude of potential evaporation displays strong divergence among models globally and regionally ( [[#Scheff--2015|Scheff and Frierson, 2015]] ). The evaluation of AED in 12 CMIP5 ESMs with pan evaporation observations in East Asia for 1961–2000 ( [[#Liu--2016|Liu and Sun, 2016]] , 2017) show that the ESMs capture seasonal cycles well, but that regional AED averages are underestimated due to biases in the meteorological variables controlling the aerodynamic and radiative components of AED. The CMIP5 ESMs also show a strong underestimation of atmospheric drying trends compared to reanalysis data ( [[#Douville--2017|Douville and Plazzotta, 2017]] ).

<div id="11.6.3.3" class="h3-container"></div>

<span id="soil-moisture-deficits-2"></span>
==== 11.6.3.3 Soil Moisture Deficits ====

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

The performance of climate models for representing soil moisture deficits shows more uncertainty than for precipitation deficits since, in addition to the uncertainties related to cloud and precipitation processes, there is uncertainty related to the representation of complex soil hydrological and boundary-layer processes ( [[#van%20den%20Hurk--2011|van den Hurk et al., 2011]] ; [[#Lu--2019|Lu et al., 2019]] ; [[#Quintana-Seguí--2020|Quintana-Seguí et al., 2020]] ). Another limitation is the lack of observations, particularly for soil moisture, in most regions ( [[#11.6.2.3|Section 11.6.2.3]] ) and the paucity of land surface property data to parametrize land surface models, in particular soil types, soil properties and depth ( [[#Xia--2015|Xia et al., 2015]] ). The spatial resolution of models is an additional limitation since the representation of some land–atmosphere feedbacks and topographic effects requires detailed resolution ( [[#Nicolai-Shaw--2015|Nicolai-Shaw et al., 2015]] ; Van Der Linden et al., 2019). In addition to climate models, land surface and hydrological models are also used to derive historical and projected trends in soil moisture and related land water variables ( [[#Albergel--2013|Albergel et al., 2013]] ; [[#Cheng--2015|Cheng et al., 2015]] ; [[#Gu--2019b|Gu et al., 2019b]] ; [[#Padrón--2020|Padrón et al., 2020]] ; [[#Markonis--2021|Markonis et al., 2021]] ; [[#Pokhrel--2021|Pokhrel et al., 2021]] ).

Overall, there are contrasting results on the performance of land surface models and climate models in representing soil moisture. Some studies suggest that soil moisture anomalies are well captured by land surface models driven with observation-based forcing ( [[#Dirmeyer--2006|Dirmeyer et al., 2006]] ; [[#Albergel--2013|Albergel et al., 2013]] ; [[#Xia--2014|Xia et al., 2014]] ; [[#Balsamo--2015|Balsamo et al., 2015]] ; [[#Reichle--2017|Reichle et al., 2017]] ; [[#Spennemann--2020|Spennemann et al., 2020]] ), but other studies report limited agreement in the representation of interannual soil moisture variability ( [[#Stillman--2016|Stillman et al., 2016]] ; [[#Yuan--2017|Yuan and Quiring, 2017]] ; [[#Ford--2019|Ford and Quiring, 2019]] ) and noticeable seasonal differences in model skill in some regions ( [[#Xia--2014|Xia et al., 2014]] , 2015). Models with good skill can nonetheless display biases in absolute soil moisture ( [[#Xia--2014|Xia et al., 2014]] ; [[#Gu--2019a|Gu et al., 2019a]] ), but these are not necessarily of relevance for the simulation of surface water fluxes and drought anomalies ( [[#Koster--2009|Koster et al., 2009]] ). There is also substantial inter-model spread ( [[#Albergel--2013|Albergel et al., 2013]] ), particularly for the root-zone soil moisture ( [[#Berg--2017a|Berg et al., 2017a]] ).

Regarding the performance of regional and global climate models, an evaluation of an ensemble of RCM simulations for Europe ( [[#Stegehuis--2013|Stegehuis et al., 2013]] ) shows that these models display overly strong drying in early summer, resulting in an excessive decrease of latent heat fluxes, with potential implications for more severe droughts in dry environments ( [[#Teuling--2018|Teuling, 2018]] ; [[#van%20Der%20Linden--2019|van Der Linden et al., 2019]] ). Compared with a range of observational ET estimates, CMIP5 models show an overestimation of ET on annual scale, but an ET underestimation in boreal summer in many Northern Hemisphere mid-latitude regions, also suggesting a tendency towards excessive soil drying ( [[#Mueller--2014|Mueller and Seneviratne, 2014]] ), consistent with identified biases in soil-moisture–temperature coupling ( [[#Donat--2018|Donat et al., 2018]] ; [[#Vogel--2018|Vogel et al., 2018]] ; [[#Selten--2020|Selten et al., 2020]] ). Land surface models used in ESMs display a bias in their representation of the sensitivity of interannual land carbon uptake to soil moisture conditions, which appears related to a limited range of soil moisture variations compared to observations ( [[#Humphrey--2018|Humphrey et al., 2018]] ).

For future projections, the spread of soil moisture outputs among different ESMs is more important than internal variability and scenario uncertainty, and the bias is strongly related to the sign of the projected change ( [[#Ukkola--2018|Ukkola et al., 2018]] ; [[#Lu--2019|Lu et al., 2019]] ; [[#Selten--2020|Selten et al., 2020]] ). The CMIP5 ESMs that project more drying and warming in mid-latitude regions show a substantial bias in soil-moisture–temperature coupling ( [[#Donat--2018|Donat et al., 2018]] ; [[#Vogel--2018|Vogel et al., 2018]] ). Although CMIP6 and CMIP5 simulations for soil moisture changes are similar overall, some differences are found in projections in a few regions ( [[#11.9|Section 11.9]] ; [[#Cook--2020|Cook et al., 2020]] ). There is still ''limited evidence'' to assess whether there are substantial differences in model performance in the two ensembles, but improvements in modelling aspects relevant for soil moisture have been reported for precipitation ( [[#11.6.3.2|Section 11.6.3.2]] ), and a better performance has been found in CMIP6 for the representation of long-term trends in soil moisture in continental USA ( [[#Yuan--2021|Yuan et al., 2021]] ). Despite the mentioned model limitations, the representation of soil moisture processes in ESMs uses physical and biological understanding of the underlying processes, which can well represent the temporal anomalies associated with temporal variability and trends in climate. In summary, there is ''medium confidence'' in the representation of soil moisture deficits in ESMs and related land surface and hydrological models.

<div id="11.6.3.4" class="h3-container"></div>

<span id="hydrological-deficits-2"></span>
==== 11.6.3.4 Hydrological Deficits ====

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

Streamflow and groundwater are not directly simulated by ESMs, which only simulate runoff, but they are generally represented in hydrological models ( [[#Prudhomme--2014|Prudhomme et al., 2014]] ; [[#Giuntoli--2015|Giuntoli et al., 2015]] ), which are typically driven in a stand-alone manner by observed or simulated climate forcing. The simulation of hydrological deficits is much more problematic than the simulation of mean streamflow or peak flows ( [[#Fundel--2013|Fundel et al., 2013]] ; [[#Stoelzle--2013|Stoelzle et al., 2013]] ; [[#Velázquez--2013|Velázquez et al., 2013]] ; [[#Staudinger--2015|Staudinger et al., 2015]] ), since models tend to be too responsive to the climate forcing and do not satisfactorily capture low flows ( [[#Tallaksen--2014|Tallaksen and Stahl, 2014]] ). Simulations of hydrological drought metrics show uncertainties related to the contribution of both GCMs and hydrological models ( [[#Bosshard--2013|Bosshard et al., 2013]] ; [[#Giuntoli--2015|Giuntoli et al., 2015]] ; [[#Samaniego--2017|Samaniego et al., 2017]] ; [[#Vetter--2017|Vetter et al., 2017]] ), but hydrological models forced by the same climate input data also show a large spread ( [[#van%20Huijgevoort--2013|van Huijgevoort et al., 2013]] ; [[#Ukkola--2018|Ukkola et al., 2018]] ). At the catchment scale, the hydrological model uncertainty is higher than both GCM and downscaling uncertainty ( [[#Vidal--2016|Vidal et al., 2016]] ), and the hydrological models show issues in representing drought propagation throughout the hydrological cycle ( [[#Barella-Ortiz--2019|Barella-Ortiz and Quintana Seguí, 2019]] ). A study on the evaluation of streamflow droughts in seven global (hydrological and land surface) models compared with observations in near-natural catchments of Europe showed a substantial spread among models, an overestimation of the number of drought events, and an underestimation of drought duration and drought-affected area ( [[#Tallaksen--2014|Tallaksen and Stahl, 2014]] ).

<div id="11.6.3.5" class="h3-container"></div>

<span id="atmospheric-based-drought-indices-2"></span>
==== 11.6.3.5 Atmospheric-based Drought Indices ====

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

A number of studies have analysed the ability of models to capture drought severity and trends based on climatic drought indices. Given the limitations of ESMs in reproducing the dynamic of precipitation deficits and AED (11.6.3.1, 11.6.3.2), atmospheric-based drought indices derived from ESM data for these two variables are also affected by uncertainties and biases. A comparison of historical trends in PDSI-PM for 1950–2014 derived from CMIP3 and CMIP5, with respective estimates derived from observations ( [[#Dai--2017|Dai and Zhao, 2017]] ) show a similar behaviour at global scale (long-term decrease), but low spatial agreement in the trends except in a few regions (Mediterranean, South Asia, north-western USA). In future projections, there is an important spread in PDSI-PM and SPEI-PM among different models ( [[#Cook--2014a|Cook et al., 2014a]] ).

<div id="11.6.3.6" class="h3-container"></div>

<span id="synthesis-for-different-drought-types-1"></span>
==== 11.6.3.6 Synthesis for Different Drought Types ====

<div id="h3-18-siblings" class="h3-siblings"></div>

The performance of ESMs used to assessed changes in variables related to meteorological droughts, agricultural and ecological droughts, and hydrological droughts, shows the presence of biases and uncertainties compared to observations, but there is ''medium confidence'' in their overall performance for assessing drought projections given process understanding. Given the substantial inter-model spread documented for all related variables, the consideration of multi-model projections increases the confidence of model-based assessments, with only ''low confidence'' in assessments based on single models.

In summary, the evaluation of ESMs, land surface and hydrological models for the simulation of droughts is complex, due to the regional scale of drought trends, their overall low signal-to-noise ratio, and the lack of observations in several regions, in particular for soil moisture and streamflow. There is ''medium confidence'' in the ability of ESMs to simulate trends and anomalies in precipitation deficits and AED, and also ''medium confidence'' in the ability of ESMs and hydrological models to simulate trends and anomalies in soil moisture and streamflow deficits, on global and regional scales.

<div id="11.6.4" class="h2-container"></div>

<span id="detection-and-attribution-event-attribution-3"></span>
=== 11.6.4 Detection and Attribution, Event Attribution ===

<div id="h2-42-siblings" class="h2-siblings"></div>

<div id="11.6.4.1" class="h3-container"></div>

<span id="precipitation-deficits-3"></span>
==== 11.6.4.1 Precipitation Deficits ====

<div id="h3-19-siblings" class="h3-siblings"></div>

There are only two AR6 regions where there is at least ''medium confidence'' that human-induced climate change has contributed to changes in meteorological droughts ( [[#11.9|Section 11.9]] ). In South-Western South America, there is ''medium confidence'' that human-induced climate change has contributed to an increase in meteorological droughts ( [[#Boisier--2016|Boisier et al., 2016]] ; [[#Garreaud--2020|Garreaud et al., 2020]] ), while in Northern Europe, there is ''medium confidence'' that it has contributed to a decrease in meteorological droughts ( [[#11.9|Section 11.9]] ; [[#Gudmundsson--2016|Gudmundsson and Seneviratne, 2016]] ). In other AR6 regions, there is inconclusive evidence in the attribution of long-term trends, but a human contribution to single meteorological events or sub-regional trends has been identified in some instances ( [[#11.9|Section 11.9]] ; see also below). In the Mediterranean region, some studies have identified a precipitation decline or increase in meteorological drought probability for time frames since the early or mid 20th century, and a possible human contribution to these trends ( [[#Hoerling--2012|Hoerling et al., 2012]] ; [[#Gudmundsson--2016|Gudmundsson and Seneviratne, 2016]] ; [[#Knutson--2018|Knutson and Zeng, 2018]] ), also on sub-regional scale in Syria from 1930 to 2010 ( [[#Kelley--2015|Kelley et al., 2015]] ). On the contrary, other studies have not identified precipitation and meteorological drought trends in the region for the long term ( [[#Camuffo--2013|Camuffo et al., 2013]] ; [[#Paulo--2016|Paulo et al., 2016]] ; [[#Vicente-Serrano--2021|Vicente-Serrano et al., 2021]] ) and also from the mid 20th century ( [[#Norrant--2006|Norrant and Douguédroit, 2006]] ; [[#Stagge--2017|Stagge et al., 2017]] ). There is evidence of substantial internal variability in long-term precipitation trends in the region ( [[#11.6.2.1|Section 11.6.2.1]] ), which limits the attribution of human influence on variability and trends of meteorological droughts from observational records ( [[#Kelley--2012|Kelley et al., 2012]] ; [[#Peña-Angulo--2020b|Peña-Angulo et al., 2020b]] ). In addition, there are important sub-regional trends showing mixed signals ( [[#11.9|Section 11.9]] ; [[#MedECC--2020|MedECC, 2020]] ). The evidence thus leads to an assessment of ''low confidence'' in the attribution of observed short-term changes in meteorological droughts in the region ( [[#11.9|Section 11.9]] ). In North America, the human influence on precipitation deficits is complex ( [[#Wehner--2017|Wehner et al., 2017]] ), with ''low confidence'' in the attribution of long-term changes in meteorological drought in AR6 regions ( [[#11.9|Section 11.9]] ; [[#Lehner--2018|Lehner et al., 2018]] ). In Africa there is ''low confidence'' that human influence has contributed to the observed long-term meteorological drought increase in Western Africa (Sections 11.9 and 10.6.2). There is ''low confidence'' in the attribution of the observed increasing trends in meteorological drought in East Southern Africa, but evidence that human-induced climate change has affected recent meteorological drought events in the region ( [[#11.9|Section 11.9]] ).

Attribution studies for recent meteorological drought events are available for various regions. In Western and Central Europe, a multi-method and multi-model attribution study on the 2015 Central European drought did not find conclusive evidence for whether human-induced climate change was a driver of the rainfall deficit, as the results depended on model and method used ( [[#Hauser--2017|Hauser et al., 2017]] ). In the Mediterranean region, a human contribution was found in the case of the 2014 meteorological drought in the southern Levant based on a single-model study ( [[#Bergaoui--2015|Bergaoui et al., 2015]] ). In Africa, there is some evidence of a contribution of human emissions to single meteorological drought events, such as the 2015–2017 southern African drought ( [[#Funk--2018a|Funk et al., 2018a]] ; [[#Yuan--2018a|Yuan et al., 2018a]] ; [[#Pascale--2020|Pascale et al., 2020]] ), and the three-year (2015–2017) drought in the western Cape Town region of South Africa ( [[#Otto--2018c|Otto et al., 2018c]] ). An attributable signal was not found in droughts that occurred in different years with different spatial extents in the last decade in North and South Eastern Africa ( [[#Marthews--2015|Marthews et al., 2015]] ; [[#Uhe--2017|Uhe et al., 2017]] ; [[#Otto--2018a|Otto et al., 2018a]] ; [[#Philip--2018b|Philip et al., 2018b]] ; [[#Kew--2021|Kew et al., 2021]] ). However, an attributable increase in 2011 long rain failure was identified ( [[#Lott--2013|Lott et al., 2013]] ). Further studies have attributed some African meteorological drought events to large-scale modes of variability, such as the strong 2015 El Niño (Box 11.4; [[#Philip--2018b|Philip et al., 2018b]] ) and increased SSTs overall ( [[#Funk--2015a|Funk et al., 2015a]] , 2018b). Natural variability was dominant in the California droughts of 2011–2012 to 2013–2014 ( [[#Seager--2015a|Seager et al., 2015a]] ). In Asia, no climate change signal was found in the record dry spell over Singapore and Malaysia in 2014 ( [[#Mcbride--2015|Mcbride et al., 2015]] ) or the drought in central south-west Asia in 2013–2014 ( [[#Barlow--2015|Barlow and Hoell, 2015]] ). Nevertheless, the South East Asia drought of 2015 has been attributed to anthropogenic warming effects ( [[#Shiogama--2020|Shiogama et al., 2020]] ). Recent droughts occurring in South America, specifically in the southern Amazon region in 2010 ( [[#Shiogama--2013|Shiogama et al., 2013]] ) and in north-east South America in 2014 ( [[#Otto--2015b|Otto et al., 2015b]] ) and 2016 ( [[#Martins--2018|Martins et al., 2018]] ) were not attributed to anthropogenic climate change. Nevertheless, the central Chile drought between 2010 and 2018 has been suggested to be partly associated to global warming ( [[#Boisier--2016|Boisier et al., 2016]] ; [[#Garreaud--2020|Garreaud et al., 2020]] ). The 2013 New Zealand meteorological drought was attributed to human influence by Harrington et al. (2014, 2016) based on fully coupled CMIP5 models, but no corresponding change in the dry end of simulated precipitation from a stand-alone atmospheric model was found by [[#Angélil--2017|Angélil et al. (2017)]] .

Event attribution studies also highlight a complex interplay of anthropogenic and non-anthropogenic climatological factors for some events. For example, anthropogenic warming contributed to the 2014 drought in North Eastern Africa by increasing east African and west Pacific temperatures, and increasing the gradient between standardized western and central Pacific SSTs, causing reduced rainfall ( [[#Funk--2015a|Funk et al., 2015a]] ). As different methodologies, models and data sources have been used for the attribution of precipitation deficits, [[#Angélil--2017|Angélil et al. (2017)]] re-examined several events using a single analytical approach and climate model and observational datasets. Their results showed a disagreement in the original anthropogenic attribution in a number of precipitation deficit events, which increased uncertainty in the attribution of meteorological droughts events.

<div id="11.6.4.2" class="h3-container"></div>

<span id="soil-moisture-deficits-3"></span>
==== 11.6.4.2 Soil Moisture Deficits ====

<div id="h3-20-siblings" class="h3-siblings"></div>

There is a growing number of studies on the detection and attribution of long-term changes in soil moisture deficits. [[#Mueller--2016|Mueller and Zhang (2016)]] concluded that anthropogenic forcing contributed significantly to soil moisture drying in the warm season in the Northern Hemisphere from 1951 to 2005 and also led to an increase in the land surface area affected by soil moisture deficits, which can be reproduced by CMIP5 models only if anthropogenic forcings are involved. [[#Gu--2019b|Gu et al. (2019b)]] similarly identified a global-scale soil moisture drying tendency in land surface model data from the Global Land Data Assimilation System 2 over the time frame 1948–2005, which was attributed to anthropogenic forcing based on evaluation with CMIP5 models using optimal fingerprinting. [[#Padrón--2019|Padrón et al. (2019)]] analysed long-term reconstructed and CMIP5 simulated dry season water availability, defined as precipitation minus ET (i.e., equivalent to soil moisture and runoff availability), also related to agricultural and ecological droughts. They found an intensification of dry-season precipitation minus evapotranspiration deficits over a predominant fraction of the land area in the last three decades, which can only be explained by anthropogenic forcing and is mostly related to increases in ET. Similarly, [[#Williams--2020|Williams et al. (2020)]] concluded that human-induced climate change contributed to the strong soil moisture deficits recorded in the last two decades in Western North America through VPD increases associated with higher air temperatures and lower air humidity. There are few studies analysing the attribution of particular episodes of soil moisture deficits to anthropogenic influence. Nevertheless, the available modelling studies coincide in supporting an anthropogenic attribution associated with more extreme temperatures, exacerbating AED and increasing ET, and thus depleting soil moisture, as observed in southern Europe in 2017 ( [[#García-Herrera--2019|García-Herrera et al., 2019]] ) and in Australia in 2018 ( [[#Lewis--2020|Lewis et al., 2020]] ) and 2019 ( [[#van%20Oldenborgh--2021|van Oldenborgh et al., 2021]] ), the latter event having strong implications in the propagation of widespread megafires ( [[#Nolan--2020|Nolan et al., 2020]] ).

<div id="11.6.4.3" class="h3-container"></div>

<span id="hydrological-deficits-3"></span>
==== 11.6.4.3 Hydrological Deficits ====

<div id="h3-21-siblings" class="h3-siblings"></div>

It is often difficult to separate the role of climate trends from changes in land use, water management and demand for changes in hydrological deficits, especially on a regional scale. However, a global study based on a recent multi-model experiment with global hydrological models and covering several AR6 regions suggests a dominant role of anthropogenic radiative forcing for trends in low, mean and high flows, while simulated effects of water and land management do not suffice to reproduce the observed spatial pattern of trends ( [[#Gudmundsson--2021|Gudmundsson et al., 2021]] ). Regional studies also suggest that climate trends have been dominant compared to land use and human water management for explaining trends in hydrological droughts in some regions, for instance in Ethiopia ( [[#Fenta--2017|Fenta et al., 2017]] ), China ( [[#Xie--2015|Xie et al., 2015]] ), and North America for the Missouri and Colorado basins, as well as in California ( [[#Shukla--2015|Shukla et al., 2015]] ; [[#Udall--2017|Udall and Overpeck, 2017]] ; [[#Ficklin--2018|Ficklin et al., 2018]] ; K. [[#Xiao--2018|]] [[#Xiao--2018|Xiao et al., 2018]] ; [[#Glas--2019|Glas et al., 2019]] ; [[#Martin--2020|Martin et al., 2020]] ; [[#Milly--2020|Milly and Dunne, 2020]] ).

In other regions, the influence of human water uses can be more important to explain hydrological drought trends (Y. [[#Liu--2016|]] [[#Liu--2016|Liu et al., 2016]] ; [[#Mohammed--2016|Mohammed and Scholz, 2016]] ). There is ''medium confidence'' that human-induced climate change has contributed to an increase of hydrological droughts in the Mediterranean ( [[#Giuntoli--2013|Giuntoli et al., 2013]] ; [[#Vicente-Serrano--2014|Vicente-Serrano et al., 2014]] ; [[#Gudmundsson--2017|Gudmundsson et al., 2017]] ), but also ''medium confidence'' that changes in land use and terrestrial water management contributed to these trends ( [[#11.9|Section 11.9]] ; [[#Teuling--2019|Teuling et al., 2019]] ; [[#Vicente-Serrano--2019|Vicente-Serrano et al., 2019]] ). A global study with a single hydrological model estimated that human water consumption has intensified the magnitude of hydrological droughts by 20–40% over the last 50 years, and that the human water use contribution to hydrological droughts was more important than climatic factors in the Mediterranean, and central USA, as well as in parts of Brazil ( [[#Wada--2013|Wada et al., 2013]] ). However, [[#Gudmundsson--2021|Gudmundsson et al. (2021)]] concluded that the contribution of human water use is smaller than that of anthropogenic climate change to explain spatial differences in the trends of low flows based on a multi-model analysis. There is still ''limited evidence'' and thus ''low confidence'' in assessing these trends at the scale of single regions, with few exceptions ( [[#11.9|Section 11.9]] ).

<div id="11.6.4.4" class="h3-container"></div>

<span id="atmospheric-based-drought-indices-3"></span>
==== 11.6.4.4 Atmospheric-based Drought Indices ====

<div id="h3-22-siblings" class="h3-siblings"></div>

Different studies using atmospheric-based drought indices suggest an attributable anthropogenic signal, characterized by the increased frequency and severity of droughts ( [[#Cook--2018|Cook et al., 2018]] ), associated to increased AED ( [[#11.6.4.2|Section 11.6.4.2]] ). The majority of studies are based on the PDSI-PM. [[#Williams--2015|Williams et al. (2015)]] and [[#Griffin--2014|Griffin and Anchukaitis (2014)]] concluded that increased AED has had an increased contribution to drought severity over the last decades, and played a dominant role in the intensification of the 2012–2014 drought in California. The same temporal pattern and physical mechanism was stressed by Z. [[#Li--2017|]] [[#Li--2017|]] [[#Li--2017|]] [[#Li--2017|Li et al. (2017)]] in central Asia. [[#Marvel--2019|Marvel et al. (2019)]] compared tree ring-based reconstructions of the PDSI-PM over the past millennium with PDSI-PM estimates based on output from CMIP5 models. The comparisons suggested a contribution of greenhouse gas forcing to the changes since the beginning of the 20th century, although characterized with temporal differences that could be driven by temporal variations in the aerosol forcing. This was in agreement with the dominant external forcings of aridification at global scale between 1950 and 2014 ( [[#Bonfils--2020|Bonfils et al., 2020]] ). In the Mediterranean region, there is ''medium confidence'' of drying attributable to antropogenic forcing as a consequence of the strong AED increase ( [[#Gocic--2014|Gocic and Trajkovic, 2014]] ; [[#Azorin-Molina--2015|Azorin-Molina et al., 2015]] ; [[#Liuzzo--2016|Liuzzo et al., 2016]] ; [[#Maček--2018|Maček et al., 2018]] ), which has enhanced the severity of drought events ( [[#Vicente-Serrano--2014|Vicente-Serrano et al., 2014]] ; [[#Stagge--2017|Stagge et al., 2017]] ; [[#González-Hidalgo--2018|González-Hidalgo et al., 2018]] ). In particular, this effect was identified to be the main driver of the intensification of the 2017 drought that affected south-western Europe, and was attributed to the human forcing ( [[#García-Herrera--2019|García-Herrera et al., 2019]] ). [[#Nangombe--2020|Nangombe et al. (2020)]] and L. [[#Zhang--2020|]] [[#Zhang--2020|]] [[#Zhang--2020|]] [[#Zhang--2020|Zhang et al. (2020)]] concluded from differences between precipitation and AED that anthropogenic forcing contributed to the 2018 droughts that affected southern Africa and south-eastern China, respectively, principally as consequence of the high AED that characterized these two events.

<div id="11.6.4.5" class="h3-container"></div>

<span id="synthesis-for-different-drought-types-2"></span>
==== 11.6.4.5 Synthesis for Different Drought Types ====

<div id="h3-23-siblings" class="h3-siblings"></div>

The regional evidence on attribution for single AR6 regions generally shows ''low confidence'' for a human contribution to observed trends in meteorological droughts at regional scale, with few exceptions ( [[#11.9|Section 11.9]] ). There is ''medium confidence'' that human influence has contributed to increases in agricultural and ecological droughts in the dry season in some regions and has led to an overall increase in the affected land area. At regional scales, there is ''medium confidence'' in a contribution of human-induced climate change to increases in agricultural and ecological droughts in the Mediterranean and Western North America ( [[#11.9|Section 11.9]] ). There ''is medium confidence'' that human-induced climate change has contributed to an increase in hydrological droughts in the Mediterranean region, but also ''medium confidence'' in contributions from other human influences, including water management and land use ( [[#11.9|Section 11.9]] ). Several meteorological and agricultural and ecological drought events have been attributed to human-induced climate change, even in regions where no long-term changes are detected ( ''medium confidence'' ). However, a lack of attribution to human-induced climate change has also been shown for some events ( ''medi'' ''um confidence'' ).

In summary, human influence has contributed to increases in agricultural and ecological droughts in the dry season in some regions due to increases in evapotranspiration ( ''medium confidence'' ). The increases in evapotranspiration have been driven by increases in atmospheric evaporative demand induced by increased temperature, decreased relative humidity and increased net radiation over affected land areas ( ''high confidence'' ). There is ''low confidence'' that human influence has affected trends in meteorological droughts in most regions, but ''medium confidence'' that they have contributed to the severity of some single events. There is ''medium confidence'' that human-induced climate change has contributed to increasing trends in the probability or intensity of recent agricultural and ecological droughts, leading to an increase of the affected land area. Human-induced climate change has contributed to global-scale change in low flow, but human water management and land-use changes are also important drivers ( ''medi'' ''um confidence'' ).

<div id="11.6.5" class="h2-container"></div>

<span id="projections-2"></span>
=== 11.6.5 Projections ===

<div id="h2-43-siblings" class="h2-siblings"></div>

The SREX (Chapter 3) asssessed with ''medium confidence'' projections of increased drought severity in some regions, including southern Europe and the Mediterranean, central Europe, central America and Mexico, north-east Brazil, and southern Africa, and ''low confidence'' elsewhere given large inter-model spread. The AR5 (Chapters 11 and 12) also assessed large uncertainties in drought projections at the regional and global scales. The assessment of drought mechanisms under future climate change scenarios depends on the model used ( [[#11.6.3|Section 11.6.3]] ). Moreover, uncertainties in drought projections are affected by the consideration of plant physiological responses to increasing atmospheric CO <sub>2</sub> (Cross-Chapter Box 5.1; [[#Milly--2016|Milly and Dunne, 2016]] ; [[#Greve--2019|Greve et al., 2019]] ; [[#Mankin--2019|Mankin et al., 2019]] ; [[#Yang--2020|Yang et al., 2020]] ), the role of soil-moisture–atmosphere feedbacks for changes in water balance and aridity ( [[#Berg--2016|Berg et al., 2016]] ; [[#Zhou--2021|Zhou et al., 2021]] ), and statistical issues related to considered drought time scales ( [[#Vicente-Serrano--2020c|Vicente-Serrano et al., 2020c]] ). Nonetheless, the extensive literature available since AR5 allows a substantially more robust assessment of projected changes in droughts, also subdivided in different drought types (meteorological drought, agricultural and ecological drought, and hydrological drought). This includes assessments of projected changes in droughts, including changes at 1.5°C, 2°C and 4°C of global warming, for all AR6 regions ( [[#11.9|Section 11.9]] ). Projected changes show increases in drought frequency and intensity in several regions as function of global warming ( ''high confidence'' ). There are also substantial increases in drought hazard probability from 1.5°C to 2°C global warming and for further additional increments of global warming ( ''high confidence'' ) (Figures 11.18 and 11.19). These findings are based on both CMIP5 and CMIP6 analyses ( [[#11.9|Section 11.9]] ; [[#Wartenburger--2017|Wartenburger et al., 2017]] ; [[#Greve--2018|Greve et al., 2018]] ; L. [[#Xu--2019|]] [[#Xu--2019|Xu et al., 2019]] ), and strengthen the conclusions of SR1.5 Chapter 3.

<div id="11.6.5.1" class="h3-container"></div>

<span id="precipitation-deficits-4"></span>
==== 11.6.5.1 Precipitation Deficits ====

<div id="h3-24-siblings" class="h3-siblings"></div>

Studies based on CMIP5, CMIP6 and Coordinated Regional Climate Downscaling Experiment (CORDEX) projections show a consistent signal in the sign and spatial pattern of projections of precipitation deficits. Global studies based on these multi-model ensemble projections ( [[#Orlowsky--2013|Orlowsky and Seneviratne, 2013]] ; [[#Martin--2018|Martin, 2018]] ; [[#Spinoni--2020|Spinoni et al., 2020]] ; [[#Ukkola--2020|Ukkola et al., 2020]] ; [[#Coppola--2021b|Coppola et al., 2021b]] ) show particularly strong signal-to-noise ratios for increasing meteorological droughts in the following AR6 regions: MED, ESAF, WSAF, SAU, CAU, NCA, SCA, NSA and NES ( [[#11.9|Section 11.9]] ). There is also substantial evidence of changes in meteorological droughts at 1.5°C versus 2°C of global warming from global studies ( [[#Wartenburger--2017|Wartenburger et al., 2017]] ; L. [[#Xu--2019|]] [[#Xu--2019|Xu et al., 2019]] ). The patterns of projected changes in mean precipitation are consistent with the changes in the drought duration, but they are not consistent with the changes in drought intensity ( [[#Ukkola--2020|Ukkola et al., 2020]] ). In general, CMIP6 projections suggest a stronger increase of the probability of precipitation deficits than CMIP5 projections ( [[#Cook--2020|Cook et al., 2020]] ; [[#Ukkola--2020|Ukkola et al., 2020]] ). Projections for the number of CDDs in CMIP6 (Figure 11.19) for different levels of global warming relative to 1850–1900 show similar spatial patterns as projected precipitation deficits. The robustness of the patterns in projected precipitation deficits identified in the global studies is also consistent with results from regional studies ( [[#Giorgi--2014|Giorgi et al., 2014]] ; [[#Marengo--2016|Marengo and Espinoza, 2016]] ; [[#Pinto--2016|Pinto et al., 2016]] ; J. [[#Huang--2018|]] [[#Huang--2018|Huang et al., 2018]] ; [[#Maúre--2018|Maúre et al., 2018]] ; [[#Nangombe--2018|Nangombe et al., 2018]] ; [[#Tabari--2018|Tabari and Willems, 2018]] ; [[#Abiodun--2019|Abiodun et al., 2019]] ; [[#Dosio--2019|Dosio et al., 2019]] ).

In Africa, a strong increase in the length of dry spells (CDD) is projected for 4°C of global warming over most of the continent, with the exception of central and eastern Africa ( [[#11.9|Section 11.9]] ; [[#Sillmann--2013a|Sillmann et al., 2013a]] ; [[#Giorgi--2014|Giorgi et al., 2014]] ; [[#Han--2019|Han et al., 2019]] ). In West Africa, a strong reduction of precipitation is projected ( [[#Sillmann--2013a|Sillmann et al., 2013a]] ; [[#Diallo--2016|Diallo et al., 2016]] ; [[#Akinsanola--2019|Akinsanola and Zhou, 2019]] ; [[#Han--2019|Han et al., 2019]] ; [[#Todzo--2020|Todzo et al., 2020]] ) at 4°C of global warming, and CDD would increase with stronger global warming levels ( [[#Klutse--2018|Klutse et al., 2018]] ). The regions most strongly affected are southern Africa (ESAF, WSAF) ( [[#Nangombe--2018|Nangombe et al., 2018]] ; [[#Abiodun--2019|Abiodun et al., 2019]] ) and northern Africa (part of the MED region), with increases in meteorological droughts already at 1.5°C of global warming, and further increases with increasing global warming ( [[#11.9|Section 11.9]] ). CDD is projected to increase more in the southern Mediterranean (northern Africa) than in the northern part of the Mediterranean region ( [[#Lionello--2020|Lionello and Scarascia, 2020]] ).

In Asia, most AR6 regions show ''low confidence'' in projected changes in meteorological droughts at 1.5°C and 2°C of global warming, with a few regions displaying a decrease in meteorological droughts at 4°C of global warming (RAR, ESB, RFE, ECA; ''medium confidence'' ), although there is a projected increase in meteorological droughts in South East Asia at 4°C ( ''medium confidence'' ) ( [[#11.9|Section 11.9]] ). In South East Asia, an increasing frequency of precipitation deficits is projected as a consequence of an increasing frequency of extreme El Niño ( [[#Cai--2014b|Cai et al., 2014b]] , 2015, 2018).

In Central America, projections suggest an increase in mid-summer meteorological drought ( [[#Imbach--2018|Imbach et al., 2018]] ) and increased CDD ( [[#Chou--2014a|Chou et al., 2014a]] ; [[#Giorgi--2014|Giorgi et al., 2014]] ; [[#Nakaegawa--2014|Nakaegawa et al., 2014]] ). In the Amazon, there is also a projected increase in dryness ( [[#Marengo--2016|Marengo and Espinoza, 2016]] ), which is the combination of a projected increase in the frequency and geographic extent of meteorological drought in the eastern Amazon, and an opposite trend in the west ( [[#Duffy--2015|Duffy et al., 2015]] ). In South-Western South America, there is a projected increase of CDD ( [[#Chou--2014a|Chou et al., 2014a]] ; [[#Giorgi--2014|Giorgi et al., 2014]] ) and in Chile, drying is projected to prevail ( [[#Boisier--2018|Boisier et al., 2018]] ). In the South America monsoon region, an increase in CDD is projected ( [[#Chou--2014a|Chou et al., 2014a]] ; [[#Giorgi--2014|Giorgi et al., 2014]] ), but a decrease is projected in South-Eastern and Southern South America ( [[#Giorgi--2014|Giorgi et al., 2014]] ). In Central America, mid-summer meteorological drought is projected to intensify during 2071–2095 for the RCP8.5 scenario ( [[#Corrales-Suastegui--2020|Corrales‐Suastegui et al., 2020]] ).

An increase in the frequency, duration and intensity of meteorological droughts is projected in south-west, south and east Australia ( [[#Kirono--2020|Kirono et al., 2020]] ; [[#Shi--2020|Shi et al., 2020]] ). In Canada and most of the USA, based on the SPI, [[#Swain--2015|Swain and Hayhoe (2015)]] identified drier summer conditions in projections over most of the region, and there is a consistent signal toward an increase in duration and intensity of droughts in southern North America ( [[#Pascale--2016|Pascale et al., 2016]] ; [[#Escalante-Sandoval--2017|Escalante-Sandoval and Nuñez-Garcia, 2017]] ). In California, more precipitation variability is projected, characterized by increased frequency of consecutive drought and humid periods ( [[#Swain--2018|Swain et al., 2018]] ).

Substantial increases in meteorological drought are projected in Europe, in particular in the Mediterranean region, already at 1.5°C of global warming ( [[#11.9|Section 11.9]] ). In southern Europe, model projections display a consistent drying among models ( [[#Russo--2013|Russo et al., 2013]] ; [[#Hertig--2017|Hertig and Tramblay, 2017]] ; [[#Guerreiro--2018a|Guerreiro et al., 2018a]] ; [[#Raymond--2019|Raymond et al., 2019]] ). In Western and Central Europe there is some spread in CMIP5 projections, with some models projecting very strong drying, and others close to no trend ( [[#Vogel--2018|Vogel et al., 2018]] ), although CDD is projected to increase in CMIP5 projections under the RCP 8.5 scenario ( [[#Hari--2020|Hari et al., 2020]] ). The overall evidence suggests an increase in meteorological drought at 4°C in the WCE region ( ''medium confidence'' ) ( [[#11.9|Section 11.9]] ).

Overall, based on global and regional studies, several hot spot regions are identified, displaying more frequent and severe meteorological droughts with increasing global warming, including several AR6 regions at 1.5°C (WSAF, ESAF, SAU, MED, NES) and 2°C of global warming (WSAF, ESAF, EAU, SAU, MED, NCA, SCA, NSA, NES) ( [[#11.9|Section 11.9]] ). At 4°C of global warming, there is also ''confidence'' in increases in meteorological droughts in further regions (WAF, WCE, ENA, CAR, NWS, SAM, SWS, SSA; [[#11.9|Section 11.9]] ), showing a geographical expansion of meteorological drought with increasing global warming. Only few regions are projected to have less intense or frequent meteorological droughts ( [[#11.9|Section 11.9]] ).

<div id="11.6.5.2" class="h3-container"></div>

<span id="atmospheric-evaporative-demand-3"></span>
==== 11.6.5.2 Atmospheric Evaporative Demand ====

<div id="h3-25-siblings" class="h3-siblings"></div>

Effects of AED on droughts in future projections is under debate. The CMIP5 models project an increase in AED over the majority of the world with increasing global warming, mostly as a consequence of strong VPD increases ( [[#Scheff--2015|Scheff and Frierson, 2015]] ; [[#Vicente-Serrano--2020a|Vicente-Serrano et al., 2020a]] ). However, ET is projected to increase less than AED in many regions due to plant physiological responses related to: i) CO <sub>2</sub> effects on plant photosynthesis; and ii) soil moisture control on ET.

Several studies suggest that increasing atmospheric CO <sub>2</sub> could lead to reduced leaf stomatal conductance, which would increase water-use efficiency and reduce plant water needs, thus limiting ET (Cross-Chapter Box 5.1; [[#Roderick--2015|Roderick et al., 2015]] ; [[#Milly--2016|Milly and Dunne, 2016]] ; [[#Swann--2016|Swann et al., 2016]] ; [[#Greve--2017|Greve et al., 2017]] ; [[#Scheff--2017|Scheff et al., 2017]] ; [[#Lemordant--2018|Lemordant et al., 2018]] ; [[#Swann--2018|Swann, 2018]] ). The implemention of a CO <sub>2</sub> -dependent land resistance parameter has been suggested for the estimation of AED ( [[#Yang--2019|Yang et al., 2019]] ). Nevertheless, there are other relevant mechanisms, as soil moisture deficits and VPD also play an important role in the control of the leaf stomatal conductance (Z. [[#Xu--2016|]] [[#Xu--2016|Xu et al., 2016]] ; [[#Menezes-Silva--2019|Menezes-Silva et al., 2019]] ; [[#Grossiord--2020|Grossiord et al., 2020]] ), and a number of ecophysiological and anatomical processes affect the response of plant physiology under higher atmospheric CO <sub>2</sub> concentrations (Cross-Chapter Box 5.1; [[#Mankin--2019|Mankin et al., 2019]] ; [[#Menezes-Silva--2019|Menezes-Silva et al., 2019]] ). The benefits of the atmospheric CO <sub>2</sub> for plant stress and agricultural and ecological droughts would be minimal precisely during dry periods given stomatal closure in response to limited soil moisture ( [[#Allen--2015|Allen et al., 2015]] ; Z. [[#Xu--2016|]] [[#Xu--2016|Xu et al., 2016]] ). In addition, CO <sub>2</sub> effects on plant stomatal conductance could not entirely compensate for the increased demand associated with warming ( [[#Liu--2017|Liu and Sun, 2017]] ); in large tropical and subtropical regions (e.g., southern Africa, the Amazon, the Mediterranean and southern North America), AED is projected to increase, even considering the possible CO <sub>2</sub> effects on land resistance ( [[#Vicente-Serrano--2020a|Vicente-Serrano et al., 2020a]] ). Moreover, these CO <sub>2</sub> effects would not affect the direct evaporation from soil and water bodies, which is very relevant in the reservoirs of warm areas ( [[#Friedrich--2018|Friedrich et al., 2018]] ). Because of these uncertainties, there is ''low confidence'' whether increased CO <sub>2</sub> -induced water-use efficiency in vegetation will substantially reduce global plant transpiration and will diminish the frequency and severity of soil moisture and streamflow deficits associated with the radiative effect of higher CO <sub>2</sub> concentrations (Cross-Chapter Box 5.1).

Another mechanism reducing the ET response to increased AED in projections is the control of soil moisture limitations on ET, which leads to reduced stomatal conductance under water stress ( [[#Berg--2018|Berg and Sheffield, 2018]] ; [[#Stocker--2018|Stocker et al., 2018]] ; [[#Zhou--2021|Zhou et al., 2021]] ). This response may be further amplified through VPD-induced decreases in stomatal conductance ( [[#Anderegg--2020|Anderegg et al., 2020]] ). However, the decreased stomatal conductance in response to soil moisture limitation and enhanced CO <sub>2</sub> would further enhance AED ( [[#Sherwood--2014|Sherwood and Fu, 2014]] ; [[#Berg--2016|Berg et al., 2016]] ; [[#Teuling--2018|Teuling, 2018]] ; [[#Miralles--2019|Miralles et al., 2019]] ), whereby the overall effects on AED in ESMs are found to be of similar magnitude for soil moisture limitation and CO <sub>2</sub> physiological effects on stomatal conductance ( [[#Berg--2016|Berg et al., 2016]] ). Increased AED is thus both a driver and a feedback with respect to changes in ET, complicating the interpretation of its role on drought changes with increasing CO <sub>2</sub> concentrations and global warming.

<div id="11.6.5.3" class="h3-container"></div>

<span id="soil-moisture-deficits-4"></span>
==== 11.6.5.3 Soil Moisture Deficits ====

<div id="h3-26-siblings" class="h3-siblings"></div>

Areas with projected soil moisture decreases do not fully coincide with areas that have projected precipitation decreases, although there is substantial consistency in the respective patterns ( [[#Dirmeyer--2013|Dirmeyer et al., 2013]] ; [[#Berg--2018|Berg and Sheffield, 2018]] ). However, there are more regions affected by increased soil moisture deficits (Figure 11.19) than precipitation deficits (Figures 2a,b,c and Cross-Chapter Box 11.1) as a consequence of enhanced AED and the associated increased ET, as highlighted by some studies ( [[#Orlowsky--2013|Orlowsky and Seneviratne, 2013]] ; [[#Dai--2018|Dai et al., 2018]] ; [[IPCC:Wg1:Chapter:Chapter-8#8.2.2.1|Section 8.2.2.1]] ). Moisture in the top soil layer is projected to decrease more than precipitation at all warming levels ( [[#Lu--2019|Lu et al., 2019]] ), extending the regions affected by severe soil moisture deficits over most of south and central Europe ( [[#Lehner--2017|Lehner et al., 2017]] ; [[#Ruosteenoja--2018|Ruosteenoja et al., 2018]] ; [[#Samaniego--2018|Samaniego et al., 2018]] ; [[#van%20Der%20Linden--2019|van Der Linden et al., 2019]] ), southern North America ( [[#Cook--2019|Cook et al., 2019]] ), South America ( [[#Orlowsky--2013|Orlowsky and Seneviratne, 2013]] ), southern Africa ( [[#Lu--2019|Lu et al., 2019]] ), East Africa ( [[#Rowell--2015|Rowell et al., 2015]] ), Southern Australia ( [[#Kirono--2020|Kirono et al., 2020]] ), India ( [[#Mishra--2014a|Mishra et al., 2014a]] ) and East Asia (Figure 11.19; [[#Cheng--2015|Cheng et al., 2015]] ). Projected changes in total soil moisture display less widespread drying than those for surface soil moisture ( [[#Berg--2017a|Berg et al., 2017a]] ), but still more than for precipitation (Cross-Chapter Box 11.1, Figures 2a,b,c). The severity of droughts based on surface soil moisture in future projections is stronger than projections based on precipitation and runoff ( [[#Dai--2018|Dai et al., 2018]] ; [[#Vicente-Serrano--2020c|Vicente-Serrano et al., 2020c]] ). Nevertheless, in many parts of the world where soil moisture is projected to decrease, the signal-to-noise ratio among models is low; only the projections in the Mediterranean, Europe, the south-western USA, and southern Africa show a high signal-to-noise ratio in soil moisture projections (Figure 11.19; [[#Lu--2019|Lu et al., 2019]] ). Increases in soil moisture deficits are found to be statistically signicant at regional scale in the Mediterranean region, southern Africa and western South America for changes as small as 0.5°C in global warming, based on differences between +1.5°C and +2°C of global warming ( [[#Wartenburger--2017|Wartenburger et al., 2017]] ). Several other regions are affected when considering changes in droughts for higher changes in global warming ( [[#11.9|Section 11.9]] and Figure 11.19). Seasonal projections of drought frequency for boreal winter (December–January–February) and summer (June–July–August), from CMIP6 multi-model ensemble for 1.5°C, 2°C and 4°C global warming levels, show contrasting trends (Figure 11.19). In the boreal winter in the Northern Hemisphere, the areas affected by drying show ''high agreement'' with those characterized by an increase in meteorological drought projections (Figures 8.14 and 12.4). On the contrary, in the boreal summer, the drought frequency increases worldwide in comparison to meteorological drought projections, with large areas of the Northern Hemisphere displaying a high signal-to-noise ratio (low spead between models). This stresses the dominant influence of ET (as a result of increased AED) in intensifying agricultural and ecological droughts in the warm season in many locations, including mid- to high latitudes.

Increased soil moisture limitation and associated changes in droughts are projected to lead to increased vegetation stress affecting the global land carbon sink in ESM projections ( [[#Green--2019|Green et al., 2019]] ), with implications for projected global warming (Cross-Chapter Box 5). There is ''high confidence'' that the global land sink will become less efficient due to soil moisture limitations and associated agricultural and ecological drought conditions in some regions in higher-emissions scenarios, specially under global warming levels above 4°C; however, there is ''low confidence'' in how these water cycle feedbacks will play out in lower-emissions scenarios (at 2°C global warming or lower; Cross-Chapter Box 5.1).

<div id="_idContainer065" class="Basic-Text-Frame"></div>

[[File:a35dd7dfa52068b8566ee043aeb0b54e IPCC_AR6_WGI_Figure_11_18.png]]

'''Figure 11.18 |''' '''Projected changes in (a) the intensity and (b) the frequency of drought under 1°C, 1.5°C, 2°C, 3°C, and 4°C global warming levels relative to the 1850–1900 baseline. (c)''' Summaries are computed for the AR6 regions in which there is at least medium confidence in an increase in agriculture/ecological drought at the 2°C global warming level (‘drying regions’), including Western North America, Central North America, North Central America, Southern Central America, Northern South America, North-Eastern South America, South American Monsoon, South-Western South America, Southern South America, West and Central Europe, Mediterranean, West Southern Africa, East Southern Africa, Madagascar, Eastern Australia, Southern Australia. Caribbean is not included in the calculation because the number of land grid points was too small. A drought event is defined as a 10-year drought event whose annual mean soil moisture was below its 10th percentile from the 1850–1900 base period. For each box plot, the horizontal line and the box represent the median and central 66% uncertainty range, respectively, of the frequency or the intensity changes across the multi-model ensemble, and the ‘whiskers’ extend to the 90% uncertainty range. The line of zero in (a) indicates no change in intensity, while the line of one in (b) indicates no change in frequency. 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 (SSP) forcing scenarios. Intensity changes in (a) are expressed as standard deviations of the interannual variability in the period 1850–1900 of the corresponding model. For details on the methods see Supplementary Material 11.SM.2. Further details on data sources and processing are available in the chapter data table (Table 11.SM.9).

<div id="11.6.5.4" class="h3-container"></div>

<span id="hydrological-deficits-4"></span>
==== 11.6.5.4 Hydrological Deficits ====

<div id="h3-27-siblings" class="h3-siblings"></div>

Some studies support wetting tendencies as a response to a warmer climate when considering globally averaged changes in runoff over land ( [[#Roderick--2015|Roderick et al., 2015]] ; [[#Greve--2017|Greve et al., 2017]] ; Y. [[#Yang--2018|]] [[#Yang--2018|]] [[#Yang--2018|]] [[#Yang--2018|]] [[#Yang--2018|Yang et al., 2018]] ), and streamflow projections respond to enhanced CO <sub>2</sub> concentrations in CMIP5 models ( [[#Yang--2019|Yang et al., 2019]] ). Nevertheless, when focusing regionally on low-runoff periods, model projections also show an increase of hydrological droughts in large world regions ( [[#Wanders--2015|Wanders and Van Lanen, 2015]] ; [[#Dai--2018|Dai et al., 2018]] ; [[#Vicente-Serrano--2020c|Vicente-Serrano et al., 2020c]] ). In general, the frequency of hydrological deficits is projected to increase over most of the continents, although with regionally and seasonally differentiated effects ( [[#11.9|Section 11.9]] ), with ''medium confidence'' of increase in the following AR6 regions: WCE, MED, SAU, WCA, WNA, SCA, NSA, SAM, SWS, SSA, WSAF, ESAF and MDG ( [[#11.9|Section 11.9]] ; [[#Forzieri--2014|Forzieri et al., 2014]] ; [[#Prudhomme--2014|Prudhomme et al., 2014]] ; [[#Giuntoli--2015|Giuntoli et al., 2015]] ; [[#Wanders--2015|Wanders and Van Lanen, 2015]] ; [[#Roudier--2016|Roudier et al., 2016]] ; [[#Marx--2018|Marx et al., 2018]] ; [[#Cook--2019|Cook et al., 2019]] ; [[#Zhao--2020|Zhao et al., 2020]] ). However, there are large uncertainties related to the hydrological/impact model used ( [[#Prudhomme--2014|Prudhomme et al., 2014]] ; [[#Schewe--2014|Schewe et al., 2014]] ; [[#Gosling--2017|Gosling et al., 2017]] ), limited signal-to-noise ratio (due to model spread) in several regions ( [[#Giuntoli--2015|Giuntoli et al., 2015]] ), and also uncertainties in the projection of future human activities, including water demand and land cover changes, which may represent more than 50% of the projected changes in hydrological droughts in some regions ( [[#Wanders--2015|Wanders and Wada, 2015]] ).

Regions dependent on mountainous snowpack as a temporary reservoir may be affected by severe hydrological droughts in a warmer world. In the southern European Alps, both winter and summer low flows are projected to be more severe, with a 25% decrease in the 2050s ( [[#Vidal--2016|Vidal et al., 2016]] ). In western USA, a 22% reduction in winter snow water equivalent is projected at around 2°C of global warming, with a further decrease of a 70% reduction at 4°C global warming ( [[#Rhoades--2018|Rhoades et al., 2018]] ). This decline would cause less predictable hydrological droughts in snowmelt-dominated areas of North America ( [[#Livneh--2020|Livneh and Badger, 2020]] ). The exact magnitude of the influence of higher temperatures on snow-related droughts is, however, difficult to estimate ( [[#Mote--2016|Mote et al., 2016]] ), since the streamflow changes could affect the timing of peak streamflows but not necessarily their magnitude. In addition, projected changes in hydrological droughts downstream of declining glaciers can be very complex to assess (Chapter 9, see also SROCC).

<div id="11.6.5.5" class="h3-container"></div>

<span id="atmospheric-based-drought-indices-4"></span>
==== 11.6.5.5 Atmospheric-based Drought Indices ====

<div id="h3-28-siblings" class="h3-siblings"></div>

Studies show a stronger drying in projections based on atmospheric-based drought indices compared to ESM projections of changes in soil moisture ( [[#Berg--2018|Berg and Sheffield, 2018]] ) and runoff ( [[#Yang--2019|Yang et al., 2019]] ). It has been suggested that this difference is due to physiological CO <sub>2</sub> effects ( [[#11.6.5.2|Section 11.6.5.2]] ; [[#Roderick--2015|Roderick et al., 2015]] ; [[#Milly--2016|Milly and Dunne, 2016]] ; [[#Swann--2016|Swann et al., 2016]] ; [[#Lemordant--2018|Lemordant et al., 2018]] ; [[#Scheff--2018|Scheff, 2018]] ; [[#Swann--2018|Swann, 2018]] ; [[#Greve--2019|Greve et al., 2019]] ; [[#Yang--2020|Yang et al., 2020]] ). Nonetheless, there is evidence that differences in projections between atmospheric-based drought indices and water-balance metrics from ESMs are not alone due to CO <sub>2</sub> -plant effects ( [[#Berg--2016|Berg et al., 2016]] ; [[#Scheff--2021|Scheff et al., 2021]] ). Differences can also be related to the fact that AED is an upper bound for ET in dry regions and conditions ( [[#11.6.1.2|Section 11.6.1.2]] ) and that soil moisture stress limits increases in ET in projections ( [[#11.6.5.2|Section 11.6.5.2]] ; [[#Berg--2016|Berg et al., 2016]] ; [[#Zhou--2021|Zhou et al., 2021]] ). In general, atmospheric-based indices show more drying than total column soil moisture ( [[#Berg--2018|Berg and Sheffield, 2018]] ; [[#Cook--2020|Cook et al., 2020]] ; [[#Scheff--2021|Scheff et al., 2021]] ), but are more consistent with projected increases in surface soil moisture deficits ( [[#Dirmeyer--2013|Dirmeyer et al., 2013]] ; [[#Dai--2018|Dai et al., 2018]] ; [[#Lu--2019|Lu et al., 2019]] ; [[#Cook--2020|Cook et al., 2020]] ; [[#Vicente-Serrano--2020c|Vicente-Serrano et al., 2020c]] ).

Atmospheric-based drought indices are not metrics of soil moisture or runoff ( [[#11.6.1.5|Section 11.6.1.5]] ) so their projections may not necessarily reflect the same trend of online simulated soil moisture and runoff. Independently of effects on the land water balance, atmospheric-based drought indices will reflect the potential vegetation stress resulting from deficits between available water and enhanced AED, even in conditions with no or low ET. Under dry conditions, the enhanced AED associated with human forcing would increase plant water stress ( [[#Brodribb--2020|Brodribb et al., 2020]] ), with effects on widespread forest dieback and mortality ( [[#Anderegg--2013|Anderegg et al., 2013]] ; [[#Williams--2013|Williams et al., 2013]] ; [[#Allen--2015|Allen et al., 2015]] ; [[#McDowell--2015|McDowell and Allen, 2015]] ; [[#McDowell--2016|McDowell et al., 2016]] , 2020), and stronger risk of megafires ( [[#Flannigan--2016|Flannigan et al., 2016]] ; [[#Podschwit--2018|Podschwit et al., 2018]] ; [[#Clarke--2019|Clarke and Evans, 2019]] ; [[#Varela--2019|Varela et al., 2019]] ). For these reasons, there is ''high confidence'' that the future projections of enhanced drought severity showed by the PDSI-PM and the SPEI-PM are representative of more frequent and severe plant stress episodes and more severe agricultural and ecological drought impacts in some regions.

Global tendencies towards more severe and frequent agricultural and ecological drought conditions are identified in future projections when focusing on atmospheric-based drought indices such as the PDSI-PM or the SPEI-PM. They expand the spatial extent of drought conditions compared to meteorological drought to most of North America, Europe, Africa, Central and East Asia and Southern Australia ( [[#Cook--2014a|Cook et al., 2014a]] ; [[#Chen--2017a|Chen and Sun, 2017a]] , b; [[#Gao--2017b|Gao et al., 2017b]] ; [[#Lehner--2017|Lehner et al., 2017]] ; [[#Zhao--2017|Zhao and Dai, 2017]] ; [[#Dai--2018|Dai et al., 2018]] ; [[#Naumann--2018|Naumann et al., 2018]] ; [[#Potopová--2018|Potopová et al., 2018]] ; [[#Gu--2020|Gu et al., 2020]] ; Vicente-Serrano et al., 2020c; [[#Dai--2021|Dai, 2021]] ). Projections in PDSI-PM and SPEI-PM are used to complement total soil moisture projections in assessing projected changes in agricultural and ecological drought ( [[#11.9|Section 11.9]] ).

<div id="11.6.5.6" class="h3-container"></div>

<span id="synthesis-for-different-drought-types-3"></span>
==== 11.6.5.6 Synthesis for Different Drought Types ====

<div id="h3-29-siblings" class="h3-siblings"></div>

The tables in [[#11.9|Section 11.9]] provide assessed projected changes in metorological drought, agricultural and ecological drought, and hydrological droughts. The assessment shows that several regions will be affected by more severe agricultural and ecological droughts even if global warming is stabilized at 2°C, including MED, WSAF, SAM and SSA ( ''high confidence'' ), and ESAF, MDG, EAU, SAU, SCA, CAR, NSA, NES, SWS, WCE, NCA, WNA and CNA ( ''medium confidence'' ). Some regions are also projected to be affected by more severe agricultural and ecological droughts at 1.5°C (MED, WSAF, ESAF, SAU, NSA, SAM, SSA, can; ''medium confidence'' ) At 4°C of global warming, even more regions would be affected by agricultural and ecological droughts (WCE, MED, CAU, EAU, SAU, WCA, EAS, SCA, CAR, NSA, NES, SAM, SWS, SSA, NCA, CNA, ENA, WNA, WSAF, ESAF and MDG). NEAF, SAS are also projected to experience less agricultural and ecological drought with global warming ( ''medium confidence'' ). Projected changes in meteorological droughts are, overall, less extended but also affect several AR6 regions, at 1.5°C and 2°C (MED, EAU, SAU, SCA, NSA, NCA, WSAF, ESAF, MDG) and 4°C of global warming (WCE, MED, EAU, SAU, SEA, SCA, CAR, NWS, NSA, NES, SAM, SWS, SSA, NCA, ENA, WAF, WSAF, ESAF, MDG). Several regions are also projected to be affected by more hydrological droughts at 1.5°C and 2°C (WCE, MED, WNA, WSAF, ESAF) and 4°C of global warming (NEU, WCE, EEU, MED, SAU, WCA, SCA, NSA, SAM, SWS, SSA, WNA, WSAF, ESAF, MDG). To illustrate the changes in both intensity and frequency of drought in the regions where strongest changes are projected, Figure 11.18 displays changes in the intensity and frequency of soil moisture drought under different global warming levels (1.5°C, 2°C, 4°C) relative to the 1851-1900 baseline based on CMIP6 simulations under different SSP forcing scenarios averaged over “drying regions”, i.e. AR6 regions for which there is at least ''medium confidence'' in increase in agricultural and ecological drought at 2°C of global warming. The 90% uncertainty ranges for the projected changes in both intensity and frequency are above zero, indicating significant increase in both intensity and frequency of drought in these regions as whole.

In summary, more regions are affected by increases in agricultural and ecological droughts with increasing global warming ( ''high confidence'' ). New evidence strengthens the SR1.5 conclusion that even relatively small incremental increases in global warming (+0.5°C) cause a worsening of droughts in some regions ( ''high confidence'' ) ''.'' Some regions are projected to be affected by more severe agricultural and ecological droughts at 1.5°C of global warming (MED, WSAF, ESAF, SAU, NSA, SAM, SSA, can; ''medium confidence'' ). A larger number of regions are projected to be affected by more severe agricultural and ecological droughts at 2°C of global warming, including MED, WSAF, SAM and SSA ( ''high confidence'' ), and ESAF, MDG, EAU, SAU, SCA, CAR, NSA, NES, SWS, WCE, NCA, WNA and CNA ( ''medium confidence'' ). At 4°C of global warming, even more regions would be affected by agricultural and ecological droughts (WCE, MED, CAU, EAU, SAU, WCA, EAS, SCA, CAR, NSA, NES, SAM, SWS, SSA, NCA, CNA, ENA, WNA, WSAF, ESAF and MDG). Some regions are also projected to experience less agricultural and ecological drought with global warming ( ''medium confidence;'' NEAF, SAS). There is ''high confidence'' that the projected increases in agricultural and ecological droughts are strongly affected by AED increases in a warming climate, although ET increases are projected to be smaller than those in AED due to soil moisture limitations and CO <sub>2</sub> effects on leaf stomatal conductance. Enhanced atmospheric CO <sub>2</sub> concentrations lead to enhanced water-use efficiency in plants ( ''medium confidence'' ), but there is ''low confidence'' that it can alleviate agricultural and ecological droughts, or hydrological droughts, at higher global warming levels characterized by limited soil moisture and enhanced AED.

Projected changes in meteorological droughts are overall less extended than for agricultural and ecological droughts, but also affect several AR6 regions, even at 1.5°C and 2°C of global warming. Several regions are also projected to be more strongly affected by hydrological droughts with increasing global warming (NEU, WCE, EEU, MED, SAU, WCA, SCA, NSA, SAM, SWS, SSA, WNA, WSAF, ESAF, MDG). Increased soil moisture limitation and associated changes in droughts are projected to lead to increased vegetation stress in many regions, with implications for the global land carbon sink (Cross-Chapter Box 5). There is ''high confidence'' that the global land carbon sink will become less efficient due to soil moisture limitations and associated drought conditions in some regions in higher-emissions scenarios, especially under global warming levels above 4°C; however, there is ''low confidence'' on how these water cycle feedbacks will play out in lower-emissions scenarios (at 2°C global warming or lower; Cross-Chapter Box5.1).

<div id="11.7" class="h1-container"></div>

<span id="extreme-storms"></span>