FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report 1 2 Cross-Chapter Paper 3: Deserts, Semi-Arid Areas and Desertification 3 4 Cross-Chapter Paper Leads: Alisher Mirzabaev (Uzbekistan), Lindsay C. Stringer (United Kingdom) 5 6 Cross-Chapter Paper Authors: Tor A. Benjaminsen (Norway), Patrick Gonzalez (USA), Rebecca Harris 7 (Australia), Mostafa Jafari (Iran), Nicola Stevens (South Africa), Cristina Maria Tirado (Spain/USA), 8 Sumaya Zakieldeen (Sudan) 9 10 Cross-Chapter Paper Contributing Authors: Elena Abraham (Argentina), Dulce Flores-Renteria 11 (Mexico), Houda Ghazi (Morocco), Pierre Hiernaux (France), Margot Hurlbert (Canada), Oksana Lipka 12 (Russian Federation), Nausheen Mazhar (Pakistan), Nicholas Middleton (United Kingdom), Uriel Safriel 13 (Israel), Ranjay K. Singh (India), Fei Wang (China) 14 15 Cross-Chapter Paper Review Editor: Taha Zatari (Saudi Arabia) 16 17 Cross-Chapter Paper Scientist: Asmita Bhardwaj (India) 18 19 Date of Draft: 1 October 2021 20 21 Note: TSU Compiled Version 22 23 24 Table of Contents 25 26 Executive Summary..........................................................................................................................................2 27 CCP3.1 Introduction...................................................................................................................................5 28 CCP3.1.1 Concepts, Definitions and Scope ............................................................................................5 29 CCP3.1.2 Key Measurement Challenges and Observed Dryland Dynamics..........................................6 30 CCP3.2 Observed Impacts of Climate Change Across Sectors and Regions ........................................7 31 CCP3.2.1 Observed Impacts on Natural Systems in Arid and Semi-arid Areas .....................................7 32 CCP3.2.2 Observed Impacts of Climate Change on Human Systems in Desert and Semi-Arid Areas 11 33 Box CCP3.1: Pastoralism and Climate Change...........................................................................................13 34 CCP3.3 Future Projections ......................................................................................................................15 35 CCP3.3.1 Projected Changes and Risks in Natural Systems ................................................................15 36 CCP3.3.2 Projected Impacts on Human Systems..................................................................................18 37 CCP3.4 Adaptations and Responses........................................................................................................19 38 FAQ CCP3.1: How has climate change already affected drylands and why are they so vulnerable? ...22 39 FAQ CCP3.2: How will climate change impact the world's drylands and their people?........................22 40 FAQ CCP3.3: What can be done to support sustainable development in desert and semi-arid areas, 41 given projected climate changes?..........................................................................................................23 42 Large Tables....................................................................................................................................................24 43 References .......................................................................................................................................................35 44 45 Do Not Cite, Quote or Distribute CCP3-1 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report 1 Executive Summary 2 3 Introduction 4 5 This cross-chapter paper on "Deserts, semi-arid areas and desertification" updates and extends Chapter 3 on 6 "Desertification" in the IPCC Special Report on Climate Change and Land (SRCCL). It assesses new 7 information and links it to the findings across the chapters of Working Group 2's contribution as well as 8 relevant chapters of Working Group 1's contribution to the IPCC Sixth Assessment Report (AR6), with an 9 added focus on deserts which were outside the scope of the SRCCL. 10 11 Where are we now: Observed impacts and adaptation responses 12 13 Deserts and semi-arid areas have already been affected by climate change, with some areas 14 experiencing increases in aridity. Mixed trends of decreases and increases in vegetation productivity 15 have been observed, depending on the time period, geographic region, detection methods used and 16 vegetation type under consideration (high confidence1). These changes have had varying and location- 17 specific impacts on biodiversity, and have altered ecosystem carbon balance, water availability and the 18 provision of ecosystem services (high confidence). There is no evidence, however, of a global trend in 19 dryland expansion based on analyses of vegetation patterns, precipitation and soil moisture, with 20 overall, more greening than drying in drylands since the 1980s (medium confidence). Deserts and semi- 21 arid areas host unique biodiversity, rich cultural heritage and provide globally valuable ecosystem services. 22 They are also highly vulnerable to climate change. The vitality of natural ecosystems in arid and semi-arid 23 regions greatly depends on water availability, as they are highly sensitive to changes in precipitation and 24 potential evapotranspiration {3.1.2; 3.2.1}, as well as to land management practices. Multiple lines of 25 evidence from 1920-2015 indicate that surface warming of 1.2篊-1.3篊 over global drylands (Section 1.1.1) 26 exceeded the 0.8篊-1.0篊 warming over humid lands. From 1982 to 2015, unsustainable land use and 27 climate change combined caused desertification of 6% of the global dryland area, while 41% showed 28 significant increases in vegetation productivity (greening) and 53% of the area had no notable change, 29 although greening rates are slowing or declining in some locations. Greening may cause biodiversity loss and 30 ecosystem service degradation in relation to livelihood systems {3.2.2}. Observed trends in deserts and semi- 31 arid areas have led to varying impacts on flora, fauna, soil, nutrient cycling, the carbon cycle and water 32 resources. Ecological changes in dryland ecosystems detected and attributed primarily to climate change 33 include tree mortality and losses of mesic tree species at specific sites in the African Sahel particularly 34 during the droughts of the 1970s and 80s, and in North Africa from 1970 to 2007 (CCP4.3.2); and losses of 35 bird species in the Mojave Desert of North America from 1908 to 2016 (CCP4.3.2). In contrast, growth in 36 herbaceous vegetation production has increased in some drylands since the 1980s. Widespread woody 37 encroachment has occurred in many shrublands and savannas in Africa, Australia, North America and South 38 America, due to a combination of land use change, changes in rainfall, fire suppression, and CO2 fertilization 39 {3.2.1, 3.2.2} which, together with unsustainable management, alters biodiversity and reduces ecosystem 40 services such as water availability and grazing potential. 41 42 The impacts of climate change have affected the ecosystem services that humans can harness from 43 drylands, with largely negative implications for livelihoods, human health and wellbeing, particularly 44 in deserts and semi-arid areas with lower adaptive capacities (high confidence). Ecosystem degradation 45 (Section 16.5.2.3.1) and desertification threaten the abilities of both natural and human systems to adapt to 46 climate change (high confidence) {3.1.1}. Changes in desert and semi-arid ecosystem services most acutely 47 affect people who are directly dependent on naatural resources for their livelihoods and survival. These 48 groups also often have lower capacities to adapt, particularly given structural limitations of some drylands 49 where healthcare, sanitation, infrastructure and efficient markets are lacking, reinforcing existing inequalities 50 (high confidence) {3.2.1, 3.2.2}. In rural drylands in tropical and Mediterranean areas, human populations 51 are steadily expanding with mixed implications for ecosystem services under climate change, while rapid 1 In this Report, the following summary terms are used to describe the available evidence: limited, medium, or robust; and for the degree of agreement: low, medium, or high. A level of confidence is expressed using five qualifiers: very low, low, medium, high, and very high, and typeset in italics, e.g., medium confidence. For a given evidence and agreement statement, different confidence levels can be assigned, but increasing levels of evidence and degrees of agreement are correlated with increasing confidence. Do Not Cite, Quote or Distribute CCP3-2 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report 1 urbanisation in new and existing dryland megacities put additional pressure on water ecosystem services 2 (high confidence). Impacts resulting from consumption of dryland ecosystem services elsewhere, alongside 3 other teleconnections associated with health, trade, conflict and migration, mean that dryland adaptive 4 capacities have far reaching implications for other locations, while other locations affect dryland adaptation 5 options {3.1.1; 3.4}. 6 7 Where are we going? Risks and adaptation under warming pathways 8 9 Some drylands will expand by 2100, while others will shrink (high confidence). Climate change affects 10 drylands through increased temperatures and more irregular rainfall, with important differences 11 between areas with different rainfall distributions linked to the dominant climate systems in each 12 location (Chapter 9). Projections are nevertheless uncertain and not well supported by observed 13 trends, while different methodological approaches and indices exhibit different strengths and 14 weaknesses (medium confidence). A fundamental methodological challenge is how to attribute projected 15 impacts to climate change when background climate variability in drylands is so high. Some projections 16 show aridity (as measured by the aridity index) to expand substantially on all continents except Antarctica. 17 Expansion of arid regions is probable in southwest North America, the northern fringe of Africa, southern 18 Africa and Australia. The main areas of semi-arid expansion are likely2 to occur on the north side of the 19 Mediterranean, southern Africa and North and South America. India, parts of northern China, eastern 20 equatorial Africa and the southern Saharan regions are projected to have shrinking drylands. Under RCP 8.5, 21 aridity zones could expand by one-quarter of the 1990 area by 2100, increasing to over half the global 22 terrestrial area. Lower greenhouse gas emissions, under RCP 4.5, could limit expansion to one-tenth of the 23 1990 area by 2100. Nevertheless, the utility of the aridity index in delineating dryland biomes is limited 24 under an increasing CO2 environment (medium confidence) and how well the index fits observed trends has 25 been questioned in recent research. The impacts of climate change on sand and dust storm activity are 26 projected to be substantial, however, there is large regional variability in terms of rainfall seasonality, land 27 management practices, as well as differences in rates of change and the scales at which the projections are 28 undertaken. The characteristics and speed of human responses and adaptations also affect future risks and 29 impacts (high confidence). Increased temperature and rainfall variability will significantly change the inter- 30 annual variability in the global carbon cycle which is strongly influenced by the world's drylands and the 31 ways they are managed (medium confidence). Increased variability of precipitation would generally 32 contribute to increased vulnerability for people in drylands, intensifying the challenges that people living in 33 deserts and semi-arid areas will face for their sustainable development (medium confidence) {3.3.1, 3.3.2}. 34 35 Contributions of adaptation measures to climate resilient development 36 37 Drivers of desert expansion and greening are numerous, are attributed to environmental and human 38 processes and differ across dryland types, yet a suite of adaptations can help to address human drivers 39 of change, support resilience and build the adaptive capacity of dryland people (medium confidence). 40 Deserts and semi-arid areas have a rich cultural heritage, Indigenous knowledge, and local knowledge which 41 enrich and influence sustainability and land use globally. Growing research evidence and experience 42 highlight the necessary features of an enabling environment for dryland adaptation (Section 8.5.2). Key 43 enablers include supportive policies, institutions and governance approaches that strengthen the adaptive 44 capacities of dryland farmers, pastoralists and other dryland resource users (high confidence), addressing 45 drivers (proximate and underlying) as well as symptoms of desertification. For instance, the skills and 46 capacities held by the mobile and adaptive approach of pastoralists may provide lessons for society at large 47 in adapting to climate change and dealing with increased uncertainty. Such a policy would stand in contrast 48 to previous attempts at settling pastoralists. There is a persistent gap in terms of scaling-up already known 49 good practices, combining nature-based, land-based, and ecosystem-based approaches that facilitate 50 sustainable land management, with contextually appropriate and responsible governance systems (e.g., 51 including those supporting communal land tenure arrangements and Indigenous knowledge) (medium 2 In this Report, the following terms have been used to indicate the assessed likelihood of an outcome or a result: Virtually certain 99�100% probability, Very likely 90�100%, Likely 66�100%, About as likely as not 33�66%, Unlikely 0�33%, Very unlikely 0�10%, and Exceptionally unlikely 0�1%. Additional terms (Extremely likely: 95� 100%, More likely than not >50�100%, and Extremely unlikely 0�5%) may also be used when appropriate. Assessed likelihood is typeset in italics, e.g., very likely). This Report also uses the term `likely range' to indicate that the assessed likelihood of an outcome lies within the 17-83% probability range Do Not Cite, Quote or Distribute CCP3-3 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report 1 confidence). Land based adaptations can help manage dryland changes including sand and dust storms and 2 desertification (high confidence), while technological options linked to water management draw from both 3 traditional practices and new innovations. Adequate financing and investment is required to harness multiple 4 benefits for managing the impacts of climate change and desertification whilst accelerating progress towards 5 sustainable development in deserts and semi-arid areas {3.4}. 6 7 Do Not Cite, Quote or Distribute CCP3-4 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report 1 CCP3.1 Introduction 2 3 CCP3.1.1 Concepts, Definitions and Scope 4 5 Deserts and semi-arid areas are in `drylands', which comprise hyper-arid, arid, semi-arid and dry sub-humid 6 areas (Figure CCP3.1). Drylands cover about 45-47% of the global land area (Prvlie, 2016; Koutroulis, 7 2019) and are home to about 3 billion people residing primarily in semi-arid and dry sub-humid areas (van 8 der Esch et al., 2017). Drylands host unique, rich biodiversity (Maestre et al., 2015) and provide important 9 ecosystem services (Bidak et al., 2015; Lu et al., 2018), while dryland people have a rich cultural and 10 historical heritage. Rural human populations are growing in some Mediterranean and tropical drylands, while 11 many are rapidly urbanizing (Guengant Jean-Pierre, 2003; Tabutin and Schoumaker, 2004; Denis and 12 Moriconi-Ebrard, 2009), with varying impacts on ecosystem services and adaptive capacities. In recent 13 decades, 6% of global megacities have been established in arid areas and 2% in hyper-arid desert areas 14 (Cherlet et al., 2018), with many of these areas suffering from severe water security challenges (Stringer et 15 al., 2021). Dryland inhabitants in many developing countries are also experiencing poverty (Section 16 16.1.4.3), hunger, poor health, land degradation, and economic and political marginalisation (Mbow et al., 17 2019; Mirzabaev et al., 2019), which sometimes limits their access to common pool resources. These 18 challenges, together with a weak enabling environment, threaten opportunities to adapt to climate change. 19 20 The terms "desert" and "desertification" are subject to various interpretations due to the diverse components, 21 processes and states they denote. Recognizing "land degradation" as a contested and perceptual term (Blaikie 22 and Brookfield, 1987; Behnke and Mortimore, 2016; Robbins, 2020), this cross-chapter paper, defines land 23 degradation as "a negative trend in land condition, caused by direct or indirect human-induced processes 24 including climate change, expressed as long-term reduction or loss of at least one of the following: biological 25 productivity, ecological integrity or value to humans" (Olsson et al., 2019). Desertification is land 26 degradation in arid, semi-arid, and dry sub-humid areas (UNCCD, 1994). Following the above definitions, 27 desertification is more common in arid and semi-arid climates than in hyper-arid climates. When 28 desertification does occur in arid and hyper-arid ecosystems it is often in oases and irrigated cultivated lands 29 (Ezcurra, 2006; Dilshat et al., 2015). Hyper-arid areas, except wetlands such as oases, wadis and riverbanks, 30 are excluded in the United Nations Convention to Combat Desertification (UNCCD) definition of 31 desertification used here, yet many of the world's deserts are in hyper-arid areas. Hyper-arid areas are 32 therefore included when discussing deserts but not when discussing desertification. Deserts are not the end 33 point in a desertification process (Ezcurra, 2006), and there is robust evidence of desertification in deserts, 34 mostly driven by human activities and climate variability, expressed as loss of biological productivity, 35 ecological integrity or value to humans to below their natural levels (Moridnejad et al., 2015). 36 37 Interactions between climate change and desertification in drylands create challenges for both ecosystem and 38 human resilience, affecting ecosystem services, biodiversity, food security, human health and wellbeing 39 (Reed and Stringer, 2016). Dryland livelihoods that heavily rely on natural ecosystems face pressures 40 including high population growth rates, weak or poor governance, low investment, unemployment and 41 poverty, market distortions and underestimates of the value of drylands (Stringer et al., 2017; Bawden, 42 2018). These pressures intersect with broader societal challenges such as conflict and civil unrest (Okpara et 43 al., 2015; Almer et al., 2017), which together, can contribute to human displacement (Section 16.2.3.8) in 44 some drylands (Warner, 2010; Abel et al., 2019). Nevertheless, evidence linking conflict with climate 45 change and desertification is weak (Benjaminsen et al., 2012) and data are insufficient to draw robust 46 conclusions. 47 48 Drylands yield important opportunities for adapting to and mitigating climate change. They offer abundant 49 solar energy which could support mitigation efforts, opportunities for cultural and nature-based tourism, rich 50 plant biodiversity in some areas (e.g. Namibia), and extensive Indigenous knowledge and experience of 51 adapting to dynamic climates (Christie et al., 2014; Stringer et al., 2017), e.g. across West Asia and North 52 Africa (Louhaichi and Tastad, 2010; Hussein, 2011). Improved understanding of challenges and 53 opportunities in drylands can be achieved by transdisciplinary, multi-scale and inter-sectoral approaches 54 encompassing links between physical, biological and socioeconomic, and institutional systems (Reynolds et 55 al., 2007; Stringer et al., 2017). 56 Do Not Cite, Quote or Distribute CCP3-5 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report 1 Chapter 3 of the IPCC Special Report on Climate Change and Land (SRCCL) focused on desertification, but 2 links between climate change and deserts, desertification and semi-arid areas have not been extensively 3 considered in recent IPCC assessment cycles. AR5 noted that desertification contributes to atmospheric dust 4 production, identifying desertification as needing consideration within climate change mitigation and 5 adaptation governance and decision-making (Boucher et al., 2013; Myhre et al., 2013). This cross-chapter 6 paper focuses on environmental and human aspects, finding that climate change impacts will intensify the 7 challenges faced by dryland populations in advancing sustainable development. However, viable options 8 exist for adapting to climate change, reducing desertification and supporting progress towards the 9 Sustainable Development Goals (SDGs), particularly by combining modern science, Indigenous knowledge, 10 and local knowledge, as well as livelihood and land management strategies that enable land-based 11 adaptation, mitigation and nature-based solutions (Section 16.3.2.3). 12 13 CCP3.1.2 Key Measurement Challenges and Observed Dryland Dynamics 14 15 Maps of dryland extent commonly employ a climate-based approach measured using the aridity index (AI), 16 or consider the extent of dryland vegetation. The two approaches sometimes do not demarcate the same 17 geographical areas as being drylands, particularly when projecting future changes (Stringer et al., 2021). 18 Dryland dynamics therefore need to be viewed specifically in relation to the definitions and measurements 19 being used. From 1982 to 2015, unsustainable land use and climate change combined caused desertification 20 of 6% of the global dryland area, while 41% showed significant greening (i.e. increased vegetation 21 productivity), and 53% of the area had no notable change (Figure CCP3.1; Burrell et al., 2020);). In contrast 22 Yuan et al. (2019) conclude that during 1999-2015, trends of vegetation production reversed globally and in 23 drylands, showing extensive declines. Thus, while overall greening has occurred, this trend now appears to 24 be declining. Analyses of vegetation, soil, and physical characteristics of over 50 000 sample points in 25 drylands around the world indicate that aridification causes ecological degradation at three successive 26 thresholds: vegetation decline at aridity index = 0.56, soil disruption at aridity index = 0.3, and loss of plant 27 cover at aridity index = 0.2 (Berdugo et al., 2020). Drylands nevertheless show different dynamics 28 depending on the index used and the variables assessed. 29 Based on the AI, some drylands are projected to expand and others to contract due to climate change. 30 However, there is no evidence of a global trend in dryland expansion based on vegetation patterns, 31 precipitation and soil moisture, based on the satellite record from the 1980s to the present (medium 32 confidence). The AI will also be of limited use under a changing CO2 environment due to higher water use 33 efficiency by some plants (Mirzabaev et al., 2019), and it overvalues the role of potential evapotranspiration 34 (PET) relative to rainfall. It also does not account for CO2 impacts on evapotranspiration, and seasonality in 35 rainfall and evapotranspiration. Higher annual PET because of increased temperatures will have little impact 36 if temperature and actual evapotranspiration are not rising during the period of vegetation growth (Stringer et 37 al., 2021). 38 39 Do Not Cite, Quote or Distribute CCP3-6 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report 1 2 Figure CCP3.1: Aridity zone extent and observed changes in dryland areas as defined by the Aridity Index. Aridity 3 zones, according to UNESCO (1979) and UNEP (1992) classifications, defined by the aridity index (AI), consider the 4 ratio of average annual precipitation to potential evapotranspiration: (i) dry sub-humid (0.5AI<0.65), (ii) semi-arid 5 (0.2AI<0.5), (iii) arid (0.05AI<0.2) and (iv) hyper-arid (AI<0.05). Drylands include land with AI <0.65, humid 6 lands are those with AI >0.6. Drylands include land with AI <0.65, humid lands are those with AI >0.65 (UNEP, 7 1992). Deserts represent a major part of the hyper-arid and arid zones. The aridity zones are shown for climate in the 8 period 1988-2017 and changes in dryland area (combined area of four aridity zones) are shown between the periods 9 1901-1930 and 1988-2017, based on climate time series at 50 km spatial resolution (Harris et al., 2020). The AI has 10 various limitations in assessing dryland expansion and different indices highlight different areas and different changes. 11 This is known as the aridity paradox (Greve et al., 2019). See SRCCL Section 3.2.1 (Mirzabaev et al., 2019) for an in- 12 depth analysis of limitations, and Stringer et al. (2021) for a summary of different measures and indices used in the 13 literature. 14 15 16 CCP3.2 Observed Impacts of Climate Change Across Sectors and Regions 17 18 CCP3.2.1 Observed Impacts on Natural Systems in Arid and Semi-arid Areas 19 20 CCP3.2.1.1 Temperature and Rainfall 21 22 Significant warming has occurred across drylands globally (WGI, 2021). Surface warming (from 1920-2015) 23 of 1.2篊-1.3篊 in global drylands has exceeded the 0.8篊-1.0篊 warming over humid lands (Huang et al., 24 2017). As measured by the AI, this has expanded the area of drylands by ~4% from 1948-2004 (Ji et al., 25 2015; Spinoni et al., 2015; Huang et al., 2016). However, as mentioned in Figure CCP3.1, the AI has various 26 limitations in assessing drylands expansion. Increases in potential evapotranspiration have exceeded 27 increases in precipitation in the last half of the period 1901-2017 (Pan et al., 2021). Observations from the 28 Sahel demonstrated that temperature seasonality changes differ from rainfall seasonality changes (Guichard 29 et al., 2015), and there has been an increase in surface water and groundwater recharge in the Sahel since the 30 1980s, referred to as "the Sahel paradox" (Favreau et al., 2009; Gardelle et al., 2010; Descroix et al., 2013; 31 Wendling et al., 2019). Research from the USA suggests that historical soil moisture levels can contribute to 32 such variability (Heisler-White et al., 2009). Studies from the Middle East show rising temperatures and 33 declining rainfall trends (ESCWA, 2017), with most decreasing aridity trends in north Sudan and most 34 increasing aridity trends in eastern Arabia over the period 1948-2018 (Sahour et al., 2020). 35 Do Not Cite, Quote or Distribute CCP3-7 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report 1 CCP3.2.1.2 Ecosystem Processes 2 3 Semi-arid ecosystems have a disproportionately large role in the global carbon cycle, driving trends and 4 inter-annual variability of the global carbon sink (Alstrom et al., 2015). These systems are highly sensitive to 5 annual precipitation and temperature variations (high confidence) (Alstrom et al., 2015, Poulter et al., 2014). 6 The positive trend in semi-arid regions is consistent with widespread woody encroachment and increased 7 vegetation greenness (Andela et al., 2013; Piao et al., 2019; Piao et al., 2020) driven by CO2 fertilisation and 8 rainfall increases (Sitch et al., 2015; Piao et al., 2020), although some trends are complicated by irrigation 9 practices (He et al., 2019). Increases in temperature and drought diminish this trend through reduced 10 vegetation productivity and increased vegetation mortality (Brandt et al., 2016; Ma et al., 2016; Fern醤dez- 11 Mart韓ez et al., 2019; Maurer et al., 2020) with indications that this trend is declining or reversing in some 12 locations (Yuan et al., 2019; Wang et al., 2020). 13 14 Changed climates have increased water constraints of vegetation growth most notably in the Mediterranean 15 (CCP1.2.3.2, CCP4.2.1) and west and central Asia (Jiao et al., 2021). Climate change and elevated CO2 have 16 both increased and decreased vegetation sensitivity to rainfall throughout drylands, with the degree of 17 variation shaped by region, land-use and vegetation traits (Haverd et al., 2017; Abel et al., 2021). Mineral 18 nitrogen production in drylands may become increasingly decoupled from consumption by plants over 19 prolonged dry periods, and more extreme hydrological events can drive multiple changes to nutrient cycling 20 (Manzoni et al., 2019). Soil biocrusts (composed of lichens, bryophytes and soil microorganisms) which 21 contribute to dryland ecosystem function including carbon uptake and soil stabilisation (Reed et al., 2019), 22 are sensitive to warming and altered rainfall in a shift in biocrust communities of mosses and lichens in 23 favour of early successional cyanobacteria-dominated biocrusts (Escolar et al., 2012; Reed et al., 2012), 24 which can increase surface albedo (Rutherford et al., 2017). 25 26 CCP3.2.1.3 Vegetation Changes 27 28 CCP3.2.1.4 Woody Cover Increase 29 30 Dryland ecosystems have shown mixed trends of decreases and increases in vegetation and biodiversity, 31 depending on the time period, geographic region, and vegetation type assessed (see Table CCP3.1 for 32 examples of observed environmental changes and impacts in drylands and the role of climate change and 33 non-climatic factors in causing these changes). 34 35 Increases in shrub cover in arid deserts and shrublands have been recorded in the North American drylands 36 (Caracciolo et al., 2016; Archer et al., 2017; Chambers et al., 2019), the Namib desert (Rohde et al., 2019), 37 the Karoo (Ward et al., 2014; Masubelele et al., 2015b), north and central Mexico (P閞ez-S醤chez et al., 38 2011; B醗z et al., 2013; Castill髇 et al., 2015; Sosa et al., 2019), large parts of the West African Sahel with 39 some local exceptions (Brandt et al., 2016), and in Central Asia (Jia et al., 2015; Li et al., 2015; Deng et al., 40 2016; Jiao et al., 2016; Wang et al., 2016). Increasing woodiness in the Namib is consistent with an increase 41 in rainfall extremes and westward expansion of convective rainfall (Haensler et al., 2010; Rohde et al., 42 2019). Increasing rainfall and rising CO2 concentrations (which improves water use efficiency) benefits 43 some shrubs (Polley et al., 1997; Morgan et al., 2004; Donohue et al., 2013). Together with changes in land 44 use (Hoffman et al., 2018), improved land management (Reij et al., 2005) and improved irrigation (He et al., 45 2019) this contributes to woody cover increases. Extensive woody encroachment has been recorded in 46 savannas (measured between 1920-2015, over the past century) in Africa (2.4% woody cover 47 increase/decade), Australia (1% increase/decade), and South America (8% increase/decade) (O'Connor et al., 48 2014; Stevens et al., 2016; Skowno et al., 2017; Venter et al., 2018; Zhang et al., 2019). Following drought 49 in the Sahel (1968-1973 and 1982-1984), a rainfall increase since the mid-1990s has been linked to increases 50 of woody cover between 1992-2011/2012 (Brandt et al., 2016; Brandt et al., 2017; Brandt et al., 2019). See 51 SRCCL section 3.2.1.1 for an evaluation of NDVI and remote sensing approaches used in these studies. Tree 52 regeneration by farmers has also increased woody cover, particularly next to villages (high confidence) (Reij 53 et al., 2005; Reij and Garrity, 2016; Brandt et al., 2018). Otherwise, savanna encroachment has been 54 attributed to combinations of increased rainfall (Venter et al., 2018; Zhang et al., 2019), warming (Venter et 55 al., 2018) and CO2 fertilisation (Kgope et al., 2010; Bond and Midgley, 2012; Buitenwerf et al., 2012; 56 Stevens et al., 2016; Quirk et al., 2019) interacting with changing land use (Archer et al., 2017; Venter et al., 57 2018) where herbivory and fire regimes are altered (O'Connor et al., 2014; Archer et al., 2017;see also Do Not Cite, Quote or Distribute CCP3-8 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report 1 discussion on fire and herbivory in Section 2.4.3.1.2). In some cases, woody increase has been balanced 2 locally by changes in runoff (Trichon et al., 2018), or by land clearing and fuel wood harvesting, as seen in 3 western Niger, northern Nigeria, and at the periphery of major towns (Montagn� et al., 2016). 4 5 CCP3.2.1.5 Tree Death and Woody Cover Decline 6 7 Field measurements have also detected tree mortality and loss of mesic tree species at some Sahel sites 8 during drought periods (Gonzalez et al., 2012; Kusserow, 2017; Brandt et al., 2018; Ibrahim et al., 2018; 9 Trichon et al., 2018; Zwarts et al., 2018; Bernardino et al., 2020; Zida et al., 2020) and a reduction of mesic 10 species in favour of drought-tolerant species (high confidence) (H鋘ke et al., 2016; Kusserow, 2017; Ibrahim 11 et al., 2018; Trichon et al., 2018; Dendoncker et al., 2020; Zida et al., 2020b), with attribution to climate 12 change (Gonzalez et al., 2012). Furthermore, vegetation productivity per unit of rainfall showed a net decline 13 of 4% in the period 2000-2015 across drylands globally, with the greatest net declines in Africa (16%) and 14 Asia (33%) (Abel et al., 2021), but with location-specific increases in vegetation-rainfall sensitivity, e.g. in 15 southern and eastern Africa and parts of the Sahel. Furthermore, NDVI declines were reported across the 16 Sahel from 1999 to 2015 (Yuan et al., 2019; Zida et al., 2020a). However, field site monitoring showed a 17 strong regeneration of the decimated woody populations except on shallow soil where the runoff system had 18 evolved towards a web of gullies (Hiernaux et al., 2009a; Trichon et al., 2018; Wendling et al., 2019) . 19 20 Other site-specific impacts include tree mortality in south-western Morocco (Le Polain de Waroux and 21 Lambin, 2012), mortality of Austrocedrus and Nothofagus forests in the dry Patagonia forest-steppe 22 (Rodr韌uez-Cat髇 et al., 2019), and a tree range contraction of Aloidendron dichotmum in Southern Africa 23 (Foden et al., 2007b). In Morocco, tree mortality was most highly correlated to an increase in aridity, 24 measured by the Palmer Drought Severity Index (PDSI), which showed a statistically significant increase 25 since 1900 due to climate change (Dai et al., 2004; Esper et al., 2007; Dai, 2011). 26 27 In deserts of the south-western United States, a drought since 2000, mainly due to climate change (Williams 28 et al., 2020), together with land use change, invasive plant species, and wildfire (Syphard et al., 2017), has 29 led to reductions in native desert plant species (Defalco et al., 2010; Conver et al., 2017) and perennial 30 vegetation cover (Munson et al., 2016a; Munson et al., 2016b). An increase in invasive exotic grasses has 31 increased wildfires in these desert ecosystems in which fire had been rare (Brooks and Matchett, 2006; 32 Abatzoglou and Kolden, 2011; Hegeman et al., 2014; Horn and St. Clair, 2017). In the Mojave Desert in the 33 United States, a loss of bird biodiversity has also been detected and attributed to increased aridity caused by 34 climate change (Iknayan and Beissinger, 2018; Riddell et al., 2019). 35 36 CCP3.2.1.6 Change in Herbaceous Cover 37 38 Changes in aridity (Rudgers et al., 2018) have caused some expansion of dominant grasses (often invasive) 39 into desert shrublands. The spread of invasive Bromus tectorum may be enhanced by altered precipitation 40 and freeze-thaw cycles (low confidence) (Collins and Xia, 2015; Rudgers et al., 2018). Arid grassland has 41 expanded (between 10-100 km) into the eastern Karoo, South Africa (high confidence) (du Toit et al., 2015; 42 Masubelele et al., 2015a; Masubelele et al., 2015b). Observations from 100-year-old grazing trials 43 demonstrate that the increase in grassiness is a product of shift in rainfall seasonality and an increase in 44 rainfall (Du Toit and O'Connor, 2014; du Toit et al., 2015; Masubelele et al., 2015a; Masubelele et al., 45 2015b; du Toit et al., 2018). These changes are causing an increase in fire frequency in these seldom burnt 46 areas (du Toit et al., 2015). The Sahara Desert was suggested to have expanded 10% from 1902 to 2013 47 (Thomas and Nigam, 2018), although herbaceous vegetation production has increased in general in the Sahel 48 since the dry 1980s (Eklundh and Olsson, 2003; Anyamba and Tucker, 2005; Herrmann et al., 2005; 49 Hutchinson et al., 2005; Olsson et al., 2005; Fensholt et al., 2006; Dardel et al., 2014; Hiernaux et al., 2016; 50 Stith et al., 2016; Benjaminsen and Hiernaux, 2019; Hiernaux and Assouma, 2020). 51 52 Trends of land degradation (Section 16.4.1.2) and desertification (as demonstrated by loss of cover or 53 reduced vegetation productivity) as an impact of changing climatic trends have been reported in Burkina 54 Faso (Zida et al., 2020), the north-western regions of China during 1975-1990 (Zhang et al., 2020) in 55 Afghanistan (Savage et al., 2009), Iran (Mahmoudi et al., 2011; Kamali et al., 2017), Argentina (Barbosa et 56 al., 2015) and India (Javed et al., 2012). Encroachment, re-greening and an increase of unpalatable plant 57 species into rangeland areas (e.g. in East Africa and southern Africa's Kalahari) all contribute to dryland Do Not Cite, Quote or Distribute CCP3-9 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report 1 degradation through the loss of open ecosystems and their services (Reed et al., 2015; Le et al., 2016; Chen 2 et al., 2019b). 3 4 5 Table CCP3.1: Observed ecological changes in drylands. 6 [INSERT TABLE CCP3.1 HERE] 7 8 9 CCP3.2.1.7 Sand and Dust Storms 10 11 Soil dust emissions are highly sensitive to changing climate conditions but also to changing land use and 12 management practices (high confidence). Distinguishing between the effects of these drivers is not 13 straightforward, even in well-documented locations (Middleton, 2019). There is limited evidence and low 14 agreement about the impacts of climate change on sand and dust storms (SDS), with studies pointing to 15 either substantial increases (+300%) or decreases (-60%) (Boucher et al., 2013). Current climate models 16 cannot adequately model the impact of climate change on SDS activity (Mirzabaev et al., 2019). However, 17 there is high confidence that land degradation, loss of vegetative cover, and drying of water bodies in semi- 18 arid and arid areas will contribute to sand and dust activity (Mirzabaev et al., 2019). 19 20 Sand and dust storms remain a major concern for desert areas under conditions of climate change and 21 desertification (Middleton, 2017). Only about 20% of deserts are covered by sand, but desert sand and dust 22 storms provide an important feedback mechanism to climate (Pu and Ginoux, 2017), with literature showing 23 that some areas have very frequent dust days (Figure CCP3.2; Ginoux et al., 2012). In some locations such as 24 the USA, desert dust can be deposited downwind on snowpacks, hastening snowmelt and altering river 25 hydrology (Painter et al., 2010). Deserts and other natural dryland surfaces produced 75-90% of atmospheric 26 dust globally in the early 21st century, with the remainder from agricultural and other land dominated by 27 human land use (Ginoux et al., 2012; Stanelle et al., 2014). 28 29 Recent changes in dust emissions and their attributions vary geographically. Warming in Iran over the period 30 1951�2013 has been associated with an increased frequency of dust events (Alizadeh-Choobari and Najafi, 31 2018) and a trend (2000�2014) towards increased fine atmospheric mineral dust concentrations in the US 32 southwest has been linked to increasing aridity (Hand et al., 2017). Conversely, increases in rainfall, soil 33 moisture, and vegetation linked to changes in circulation strength of the Indian summer monsoon since 2002 34 have led to a substantial reduction of dust in the Thar Desert and surrounding region, showing agreement 35 with findings from the Sahel and the West African Monsoon (Kergoat et al., 2017). A decreasing trend in the 36 number and intensity of SDS in spring (2007-2016) in East Asia has also responded to higher precipitation 37 and soil moisture, related to a decrease in the intensity of the polar vortex, favouring higher vegetation cover 38 during the period studied (An et al., 2018). Global climate change, transboundary movement of aeolian 39 material by atmospheric flows from Central Asia, dynamics of the Caspian Sea regime, erosion, salinization, 40 as well as the loss of land as a result of the placement of industrial facilities have expanded the land area 41 prone to desertification in Russia. Desertification has been observed to some extent in 27 sub-regions of the 42 Russian Federation on territory of more than 100 million hectares (Kust et al., 2011; also recently confirmed 43 by National Report, 2019). Eastern and south-eastern regions of Kalmykia, Russia, serve as dust sources, 44 while dust and sand masses from the areas of the Black Land sometimes move far beyond to parts of Rostov, 45 Astrakhan, Volgograd, and Stavropol regions. Agricultural land in these areas can become covered with dust 46 and sand 10 cm or more thick, with negative impacts on yields (Tsymbarovich et al., 2020). High dust day 47 frequency is occurring also in the High Latitude Dust (HLD) source areas not reported in Figure CCP3.2 48 such as in Iceland, Patagonia, Canada, Alaska, and based on in situ measurements in Antarctica (Dagsson- 49 Waldhauserov� et al., 2014; Bullard et al., 2016; Dagsson-Waldhauserova and Meinander, 2019; Bachelder 50 et al., 2020). Active HLD dust sources cover at least 500,000 km2 and produce at least 5% of global dust 51 budget (Bullard et al., 2016). HLD has negative impacts on the cryosphere via albedo changes and snow/ice 52 melting (Boy, 2019; Dagsson-Waldhauserova and Meinander, 2019). 53 Do Not Cite, Quote or Distribute CCP3-10 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report 1 2 Figure CCP3.2: Frequency of high dust days (dust optical depth >0.2) during the dust season, based on 2003-2009 3 remote sensing, the most recent data analysed, and divided into areas primarily in agriculture and areas dominated by 4 natural land cover (Ginoux et al., 2012). Dust seasons: Africa (North), Year-round; Africa (South), September- 5 February; America (North), March-May; America (South), December-February), Asia, March-May; Australia, 6 September-February. 7 8 9 CCP3.2.1.8 Water Scarcity 10 11 Climate change and desertification have been linked to water loss (Bayram and 謟t黵k, 2014; Schwilch et 12 al., 2014; Mohamed et al., 2016), decreases in water quantity for irrigation, and contamination of surface 13 water bodies (Middleton, 2017). Increased runoff in areas in the Sahel with shallow soils increased water 14 flows to lakes and the recharge of water tables (Favreau et al., 2009; Gardelle et al., 2010; Descroix et al., 15 2013; Kaptu� et al., 2015; Gal et al., 2017). Water scarcity (Section 16.5.2.3.7) was among the first impacts 16 of climate change recognized in North African countries such as Morocco which have extensive dryland 17 areas, with countries such as Turkey, Libya, USA and China carrying out large-scale water transfer projects 18 (Sternberg, 2016; Stringer et al., 2021). The decrease in water availability in Morocco was substantial in 19 terms of both surface water supply (Rochdane et al., 2012; Choukri et al., 2020) and groundwater (Bahir et 20 al., 2020), threatening agricultural production. 21 22 CCP3.2.2 Observed Impacts of Climate Change on Human Systems in Desert and Semi-Arid Areas 23 24 Climate change and desertification, alongside other drivers of degradation, reduce dryland ecosystem 25 services, leading to losses of biodiversity, water, food, and impacts on human health (CCP4.2.3, WG2) and 26 well-being (high confidence) (Mirzabaev et al., 2019) resulting in disruption to the economic structures and 27 cultural practices of affected communities (Elhadary, 2014; Middleton, 2017). 28 29 CCP3.2.2.1 Sand and Dust Storms 30 31 Desertification and SDS can cause substantial socioeconomic damage in drylands (UNEP, 1992; Opp et al., 32 2021) over both the short and long term. Short-term impacts occur on health, food production systems, 33 infrastructure (damaging buildings, energy systems, and communications), transport and related economic 34 productivity, air and road traffic, and costs are incurred in clearing sand and dust from deposition areas 35 (Mirzabaev et al., 2019). In the Arab region increasing frequency of SDS events is projected to further 36 exacerbate water scarcity and drought (ESCWA, 2017). Longer-term costs include loss of ecosystem Do Not Cite, Quote or Distribute CCP3-11 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report 1 services, biodiversity and habitat, chronic health problems, soil erosion and reduced soil quality (particularly 2 through nutrient losses and deposition of pollutants), and disruption of global climate regulation (Middleton, 3 2018; Allahbakhshi et al., 2019). Dust deposition nevertheless can offer environmental and economic 4 benefits, bringing important nutrients that improve and sustain soil fertility (Marticorena et al., 2017). 5 Preventing and reducing SDS entails upfront investment costs but full cost-benefit analyses of different 6 measures compared to the costs of inaction are scarce and need to consider the likely frequency and 7 magnitude of SDS events (Tozer and Leys, 2013). 8 9 CCP3.2.2.2 Human Health 10 11 Potential impacts of climate change, recurrent droughts and desertification on human health in drylands 12 include: higher risks from water scarcity (linked to deteriorating surface and ground water quality and water- 13 borne diseases; Stringer et al., 2021), food insecurity and malnutrition (Section 16.2.3.3) in the absence of 14 sufficient imports; respiratory, cardiovascular and infectious diseases caused by SDS (Mirzabaev et al., 15 2019), potential displacement and migration and mental health consequences (Chapter 7; Stringer et al., 16 2021) and heat stress (Dunne et al., 2013; Zhao et al., 2015; Russo et al., 2016). SDS negatively impact 17 human health through various pathways, causing respiratory, cardiovascular diseases and facilitating 18 infections (high confidence) (D韆z et al., 2017; Goudarzi et al., 2017; Allahbakhshi et al., 2019; M黱zel et 19 al., 2019). SDS can cause mortality and injuries related to transport accidents (Goudie, 2014). Research from 20 China suggests that prenatal exposure to SDS can affect children's cognitive function (Li et al., 2018). The 21 pollutants that are entrained and ingested or inhaled closely link to the land management strategies in source 22 areas. 23 24 Droughts (TS.2.6 L, AR6, WGI) are among the natural hazards with the highest adverse impacts on human 25 populations (Mishra and Singh, 2010). Although droughts just represented 4% of hazard events, their 26 impacts amounted to 31% of affected people (29 million) (Louvain, 2019). Drought exposure relates to a 27 higher risk of undernutrition (Section 16.5.2.3.6), among vulnerable populations (Kumar, 2016), particularly 28 children (IFPRI, 2016) for whom the impacts can lead to lifelong consequences through stunted growth, 29 impaired cognitive ability and reduced future educational and work performance (UNICEF/WHO/WBG, 30 2019). The corresponding costs of children stunting in terms of lost economic growth can be of the order of 31 7% of per capita income in developing countries (Galasso and Wagstaff, 2018). 32 33 CCP3.2.2.3 Agro-ecological Food Systems, Livelihoods and Food Security 34 35 Rising temperatures, variation in rainfall patterns and frequent extreme weather events associated with 36 climate change have adversely affected agro-ecological food systems and pastoral systems in some drylands 37 (Section 16.3.2.6Zhu et al., 2013; Amin et al., 2018), especially in developing countries (Haider and Adnan, 38 2014; Ahmed et al., 2016; ur Rahman et al., 2018) where desertification is a key challenge to agricultural 39 livelihoods. Recurrent droughts in recent decades, coupled with wind erosion (particularly of fine sediment 40 which gives soil its water holding capacity and nutrients), affected vast areas in Argentina, leading to land 41 abandonment and agricultural fields being covered by sand and invasive plants (Abraham et al., 2016). 42 Temperature increases have contributed to reduced wheat yields in arid, semi-arid and dry sub-humid zones 43 of Pakistan (Sultana et al., 2019). Agricultural production in the drylands of South Punjab is experiencing 44 irreversible impacts since the grain formation phase has become swifter with a warmer climate, leading to 45 improper growth and reduced yields (Rasul et al., 2011). 46 47 Aslam et al. (2018) regard climate change impacts to be particularly threatening to the livestock sector, water 48 and food security and the economy beyond agriculture in South Punjab, particularly as yields decrease. In the 49 livestock sector across global drylands (TS.4.3.2.10, AR 6, WGI), observed impacts include reduction of 50 plant cover in rangelands, reduced livestock and crop yields, loss of biodiversity and increased land 51 degradation and soil nutrient loss (Van de Steeg, 2012; Mganga et al., 2015; Ahmed et al., 2016; Mohamed 52 et al., 2016; Eldridge and Beecham, 2018) as well as injury and livestock death due to SDS. This is 53 particularly worrisome for traditional pastoralists who find themselves with fewer safety nets and limited 54 adaptive capacities than in the past, particularly where mobility, access and tenure rights are becoming 55 restricted (Section 16.3.2.6; Box CCP3.1) and where use of technologies such as mobile phones can result in Do Not Cite, Quote or Distribute CCP3-12 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report 1 mixed effects, as found in Morocco (Vidal-Gonz醠ez and Nahhass, 2018). Observed SDS impacts can 2 increase food production costs and threaten sustainability more generally (Middleton, 2017). 3 4 Woody-plant encroachment and greening may be masking underlying land degradation processes and losses 5 of ecosystem services, livelihood and adaptation options in pastoral livelihood systems (Reed et al., 2015; 6 Chen et al., 2019a). Woody encroachment alters ecosystem services, particularly in rangelands, resulting in 7 reduction of grass cover, hindering livestock production (Anad髇 et al. 2014), reducing water availability 8 (Honda and Durigan 2016, Stringer et al., 2021) but increasing availability of wood (Mograbi et al., 2019). 9 10 11 [START BOX CCP3.1 HERE] 12 13 Box CCP3.1: Pastoralism and Climate Change 14 15 Pastoralism is a livestock keeping system based on the herding of animals. Migrations often take place over 16 long distances to track variable and unpredictable plant growth that tends to be patchy in space and variable 17 in time (Homewood, 2018). Pastoralism has a considerably lower carbon budget than other livestock- 18 keeping systems, with research on pastoralism in the Sahel concluding that this system may be carbon 19 neutral (Assouma et al., 2019), despite contributing directly to greenhouse gas emissions via methane enteric 20 emissions and indirectly through faeces-driven CO2, CH4 and N2O emissions during mineralisation 21 (Assouma et al., 2017). Efforts to sedentarize and villagize pastoralists can lead to land degradation and 22 higher overall emissions from the sector (Section 16.3.2.6). 23 24 Pastoralists migrate with their animals in some of the most remote and marginal environments on the planet. 25 Globally, mobile pastoralists number about 200 million households and use about 25% of the Earth's 26 landmass (Dong, 2016). Many pastoralists operate in non-equilibrial environments that are unstable, 27 fluctuating and generally uncertain, and driven more by climatic variation than livestock numbers and 28 grazing pressure (Behnke et al., 1993). Examples of such systems are grazing areas in the dry tropics 29 (Sandford, 1983; Turner, 1993; Sullivan and Rohde, 2002; Benjaminsen et al., 2006; Hiernaux et al., 2016), 30 and rangelands in the Arctic (Behnke, 2000; Tyler et al., 2008; Benjaminsen et al., 2015; Marin et al., 2020). 31 32 Over many generations, pastoralists have accumulated practical experience and knowledge to cope with 33 uncertainty and value variability (Kr鋞li and Schareika, 2010), mainly through a mobile and flexible 34 approach. While pastoralists are also at risk of climate change impacts, they may be better able to adapt to a 35 changing climate than other land users (Davies and Nori, 2008; Kr鋞li and Schareika, 2010; Jones and 36 Gutzler, 2016). 37 38 While pastoralists possess substantial adaptive capacity as a result of their Indigenous knowledge, this has 39 been under pressure during the last few decades through continued loss of livestock corridors (essential to 40 mobility) and pastures in general due to competing land-uses such as farming, mining, crop expansion and 41 the establishment or extension of protected areas (Th閎aud and Batterbury, 2001; Brockington, 2002; 42 Benjaminsen and Ba, 2009; Upton, 2014; Johnsen, 2016; Tappan, 2016; Homewood, 2018; Weldemichel 43 and Lein, 2019; Bergius et al., 2020). Many of these competing land uses erect fences and exclude other 44 uses, while property rights often privilege sedentary farming. 45 46 Modern states have typically tried to settle pastoralists and confine their movements within clearly defined 47 boundaries, claiming that pastoral land-use is neither ecologically sustainable nor economically productive. 48 Based on such negative and often flawed views, stall-feeding and ranching are often presented by 49 policymakers as successful models of livestock keeping in contrast to the pastoral way of life (Steinfeld et 50 al., 2006; Chatty, 2007). 51 52 Current pressures and processes of pastoral change are spatially variable and complex, and tend to result in 53 further economic and political marginalization of pastoralists, with adverse effects on livelihoods and 54 landscapes. With climate change, which is projected to lead to higher temperatures and more frequent 55 fluctuations in precipitation, maintaining flexibility and resilience in pastoral land use is essential. However, 56 current processes of marginalization, in addition to increased insecurity in some drylands (e.g. the Sahel), 57 make pastoralists more vulnerable, and constrain them from fully employing their adaptive capacities Do Not Cite, Quote or Distribute CCP3-13 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report 1 (Davies and Nori, 2008). The skills and capacities held by pastoralists may, however, offer lessons for 2 society at large in its struggle to adapt to climate change and deal with increased uncertainty (Davies and 3 Nori, 2008; Scoones, 2009; Nori and Scoones, 2019). 4 5 [END BOX CCP3.1 HERE] 6 7 8 CCP3.2.2.4 Gender Differentiated Impacts 9 10 Impacts of desertification, climate change, and environmental degradation, as well as vulnerability and 11 capacity to adapt, are gendered. Differences are determined by socially structured gender-specific roles and 12 responsibilities, ownership of, access to and control over natural resources and technology, decision making, 13 and capacity to cope and adapt to long-term changes (Mirzabaev et al., 2019; Cross-Chapter Box GENDER 14 in Chapter 18). Assessments of the gender dimension of desertification and climate change impacts and 15 responses are scarce, and highly context specific. For example, in many lower income countries, rural 16 women produce most of the household food, and are responsible for food preparation and collecting 17 fuelwood and water from increasingly distant sources (Mekonnen et al., 2017; Droy, 2020). Drought and 18 water scarcity particularly affect women and girls in drylands because they need to spend more time and 19 energy collecting water and fuelwood, have less time for education or income generating activities, and may 20 be more exposed to violence (Sommer et al., 2014) and less able to migrate as an adaptation option. Women 21 are also commonly excluded from family and community decision making on actions to address 22 desertification and climate change, yet their engagement in climate adaptation is critical. International policy 23 efforts are currently seeking to better recognise and address this challenge (Okpara et al., 2019). 24 25 CCP3.2.2.5 Climate Change, Migration and Conflict 26 27 Dryland populations pursuing traditional land-based livelihood options are generally mobile due to a highly 28 fluctuating resource base (Box CCP3.1). Many rural dwellers in drylands also move to urban areas for 29 seasonal work which can have positive impacts in terms of remittances. While reasons for migration vary 30 and can be positive or negative, oppression and human rights abuses, lack of livelihood opportunities and 31 food insecurity tend to be among the main push factors, while emerging opportunities at the rural-urban 32 nexus present lucrative pull factors (Cross-Chapter Box MIGRATE in Chapter 7). In a survey in Libya in 33 2016, 80% of migrants interviewed said they had left home because of economic hardship (Hochleithner and 34 Exner, 2018), which in drylands under water scarcity linked to climate change, would be exacerbated. 35 36 Causes of migration and violent conflict need to be seen in a wider historical, agrarian, political, economic 37 and environmental context, in a multi-scalar perspective integrating levels of analysis from the local to the 38 global (Glick Schiller, 2015). Quantitative studies tend to conclude that climate change has so far not 39 significantly impacted migration including in drylands (Owain and Maslin, 2018), although with some 40 disagreement (Lima et al., 2016; Missirian and Schlenker, 2017). In a study of the climate change-migration- 41 conflict interface, Abel et al. (2019) found limited empirical evidence supporting a link between climatic 42 shocks, conflict and asylum-seeking for the period 2006�2015 from 157 countries. The authors found 43 evidence of such a link for the period 2010�2012 relating to some countries affected by the Arab Spring and 44 concluded that the impact of climate on conflict and migration is limited to specific time periods and 45 contexts. 46 47 The same lack of general causality is largely concluded on the specific link between climate change and 48 conflict (Buhaug et al., 2014; Buhaug et al., 2015; von Uexkull et al., 2016; Koubi, 2019), but a minority of 49 quantitative studies argue for a stronger causal association (Hsiang et al., 2013). Mach et al. (2019) found 50 considerable agreement among experts that climate variability and change have influenced the risk of 51 organized armed conflict within countries, but they also agreed that other factors, such as state capacity and 52 level of socioeconomic development, played a much larger role. These factors also play a role in determining 53 adaptation possibilities and in shaping the enabling environment (Section 8.5.2). 54 55 Qualitative case studies tend to frame conflict and migration within a larger political, economic and 56 historical context. A number of studies from African drylands find that land dispossession is a key driver of 57 both migration and conflict resulting from large-scale resource extraction or land encroachment often Do Not Cite, Quote or Distribute CCP3-14 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report 1 associated with processes of elite capture and marginalization (Benjaminsen and Ba, 2009; Benjaminsen et 2 al., 2009; Cross, 2013; Glick Schiller, 2015; Nyantakyi-Frimpong and Bezner Kerr, 2017; Obeng-Odoom, 3 2017; Bergius et al., 2020). By undermining livelihoods, exacerbating poverty, and setting rural population 4 groups adrift, land dispossession in the Sahel may lead to increased migration to urban areas, to rural sites of 5 non-farm employment (e.g. mines) (Chevrillon-Guibert et al., 2019) or out of the country. In addition, it may 6 lead to other types of reactions including violent resistance (Oliver-Smith, 2010; Cavanagh and 7 Benjaminsen, 2015; Hall et al., 2015) as already seen in the Sahel in terms of the emergence of jihadist 8 armed groups (Benjaminsen and Ba, 2019). Major drivers of the current crisis in Mali include decades of 9 bureaucratic mismanagement and widespread corruption, the spill-over of jihadist groups from Algeria after 10 the civil war there in the 1990s and the current civil war in Libya. Climate change has played a marginal role 11 as a driver of conflicts in the Sahel (Benjaminsen et al., 2012; Benjaminsen and Hiernaux, 2019) but has 12 potential to exacerbate the situation in the future with regards to migration and conflict (Owain and Maslin, 13 2018). 14 15 16 CCP3.3 Future Projections 17 18 CCP3.3.1 Projected Changes and Risks in Natural Systems 19 20 CCP3.3.1.1 Temperature 21 22 Globally, warming rates have been twice as high in drylands compared to humid lands, because the sparse 23 vegetation cover and lower soil moisture of dryland ecosystems amplify temperature and aridity increases 24 (Huang et al., 2016). This enhanced warming is expected to continue in the future. Surface warming over 25 drylands is projected to reach 6.5癈 (3.5癈) under the high RCP8.5 (low-moderate RCP4.5) emissions 26 scenario by the end of this century, relative to the historical period (1961-1990) (Huang et al., 2016; Huang 27 et al., 2017). Exploring the spatial variations between the aeolian desertification response in selected climate 28 change scenarios, Wang et al. (2017) reported that temperature rise could trigger aeolian desertification in 29 West Asia, Central China and Mongolia. The number of extremely hot days with temperatures above 40癈 is 30 projected to increase considerably across the Arab region by the end of the 21st century (ESCWA, 2017). 31 32 CCP3.3.1.2 Rainfall, Evaporation and Drought 33 34 Drylands are highly sensitive to changes in precipitation and evapotranspiration. Potential evapotranspiration 35 (PET) is projected to increase in all regions globally, under all RCPs, as a result of increasing temperatures 36 and surface water vapour deficit (Mirzabaev et al., 2019). Simulations based on coupled land surface, energy 37 and water and vegetation models in the Central Sahel showed a strong response of the water budget. Under 38 +2癈 and +4癈 warming scenarios, decreased evapotranspiration, runoff and drainage were found for all 39 scenarios except those with the highest precipitation (L閍uthaud et al., 2015). 40 Globally, soil moisture declined over the 20th century (Gu et al., 2019), a trend that is projected to continue 41 under all emissions scenarios (WGI AR6). Projected drier soils can further amplify aridity through feedbacks 42 with land surface temperature, relative humidity and precipitation (Berg et al., 2016). 43 Drought conditions (frequency, severity and duration) are expected to substantially worsen in global 44 drylands, driven by a higher saturation threshold and more intense and frequent dry spells under rising 45 temperatures (Liu et al., 2019a; Liu et al., 2019b). In a +1.5癈 world, historical 50-year droughts (based on 46 the Standardised Precipitation-Evapotranspiration Index (SPEI)) could occur twice as frequently across 58% 47 of global landmasses relative to the 1976�2005 period, an area that increases to 67% under 2癈 warming (Gu 48 et al., 2020). Multi-year drought events of magnitudes exceeding historical baselines will increase by 2050 in 49 countries with drylands including Australia, Brazil, Spain, Portugal, and the USA (Jenkins and Warren, 50 2015). The magnitude of drought stress in different regions differs depending on the metric used. Projections 51 based on the Palmer Drought Severity Index (PDSI) suggest drought stress will increase by more than 70% 52 globally, while a substantially lower estimate of 37% is found when precipitation minus evapotranspiration 53 (P-E) is used (Swann et al., 2016). However, the two metrics agree on increasing drought stress in regions 54 with more robust decreases in precipitation, such as southern North America (Section 14.4.3.1), north- 55 eastern South America (Section 12.3.1.1) and southern Europe (Section 13.1.3; Swann et al., 2016). Do Not Cite, Quote or Distribute CCP3-15 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report 1 2 CCP3.3.1.3 Aridity 3 4 Studies based on the AI (the ratio of annual potential evapotranspiration to precipitation), almost always 5 project conditions of increasing aridity under climate change, and associated widespread expansion of 6 drylands (Huang et al., 2016). The limitations of the AI are widely reported (Mirzabaev et al., 2019), with 7 alternative indices that consider different variables including the Ecohydrological Index, PDSI, Standardised 8 Precipitation Index and SPEI (Stringer et al., 2021). AI projections indicate potentially severe aridification in 9 the Amazon, Australia, Chile, the Mediterranean region, northern, southern and west Africa, south-western 10 United States, and South America (medium confidence) (Feng and Fu, 2013; Greve and Seneviratne, 2015; 11 Jones and Gutzler, 2016; Park et al., 2018). However, the AI does not incorporate potential changes to plant 12 transpiration under increasing CO2 concentration and therefore overestimates drought conditions and aridity. 13 Additionally, it does not reflect seasonality in rainfall and evapotranspiration, which is important in regions 14 where temperature and actual evapotranspiration are not increasing during the wet season when vegetation 15 growth is occurring. Mirzabaev et al. (2019) concluded that while aridity will increase in some places (high 16 confidence), there is insufficient evidence to suggest a global change in dryland aridity (medium confidence). 17 Nevertheless, a comparison of several metrics of aridity showed robust aridity increases are projected for 18 several hotspots such as the Mediterranean region and South Africa (Greve et al., 2019). Under RCP8.5, 19 aridity zones could expand by one-quarter of the 1990 area by 2100, increasing to over half of the global 20 terrestrial area (Huang et al., 2016; Lickley and Solomon, 2018). Lower greenhouse gas emissions, under 21 RCP4.5, could limit expansion to one-tenth of the 1990 area by 2100 (Huang et al., 2016). Aridity could 22 expand substantially on all continents except Antarctica (Huang et al., 2016), with expansion first 23 manifesting in the Mediterranean region, southern Africa, southern South America, and western Australia 24 (Lickley and Solomon, 2018). In the Northern Hemisphere, aridity zones could expand poleward as much as 25 11 degrees of latitude (Rajaud and Noblet-Ducoudr�, 2017). By 2100, the population of dryland areas could 26 increase by 700 million people and, under RCP8.5, three billion people might live in areas with a 25% or 27 greater increase in aridity (Lickley and Solomon, 2018). Many studies point to an increasing dryland area 28 based on the AI, but there is low agreement on the actual amount and area of change (Feng and Fu, 2013; 29 Scheff and Frierson, 2015; Huang et al., 2017). The inconsistency between studies is largely due to the 30 substantial internal climate variability in regional precipitation. Changes in annual precipitation have been 31 shown to range from -30% to 25% over drylands. Consistent changes in precipitation are only found at high 32 latitudes, while total PET is projected to increase over most land areas (Feng and Fu, 2013). This leads to 33 more consistent, widespread drying in the tropics, subtropics and mid-latitudes in most models (Feng and Fu, 34 2013; Cook et al., 2014; Scheff and Frierson, 2015; Zhao and Dai, 2015). 35 36 CCP3.3.1.4 Dryland Extent 37 38 Global dryland area (based on the AI) is projected to expand by ~10% by 2100 compared to 1961-1990 39 under a high emission scenario (Chapter 12, WGI). However, there are significant regional differences in the 40 drivers of dryland expansion and subsequent estimates of change in dryland extent. Subtropical drylands are 41 projected to expand as the climate in these regions shifts from temperate to subtropical and aridity increases 42 in currently sub-humid subtropical regions, resulting in the loss of temperature-controlled seasonal cycles 43 (Schlaepfer et al., 2017). Observed and projected warming and drying trends are most severe in transitional 44 climate regions between dry and wet climates, with some exceptions (Nkrumah et al., 2019), which are often 45 highly populated agricultural regions with fragile ecosystems (Cheng and Huang, 2016). In contrast, P-E 46 predicts decreasing drought stress across temperate Asia and central Africa (Swann et al., 2016). Expansion 47 of arid regions is anticipated in southwest North America, the northern fringe of Africa, southern Africa and 48 Australia. The main areas of semi-arid expansion are expected to occur in the north side of the 49 Mediterranean, southern Africa and North and South America. In contrast, India, eastern equatorial Africa 50 and other areas of the southern Saharan regions are projected to have shrinking drylands (Biasutti and 51 Giannini, 2006; Biasutti, 2013; Rowell et al., 2016). Future projections may underestimate dryland 52 expansion, since the Coupled Model Intercomparison Project (CMIP) 5 models underestimate historical 53 warming (Huang et al., 2016) and overestimate precipitation over drylands, particularly in the semi-arid and 54 dry sub-humid regions (Ji et al., 2015). However, estimates vary depending on the metric used (Swann et al., 55 2016; Berg et al., 2017b). Studies based on off-line aridity and drought metrics (calculated from model 56 output of precipitation, evapotranspiration or temperature) project strong surface drying trends (Cook et al., 57 2014; Scheff and Frierson, 2015; Zhao and Dai, 2015), while projections based on total soil water Do Not Cite, Quote or Distribute CCP3-16 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report 1 availability from CMIP5 models show weaker and less extensive drying (Berg et al., 2017a). In contrast, 2 projections in southern Africa may overestimate future drying, with systematic rainfall biases being found in 3 the present-day climatology in models that simulate extreme future drying (Munday and Washington, 2019). 4 Improvements in projections of future changes in aridity require better understanding of seasonality, land 5 hydrology, and the feedbacks between projected soil moisture decrease on land surface temperature, relative 6 humidity and precipitation (Huang et al., 2016). 7 8 Higher dust emissions are consistent with climate change projections indicating an expansion in the global 9 area of drylands (Feng and Fu, 2013; Huang et al., 2016) and increased drought risk (Cook et al., 2014; Xu et 10 al., 2019), but future trends in dust event frequency and intensity as a result of climate change are uncertain 11 and will vary geographically (Jia, 2019). Combined effects of climate change and anthropogenic activities 12 are projected to increase sand encroachment and extreme dust storms (Omar Asem and Roy, 2010; Sharratt 13 et al., 2015; Pu and Ginoux, 2017) as a result of increased aridity, accelerating soil erosion (Section 4.4.8; 14 Sharratt et al., 2015) and loss of biomass (Sharratt et al., 2015; Middleton and Kang, 2017). Shifts in dust 15 storm timings are also projected in some regions (Hand et al., 2016). Dustiness is projected to increase in the 16 southern US Great Plains in the late 21st century under the RCP8.5 climate change scenario but decrease over 17 the northern Great Plains (Pu and Ginoux, 2017). A declining trend in dust emission and transport from the 18 Sahara under RCP8.5 was detected by Evan et al. (2016) but regional climate model experiments conducted 19 by Ji et al. (2018) under the same scenario indicated that overall dust loadings would increase by the end of 20 the 21st century over West Africa. New dust sources may emerge with changing climate conditions, as 21 Bhattachan et al. (2012) indicate for the Kalahari Desert in southern Africa, due to vegetation loss and dune 22 remobilization. There is overall low confidence on future atmospheric dust loads at the global and regional 23 scale. Models of future dust emissions are limited by the low accuracy of models of present anthropogenic 24 dust emissions, which range from 10% and 60% of the total atmospheric dust load (Webb and Pierre, 2018). 25 A global compilation of data from sedimentary archives (ice cores), remote sensing, airborne sediment 26 sampling and meteorological station data estimated that anthropogenic dust emissions have at least doubled 27 over the past 250 years (Hooper and Marx, 2018). While future emissions of natural dust sources are 28 projected to decrease (Mahowald et al., 2006) or remain stable (Ashkenazy et al., 2012), when sources of 29 human emissions are included, projections of future atmospheric dust loads suggest that emissions may 30 increase (Stanelle et al., 2014). 31 32 The relative contribution of albedo and evapotranspiration to regional trends in surface temperature 33 (Charney, 1975) remains unresolved, and may be determined by different mechanisms in different systems, 34 depending on site-specific conditions such as snow coverage, vegetation and soil moisture (Yu et al., 2017). 35 For example, the vegetation-albedo feedback mechanism may dominate in the Arctic (Blok et al., 2011; te 36 Beest et al., 2016), while the vegetation-evaporation feedback may drive change in other regions. Actions 37 that increase forest cover across Africa could thus, theoretically, moderate projected future temperature 38 increases (Wu et al., 2016; Diba et al., 2018), but with potentially negative effects on biodiversity (Chapter 39 2). Soil drying exacerbates atmospheric aridity, which causes more soil drying in a self-reinforcing land� 40 atmosphere feedback that could intensify under RCP8.5 (Zhou et al., 2019). 41 42 Changes to the composition, structure and functioning of natural communities in deserts and dryland 43 ecosystems are key risks resulting from water stress, drought intensity and continued habitat degradation, 44 greater frequency of wildfire, biodiversity loss and the spread of invasive species (Hurlbert et al., 2019). Not 45 all these stresses occur at the same time in a particular environment, with some areas more exposed to e.g. 46 wildfire than others, especially in areas with high amounts of dry herbaceous biomass. Grassland 47 composition may shift as C3 plants are replaced by C4 species, which have higher optimal temperatures and 48 higher water use efficiency (although seasonality of precipitation also plays a role) (Knapp et al., 2020). 49 Many desert species have morphological, physiological and/or behavioural adaptations to cope with climatic 50 extremes, including rapid regeneration following droughts (Boudet, 1977; Hiernaux and Le Hou閞ou, 2006), 51 leaf dropping during the dry season to reduce water loss (Santos et al., 2014), alongside long histories of 52 adaptation to climate change (Brooks et al., 2005; Ballouche and Rasse, 2007), while many animals live near 53 their physiological limits (Vale and Brito, 2015). Substantial ecological effects may occur when extreme 54 events such as heatwaves or droughts are superimposed on the warming trend, pushing species beyond their 55 physiological and mortality thresholds (Hoover et al., 2015; Harris et al., 2018). 56 Do Not Cite, Quote or Distribute CCP3-17 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report 1 Climate change increases risks of continued range retractions of Karoo succulents in South Africa (Young et 2 al., 2016), dry argan woodlands in Morocco (Alba-S醤chez et al., 2015), epiphytic cacti in Brazil 3 (Cavalcante and Duarte, 2019; Cavalcante et al., 2020) and other plant species exposed to higher aridity. 4 Projected increases in heat and aridity could increase mortality of trees and shrubs in Sonoran Desert 5 ecosystems in the United States (Munson et al., 2012; Munson et al., 2016b), reduce sagebrush in arid 6 ecosystems of the western United States (Renwick et al., 2018), and contribute to the replacement of 7 perennial grasses with xeric shrubs in the south-western United States (Bestelmeyer et al., 2018). CO2 8 fertilization and warmer conditions, combined with changes in timing and availability of moisture, could 9 increase invasive grasses and wildfire in desert ecosystems of Australia and the south-western United States 10 where wildfire has historically been absent or infrequent (Abatzoglou and Kolden, 2011; Horn and St. Clair, 11 2017; Klinger and Brooks, 2017; Syphard et al., 2017). Trends of woody encroachment may continue in 12 some North American and African drylands or at least not reverse (Higgins and Scheiter, 2012; Caracciolo et 13 al., 2016). Impacts of woody encroachment on drylands may show a slight increase in carbon, but a decline 14 in water and huge negative impacts on biodiversity, with a tendency for open ecosystem species to be most 15 affected (Archer et al., 2017). Expansion of grasses into these arid shrublands has the potential to transform 16 them rapidly, especially through the acceleration of the fire cycle (Bradley et al., 2016). While the impact of 17 increased aridity may be offset by changing water use efficiency by plants under high CO2 concentrations, 18 limiting the expansion of dryland ecosystems (Swann et al., 2016; Mirzabaev et al., 2019), increased plant 19 growth in response to elevated CO2, which results in increased water consumption, may counteract this. 20 Increased water use efficiency is therefore not expected to counterbalance increased evaporative demand 21 (Chapter 8). There is medium confidence that succulent species will be particularly vulnerable to increased 22 heat and aridity due to reduced physiological performance, loss of seed banks, lower germination rates and 23 increased mortality (Table CCP3.1; Musil et al., 2005; Arag髇-Gast閘um et al., 2014; Shryock et al., 2014; 24 Martorell et al., 2015; Carrillo-Angeles et al., 2016; Arag髇-Gast閘um et al., 2017; Komiska et al., 2019). 25 26 CCP3.3.2 Projected Impacts on Human Systems 27 28 Across many drylands, human-induced causes of desertification, SDS, climate change and unsustainable 29 land use, are projected to become more pronounced over the next several decades with global consequences. 30 Future climate changes with increasing frequency, intensity and scales of droughts and heatwaves, are 31 projected to further exacerbate the vulnerability and risk to humans from desertification (Hurlbert et al., 32 2019). 33 34 Sand and dust storms exert a wide range of impacts on people, within deserts and semi-deserts but also 35 outside dryland environments because of long-range dust transport (Middleton, 2017). Research on the 36 economic impacts of SDS is lacking, while studies that have been conducted lack consistency in data 37 collection methods and analysis (Middleton, 2019). Although projections are rarely modelled, estimated 38 economic damages of increased dust-related health impacts and mortality under RCP8.5 could total $47 39 billion/year additional to the 1986-2005 value of $13 billion/year in southwest USA (Allahbakhshi et al., 40 2019). 41 42 Projected impacts of climate change on the risk of food insecurity are a particular concern for the developing 43 world drylands (Chapter 16, WGI; Mirzabaev et al., 2019), potentially leading to breakdown of food 44 production systems, including crops, livestock, and fisheries, as well as disruptions in food supply chains and 45 distribution (Myers et al., 2017; Lewis and Mallela, 2018). Developing country drylands are particularly 46 vulnerable due to a higher share of populations with lower income, lower physical access to nutritious food, 47 social discrimination as well as other environmental factors that link to climate change. For example, 48 countries such as Somalia, Yemen and Sudan faced recent and resurging challenges from an increase in 49 desert locusts, the effects of which in 2020 extended from East Africa through the Arabian Peninsula and 50 Iran as far as India and Pakistan. Meynard et al. (2020) note that under climate change, some areas suffering 51 from previous outbreaks may see changes in formation of swarms of Schistocerca gregaria. Salih et al. 52 (2020) recognise that attributing the 2020 swarms as a single event to climate change remains challenging, 53 but highlight that projected temperature and rainfall increases in deserts and strong tropical cycles can create 54 conditions conducive to the development, aggregation, outbreak and survival of locusts. Mandumbu et al. 55 (2017) highlight how crop parasites such as Striga spp. in southern Africa may benefit from higher 56 temperatures and rainfall activating dormant seeds, while high winds aid their dispersal. Combined with 57 increasing risks of erosion and soil fertility losses (Striga is able to tolerate drought and a low nitrogen Do Not Cite, Quote or Distribute CCP3-18 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report 1 environment), this can have important impacts on the yields of key dryland crops such as maize and pearl 2 millet. 3 4 Human responses can exacerbate desertification processes under climate change conditions, even in deserts. 5 Exploitation of mineral resources (e.g. lithium mining in Chile's Atacama Desert) can cause human 6 population changes as people flock to the area for work (Liu et al., 2019), increasing vulnerability due to e.g. 7 soil erosion and salinisation, as well as increasing pressure on potable water for human consumption 8 (Stringer et al., 2021) and exhausting aquifers. Salinisation is projected to increase in the drylands due to 9 climate change impacts in future (Mirzabaev et al., 2019). For example, in India, about 7 million ha arable 10 land area is currently salt-affected (Sharma et al., 2015; Sharma and Singh, 2015). It is projected that 11 unsustainable use of marginal quality waters in irrigation and neglect of drainage, combined with 12 climate change impacts, will accelerate land salinization in India, rendering another 9 million ha area 13 salty and less productive by 2050 (ICAR-CSSRI, 2015). This has important cost implications given that 14 annually, 16.84 million tonnes of farm production valued at INR 230.19 billion is already lost in India 15 due to salinity and associated problems (Sharma and Singh, 2015). The literature further shows evidence 16 of desertification of oases and irrigated lands in parts of northern China's drylands (Wang et al., 2020), the 17 Indian subcontinent's deserts, as well as the Mesopotamian Arabian Desert (Ezcurra, 2006; Dilshat et al., 18 2015). 19 20 21 CCP3.4 Adaptations and Responses 22 23 Adaptations to climate change impacts in human systems vary depending on exposure to risks, types of risks 24 and responses, underlying social vulnerabilities and adaptive capacities, including access to resources, the 25 extent of adaptation responses and the potential of these responses to reduce risk/vulnerability (Chapter 16 ; 26 Singh and Chudasama 2021). Adaptations tend to be applied locally, tackling symptoms of the problem and 27 proximate drivers (e.g. of desertification), rather than distant or external drivers (Morris et al., 2016; Adenle 28 and Ifejika Speranza, 2021). Different groups require different kinds of supports and levers to enable them to 29 follow adaptive pathways (Stringer et al., 2020; M鴏ler et al., 2017) and face different barriers and limits to 30 adaptation (Chapter 18, WG2). What constitutes an incremental adaptation in one location may be 31 transformational in another. Spatial patterns of dryland resilience and adaptive capacity can be partly 32 explained by access to livelihood capitals (Mazhar et al., 2021) and are shaped by prevailing structures and 33 power dynamics. Supportive policies, institutions and good governance approaches can strengthen the 34 adaptive capacities of dryland farmers, pastoralists and other resource users (high confidence) (Stringer et al., 35 2017). Table CCP3.2 provides examples of illustrative adaptation options responding to major challenges of 36 climate change and desertification in deserts and semi-arid areas. Some adaptations present no-regrets 37 options while others tackle desertification and/ or climate changes to different extents. 38 39 40 Table CCP3.2: Synthesis of adaptation measures and responses to risks in deserts and semi-arid areas. Appropriateness 41 of measures is context dependent and some adaptations will be incremental or even maladaptive in some dryland 42 contexts while being transformational in other locations. 43 [INSERT TABLE CCP3.2 HERE] 44 45 46 Adaptations to climate change, desertification, drought management (Section 17.2.2.2) and sustainable 47 development activities largely overlap in drylands, pointing to synergies between them (Reichhuber et al., 48 2019). For example, support for communal and flexible land tenure could bring about benefits across 49 multiple dimensions, while attention to water as a limiting factor in drylands can link to multiple SDGs 50 (Stringer et al., 2021) as well as adaptations in natural systems, where improved forecasting and anticipatory 51 science and management can be appropriate (Bradford et al., 2018). Currently, more than 125 countries 52 around the world, particularly in drylands, are setting land degradation neutrality (LDN) targets. LDN and its 53 hierarchical response mechanisms of avoiding, reducing and reversing land degradation, can provide an 54 overarching resilience-based framework for adaptation at the national level (Mirzabaev et al., 2019; Orr et 55 al., 2017b; Cowie et al., 2018) and support biodiversity conservation (Akhtar-Schuster et al., 2017). 56 However, achieving LDN will require a transparent decision and prioritisation process (Dallimer and 57 Stringer, 2018), anchored in a socio-ecological systems approach (Okpara et al., 2018), with investment in Do Not Cite, Quote or Distribute CCP3-19 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report 1 all dimensions of an enabling environment, including inclusive policies and regulations, sustainable 2 institutions, accessible finance and effective science-policy communications and interactions (Verburg et al., 3 2019; Allen et al., 2020). LDN calls for integrated land use planning to ensure land uses are optimized at a 4 landscape scale to help balance competition for limited land resources and harness multiple benefits (Cowie 5 et al., 2018, Verburg et al., 2019), recognising that adaptations present synergies and trade-offs along various 6 dimensions of sustainable development such as poverty reduction, enhancing food security and human health 7 or providing improved access to clean energy, land water, and finance (see Section 8.6). Distributional 8 effects of adaptation options also may vary between different socio-economic groups within countries or 9 locally among communities, pushing social justice concerns to the fore (Section 8.4). Measures promoting 10 particular adaptations need to take into account such consequences as well as the potential for some 11 adaptations to become maladaptive at scale. 12 13 Natural systems are also able to adapt to climate change, be adapted and become more resilient to 14 desertification. For example, the root network architecture of the hyper-arid Negev Desert acacia trees has 15 enabled them to withstand intensive cultivation and climate-change driven desertification (Winter et al., 16 2015) while vegetation-induced sand mounds ("coppice dunes") in the Arabian Desert have reduced 17 desertification through reducing wind erosion and enriching sand desert land with water and nutrients (Quets 18 et al., 2017). Vegetation cover of psammophyte shrub species (in the "desert oasis transitional area") 19 surrounding the Dunhuang Oasis (northwest China) reduces oasis land degradation risk by reducing sand 20 grain size and velocity of winds from the aeolian desert (Zhang et al., 2007); while land use planning in 21 Israel's Negev Desert taking a `sharing' approach between cultivation and urbanization has helped to 22 minimise the external degrading effects of adjacent desert land ecosystems (Portnov and Safriel, 2004). 23 Scholars are nevertheless questioning the wider suitability of tree planting in drylands, given concerns for 24 water availability and other ecosystem services (Veldman et al., 2015; Bond et al., 2019; Veldman et al., 25 2019). How natural dryland systems are managed following disturbances such as wildfire is important too. 26 van den Elsen et al. (2020) found that establishing vegetation and mulch cover after a fire in a Mediterranean 27 dryland ecosystem reduced soil erosion, helping maintain soil fertility and nutrients. However, different 28 management objectives require different adaptations. For example, adaptation measures that reduce land 29 degradation through reforestation could increase vulnerability to fire if they exclude ecologically sound fire 30 management or are based on plant species that are fire prone. Combinations of different land management 31 practices and governance approaches tackling a range of different stresses appear to best support 32 sustainability and adaptation over the long term (van den Elsen et al., 2020). 33 34 Collective action can facilitate the implementation of adaptation responses and help tackle challenges 35 associated with upscaling of successful land-based adaptations (Thomas et al., 2018). However, a lack of 36 coordination between stakeholders and across sectors can be problematic (Amiraslani et al., 2018), showing 37 the importance of multi-stakeholder engagement (De Vente et al., 2016). Multi-stakeholder engagement is 38 recognized as an essential part of desertification control, as well as vital in tackling climate change (Reed 39 and Stringer, 2016), with participation taking place to different extents in different drylands according to the 40 prevailing governance system. In China, the Grain for Green programme is an example of a large-scale 41 ecological restoration programme securing local engagement through payments for ecosystem services 42 (Kong et al., 2021), while transdisciplinary stakeholder engagement involving researchers and central and 43 local governments in the Heihe River Basin in China's arid and semi-arid northwest, using an 44 interdisciplinary `web' approach, enabled basin restoration. Multi-stakeholder efforts saw improvement in 45 the condition of Juyan Lake and the surrounding catchment, increasing both the lake surface area and 46 groundwater in downstream locations (Liu et al., 2019). 47 48 In the short- to medium-term, monitoring, prediction and early warning can support adaptation and e.g. help 49 reduce negative impacts of SDS by mobilising emergency responses. Daily dust forecasts enable preparation 50 to minimise risks from sand/dust storms to both human and natural systems (e.g. the WMO Sand and Dust 51 Storm Warning Advisory and Assessment System: https://sds-was.aemet.es/forecast-products/dust- 52 forecasts). Preparedness and emergency response procedures benefit from covering diverse sectors, such as 53 public health surveillance, hospital services, air and ground transportation services, water and sanitation, 54 food production systems and public awareness, suggesting the need for a coherent, multi-sector governance 55 approach. Longer-term actions include prioritizing sustainable land management (Middleton and Kang, 56 2017), based on Indigenous knowledge and local knowledge, and modern science (Verner, 2012), along with 57 the investment of financial and human capital in supporting these measures. Devolved adaptation finance in Do Not Cite, Quote or Distribute CCP3-20 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report 1 dryland areas of e.g. Kenya (Nyangena and Roba, 2017) and Mali (Hesse, 2016) has yielded promising 2 insights, highlighting the importance of climate information services and local government support for 3 community prioritisation of adaptation activities. Such actions can enable substantial benefits for poor and 4 marginalised men and women. Among international institutional measures, a global coalition to combat SDS 5 was launched at the United Nations Convention to Combat Desertification Conference of Parties (UNCCD 6 COP14) in 2019, which could help to better mobilize a global response to SDS. Similarly, there have been 7 calls for increased investment in regional institutions such as the Desert Locust Control Organisation for 8 Eastern Africa to both pre-empt and tackle locust plagues (Salih et al., 2020), requiring trans-boundary 9 cooperation. 10 11 There is high agreement and robust evidence that shifting emphasis to proactive risk mitigation, including 12 solutions for drought, flooding erosion and dust management, instead of exclusive focus on disaster 13 management, reduces vulnerability and improves adaptive capacity (Section 16.4.3.2 and 16.5.2.3.4; 14 Sivakumar, 2005; Grobicki et al., 2015; Wieriks and Vlaanderen, 2015; Aguilar-Barajas et al., 2016; 15 Runhaar et al., 2016; Wilhite and Pulwarty, 2018; Wilhite, 2019). It also underscores the LDN response 16 hierarchy avoid > reduce > reverse (Orr et al., 2017a). Nevertheless, ex ante drought and flood risk 17 mitigation has been adopted in limited dryland settings, despite that it is preferable to increase preparedness 18 before it happens, provide incentives for adaptation instead of insurance, provide insurance instead of relief, 19 and provide relief instead of regulation (Sivakumar, 2005). Yet, providing disaster relief is often more 20 publicly visible and politically expedient, despite its social, economic and environmental challenges. The 21 absence of proactive risk mitigation and resulting crisis management increases vulnerability, increases 22 reliance on government support, reduces self-reliance and increases costs (Grobicki et al., 2015; Wilhite, 23 2019), as well as hindering progress towards the SDGs. In the case of drought and flooding, major obstacles 24 for the transition from reactive management to proactive drought risk mitigation include path dependencies 25 and lack of knowledge about relative costs and benefits of reactive versus proactive approaches. This lack of 26 information can deter large-scale and long-term investments into proactive approaches (Mirzabaev, 2016). 27 28 A range of risk mitigation and adaptation measures can be taken, to address drought, desertification and 29 other climate change-related challenges in deserts and semi-arid areas, some of which can be both proactive 30 and reactive. These include inter alia: 31 i) Policies, public advocacy, and social media campaigns that improve water use 32 efficiency, especially in agriculture and industry, which can foster behavioural changes 33 and reduce water consumption (Yusa et al., 2015; Tsakiris, 2017; Booysen et al., 2019), 34 ii) Integrating access to insurance, financial services, savings programs, and cash transfers 35 into policies to increase the effectiveness of e.g. drought responses. Such efforts can 36 result in significant cost savings (Berhane et al., 2014; Bazza et al., 2018 ; Guimar鉫s 37 Nobre et al., 2019), 38 iii) Development of robust early warning systems that provide information and improve 39 knowledge surrounding drought and SDS to enable early recovery (Wilhite, 2019), 40 considering also vulnerability and impact assessments (i.e. who is at greatest risk), 41 iv) Water management and storage, including using methods that draw on Indigenous 42 knowledge (Stringer et al., 2021), water transfers, and trade, all of which can reduce 43 costs and provide timely adaptations to drought, supporting agricultural productivity and 44 rural livelihoods (Harou et al., 2010; Hurlbert, 2018), 45 v) Restoration, reclamation, and landscape heterogeneity strategies, promoting ecosystem 46 resilience to wind erosion and dust abatement (Duniway et al., 2019) as well as restoring 47 important ecosystem services at a catchment scale, 48 vi) Prevention of soil erosion, provision of dust abatement and enhanced biodiversity by 49 changing grazing techniques (e.g. rotational grazing), facilitating herd mobility, 50 protecting rangeland areas from fragmentation, promoting common tenure and access 51 rights on grazing land, enabling rapid post fire restoration efforts, minimum tillage, 52 sustainable land management, integrated landscape management, planting and caring for 53 non-irrigated indigenous trees and other vegetation (Middleton and Kang, 2017); and 54 vii) Creation of drought tolerant food crops through participatory plant breeding (Grobicki et 55 al., 2015) and investment in research and development of drought resistant varieties 56 (Basu et al., 2017; Mottaleb et al., 2017; Dar et al., 2020), alongside adjusted planting 57 and harvesting periods (Frischen et al., 2020). Similar to other adaptations, the net Do Not Cite, Quote or Distribute CCP3-21 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report 1 economic benefits of ex ante resilient plant development far outweigh the research 2 investment (Basu et al., 2017; Mottaleb et al., 2017; Dar et al., 2020). 3 4 Many of these measures can also support climate change mitigation efforts in drylands. Uptake of adaptation 5 measures is often grounded in clear communications and information provision to support behavioural 6 changes, taking into account local risk aversion and risk perceptions (Zeweld et al., 2018; Jellason et al., 7 2019). Building capacity by improving the knowledge base and access to information as well as to financial 8 and other resources, encourages vulnerable economic sectors and people to adopt more self-reliant measures 9 that promote more integrated and sustainable use of natural resources (high confidence) (Sivakumar, 2005; 10 Wieriks and Vlaanderen, 2015; Aguilar-Barajas et al., 2016; Middleton and Kang, 2017; Wilhite, 2019). 11 Engaging natural resource users as active participants in planning and technology adoption using extension 12 services, financial grants and services geared to the local area, can build resilience and drive changes in 13 practices (Webb and Pierre, 2018), while approaches such as Integrated Water Resources Management 14 (IWRM) can support adaptation and drought risk management, including in dryland urban megacities 15 (Stringer et al., 2021) and in deserts and semi-arid areas where precipitation trends remain stable yet other 16 pressures on water are growing (Reichhuber et al., 2019). 17 18 19 [START FAQ CCP3.1 HERE] 20 21 FAQ CCP3.1: How has climate change already affected drylands and why are they so vulnerable? 22 23 Human-caused climate change has so far had mixed effects across the drylands, leading to fewer trees and 24 less biodiversity in some areas and increased grass and tree cover in others. In those dryland areas with 25 increasing aridity, millions of people face difficulties in maintaining their livelihoods particularly where 26 there is water scarcity. 27 28 Drylands include the hottest and most arid areas on Earth. Human-caused climate change has been 29 intensifying this heat and aridity in some places, increasing temperatures more across global drylands than in 30 humid areas. In areas which are hotter and drier, tree death has occurred and in some locations bird species 31 have been lost. Climate change has reduced rainfall in some dryland areas and increased rainfall in other 32 areas. Increased rainfall, combined with the plant-fertilizing effect of more carbon dioxide in the atmosphere, 33 can increase grass and shrub production in dryland areas. Because water is scarce in drylands and aridity 34 limits the productivity of agriculture, millions of people living in drylands have faced severe difficulties in 35 maintaining their livelihoods. This challenge is exacerbated by non-climate change factors, such as low 36 levels of infrastructure, remoteness, and limited livelihood options that are less dependent on scarce natural 37 resources. High temperatures in drylands increase the vulnerability of people to potential heat-related 38 illnesses and deaths from heat under continued climate change. 39 40 [END FAQ CCP3.1 HERE] 41 42 43 [START FAQ CCP3.2 HERE] 44 45 FAQ CCP3.2: How will climate change impact the world's drylands and their people? 46 47 Climate change is projected to lead to higher temperatures across global drylands. Many drylands also risk 48 more irregular rainfall leading to increased irregularity in crop yields, and increased water insecurity 49 where less rainfall is projected, which may have profound implications for both dryland ecosystems and 50 their human inhabitants. 51 52 There is, however, considerable uncertainty about the changes that may occur in drylands in the future and 53 how people and ecosystems will be affected. In some drylands, higher temperatures and declining rainfall 54 have increased aridity. However, this is not a global trend as many drylands are experiencing increases in 55 vegetation cover and rainfall. Both the amount of rainfall and its seasonality have changed in many dryland 56 areas, associated with natural variability and warming. 57 Do Not Cite, Quote or Distribute CCP3-22 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report 1 Most climate models project increased rainfall in tropical drylands, but more variability. High natural 2 climatic variability in drylands makes predictions uncertain. Understanding future impacts is further 3 complicated by many interacting factors such as land use change and urbanisation that affect the condition of 4 drylands. Future trends in sand and dust storm activity are also uncertain and will not be the same 5 everywhere, but there will likely be increases in some regions (e.g. the United States) in the long-term. The 6 impacts of climate change in deserts and semi-arid areas may have substantial implications globally: for 7 agriculture, biodiversity, health, trade and poverty, as well as potentially, for conflicts and migration. 8 Increasing temperatures and more irregular rainfall are expected to affect soil and water and contribute to 9 tree death and loss of biodiversity. In other places, woody encroachment onto savannas may increase, in 10 response to the combination of land use change, changes in rainfall, fire suppression, and CO2 fertilization. 11 Crop yields are projected to decline in some areas, with adverse impacts on food security. The potential for 12 conflicts and migration is primarily associated with socioeconomic development, while links to climate 13 change remain uncertain and lack evidence. 14 15 [END FAQ CCP3.2 HERE] 16 17 18 [START FAQ CCP3.3 HERE] 19 20 FAQ CCP3.3: What can be done to support sustainable development in desert and semi-arid areas, 21 given projected climate changes? 22 23 Water is a major limiting factor in drylands. Many efforts to support sustainable development aim to 24 improve water availability, access and quality, ranging from large engineering solutions that move or 25 desalinise water; to herders' migrations with their animals to locations that have water; to land 26 management and water harvesting practices that conserve water and support land cover. These solutions 27 draw on Indigenous knowledge, local knowledge and innovative science, and can help to address multiple 28 Sustainable Development Goals. 29 30 Different desert and semi-arid areas can benefit from different incremental and transformational solutions to 31 move toward sustainable development under climate change. In some dryland areas facing critical water 32 shortages, transformational adaptations may be needed - for example, large-scale water desalination when 33 they have access to sea water, despite high energy use and negative environmental impacts of waste brine. In 34 dryland agricultural areas across the world, incremental adaptations include water conservation measures, 35 use of improved crop varieties or increasing herd mobility. What counts as a transformational change in 36 some places may be incremental in others. Often solutions can target multiple development goals. For 37 example, water harvesting can make water available during drought, buffering water scarcity impacts, while 38 also supporting food production, agricultural livelihoods and human health. Land based approaches, e.g. 39 restoration of grassland, shrubland, and savanna ecosystems, are important for ensuring ecological integrity, 40 soil protection and preventing livelihoods from being undermined as a result of growing extreme weather 41 events. It is important that policies, investments and interventions that aim to support sustainable 42 development take into account which groups are likely to be most affected by climate change. Those people 43 directly dependent on natural resources for their survival are generally most vulnerable but least able to 44 adapt. The capacity to translate local and Indigenous knowledge and experience into actions can require 45 external support. Governments and other stakeholders can help by investing in early warning systems, 46 providing climate information, realigning policies and incentives for sustainable management, investing in 47 supporting infrastructures, alongside developing alternative livelihood options that are less exposed and 48 sensitive to climate change. Involving all relevant stakeholders is important. For example, in China the Grain 49 for Green programme secured local engagement by paying people to manage the environment more 50 sustainably. At a global level important groups have emerged to cooperate and offer solutions around issues 51 such as sand and dust storms, and integrated drought management. Efforts are needed across all scales from 52 local to global to support sustainable development in desert and semi-arid areas, given projected climate 53 changes. 54 55 56 Do Not Cite, Quote or Distribute CCP3-23 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report Large Tables Table CCP3.1: Observed ecological changes in drylands. Region Observed change Climate change Attribution Non-climate Confidence in References factors to climate change factors observed change change Hyper arid Asian hyper Loss of shallow rooted desert Increase in Yes. Livestock, human- Medium Li et al. (2015) arid regions plants extreme warm Analyses of ignited fires Medium Iknayan and Beissinger (Gobi) temperatures causal (2018); Riddell et al. (2019) North Loss of mesic bird species Decreased rainfall factors find High America - decreased High Lovich et al. (2014) Mojave rainfall more Desert important Defalco et al. (2010); Munson than non- et al. (2016b); Conver et al. climate (2017) factors. Arid Decline of desert tortoise Decreased rainfall Land use change, African Sahel (Gopherus agassizii) invasive plant population 90% from 1993 to Increased species 2012 at one site in the temperature, Mojave decreased rainfall, Restoration Reduced perennial vegetation wildfire planting cover, including trees and Agroforestry cacti, in the Mojave and Increase in Sonoran deserts of the rainfall since the southwestern United States mid-1990s (compared to Woody cover increase in 1968-1993)and parts of the Sahel increased CO2 Do Not Cite, Quote or Distribute CCP3-24 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report Increase in grass production Increases in Medium Hiernaux et al. (2009a); across Sahel rainfall since the Hiernaux et al. (2009b); mid-1990s Dardel et al. (2014); Venter et (compared to al. (2018); Zhang et al. (2018); 1968-1993) and Brandt et al. (2019); increased CO2 Bernardino et al. (2020) Decline of mesic tree species Decreased rainfall Yes. Land clearing for High Gonzalez (2001); Wezel and at field sites across the Sahel from 1901 to multivariate cropland expansion, Lykke (2006); Maranz (2009); 2002 increased statistical Increase pressure Gonzalez et al. (2012); H鋘ke Increased tree mortality at temperature analyses on wood resources et al. (2016); Kusserow field sites across the Sahel find climate (rural demography, (2017); Ibrahim et al. (2018); Decreased rainfall factors more urbanization) Zida et al. (2020b) from 1901 to important 2002, increased than non- Agricultural High Helld閚 (1984); Gonzalez, temperature climate expansion, (2001); Wezel and Lykke factors. modified runoff on (2006); Maranz (2009); Yes. shallow soils Vincke et al. (2010); H鋘ke et multivariate al. (2016); Trichon et al. statistical (2018); Zwarts et al. (2018); analyses Wendling et al. (2019); find climate Bernardino et al. (2020); Zida factors more et al. (2020a) important than non- Boudet (1977); Tucker and climate Nicholson (1999); Gonzalez, factors. (2001); Hiernaux and Le Hou閞ou (2006); Hiernaux et Latitudinal biome shift of the Decreased Yes. High al. (2009a); Maranz (2009); Sahel rainfall, increased multivariate Medium Gonzalez et al. (2012) temperature statistical Increase in woody plant analyses Morgan et al. (2004); Haensler cover and a shift of mesic Increase in find climate et al. (2010); Donohue et al. amount of fog factors more (2013); Rohde et al. (2019) from westward important than non- climate factors. Namib desert Do Not Cite, Quote or Distribute CCP3-25 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report species into more arid expansion of Medium Du Toit and O'Connor (2014); regions convective du Toit et al. (2015); rainfall and Masubelele et al. (2015a); Southern increase in Masubelele et al. (2015b) Africa - number of Nama-Karoo extreme rainfall Davis-Reddy (2018) events. Elevated CO2 and warming Polley et al. (1997); Morgan et effects on the al. (2004); Donohue et al. Bengula (2013); Ward et al. (2014); upwelling system Masubelele et al. (2015a); Shifting rainfall Hoffman et al. (2018) seasonality Foden et al. (2007a); Jack et (debate if its al. (2016) cyclical or directional); Karmaoui et al. (2014) elevated CO2 Eastern Karoo has Shift in rainfall Low experienced a significant seasonality and increase in the end of the increase in MAP Changing land use Medium growing season length Rising and herbivore Woody encroachment has concentration of management been observed throughout the CO2 Southern Nama-Karoo in valley Medium Africa - bottoms, ephemeral stream Warming and Succulent banks and the slopes of drying Agricultural Medium Karoo Karoo hills. growth, high Succulent Karoo: Range shift High temperature population growth Northern in tree aloe Aloidendron and reduced Africa - dichotomum with mortality in precipitation Morocco the warmer and drier range and increase in recruitment in the cooler southern range, populations have positive growth rates, possibly due to warming, although this finding has been challenged Increased vulnerability of oasis's, and reduced ecosystem service provision Do Not Cite, Quote or Distribute CCP3-26 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report causing soil and and unregulated water salinization, and indiscriminate drying up of land development surface water. Hot Reduced surface water winds and High demand Medium Rochdane et al. (2012); availability sandstorms. Choukri et al. (2020) Increased (population growth) Reduction of resilience of temperature and Navarro-Cerrillo et al. (2020) Abies pinasapo- Cedrus reduced and landuse change atlantica forests to precipitation Collins and Xia (2015); subsequent droughts Successive Medium Rudgers et al. (2018) Drought adapted species are droughts increasing in Chihuahuan Medium Caracciolo et al. (2016); deserts Increase in aridity Archer et al. (2017) North and increased Fire suppression High American Widespread woody plant inter-annual and altered Rutherford et al. (2000) drylands encroachment. Prosopis sp variation in grazing/browsing encroachment in arid desert climate trends regimes, P閞ez-S醤chez et al. (2011); South regions (Chihuahuan and Increasing Castill髇 et al. (2015); Sosa et Chihuahuan Sonoran Desert) at a rate of temperature, Medium al. (2019) Desert - ~3% per decade. elevated CO2 and North and Plant desert community shift changing rainfall Urban growth, High central changes the albedo through mechanized Mexico the reduction in dark Warming and agriculture, and biocrusts drought changes in land use Shrub encroachment of grassland (Berberis Decreased trifoliolata, Ephedra aspera, rainfall, Larrea tridentata) changes increase in on dominant species in shrub temperature and areas loss of less resistant increase CO2 shrubby species (Leucophyllum laevigatum, Lindleya mespiloides, Setchellanthus caeruleu). Shrub encroachment of mesic and temperate areas Do Not Cite, Quote or Distribute CCP3-27 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report changes in land use Low Shifts on soil microbial decreased rainfall Vargas-Gast閘um et al. (2015) community to more abundant and Medium in fungi (Ascomycota and increase in Sosa et al. (2019) Pleosporales) temperature Cattle grazing, High Limited ecological decreased rainfall looting Arag髇-Gast閘um et al. connectivity of shrubby + (2014); Shryock et al. (2014); Arid and populations increase in human-induced Low Martorell et al. (2015); semi-arid temperature Carrillo-Angeles et al. (2016); territories in Loss of Cacti species decreased rainfall land degradation Arag髇-Gast閘um et al. (2018) Argentina (Echinocactus platyacanthus, + Argentina Pediocactus bradyi, increase in Land use change Medium Barbosa et al. (2015) Chaco Coryphantha werdermannii, temperature Overexploitation of High Region Astrophytum) due to decline water resources Amdan et al. (2013); South in physiological Decreased rainfall Marchesini et al. (2017) America Arid performance, loss of seed High Diagonal banks and lower germination changes in rainfall Total pages: 53 Bianchi et al. (2017); Rivera rates and Penalba (2018); Masiokas South Decreases in vegetation Decrease in et al. (2019); Rodr韌uez- American indexes precipitation in Morales et al. (2019) Andes the upper Andes. Dryland salinity The Dussaillant et al. (2019); unprecedented 10- Falaschi et al. (2019); Marked reduction in year extreme dry Masiokas et al. (2019) streamflow from the Andes period has been mountain "water towers" due called the "Mega- to the persistent reduction in drought precipitation." Increasing sub- continental Extensive glacier retreat temperature and across the Andes regional reduction in snow precipitation Do Not Cite, Quote or Distribute CCP3-28 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report High Patagonian Widespread tree mortality of Increase in High Rodr韌uez-Cat髇 et al. (2019) Andes Austrocedrus and Nothofagus extreme drought forests in the dry ecotone events Medium (Srur et al. (2016); Srur et al. Central Asian forest-steppe across (2018) arid lands Patagonia Increase in The land use and High Increase in elevation of the temperature and cover change, High Li et al. (2015) Loess upper-forest Nothofagus duration of the ecological Plateau, treeline across Patagonia growing season at restoration, mainly Jia et al. (2015); Wang et al. China high elevation in induced by Grain (2015); Deng et al. (2016); Shrub encroachment into arid the Patagonian for Green Project Jiao et al. (2016) The Three- grasslands within the past 10 Andes River Source years Temperature of Grassland Xu et al. (2019) Region of the central Asian arid protection Tibetan Widespread vegetation regions Donohue et al. (2013) Plateau, greening in the Loess Plateau experienced a Removal of mega- Medium Skowno et al. (2017); Stevens China region; soil moisture sharp increase herbivores, fire High et al. (2017); Venter et al. Semi-arid declining widely, and deficit since 1997 and suppression, (2018); Garc韆 Criado et al. Australian in forests and orchards. The has been in a state 2020) arid lands runoff of the Yellow River is of high variability African declining since then savanna The runoff increases, the Significant total water storage and warming, slight groundwater increasing. NPP increase in increase precipitation. Widespread greening The precipitation increasing and Doubling of tree cover from evapotranspiratio 1940 � 2010 in South Africa n (ET) slight changing land use), and 20% decreasing increase in spread of woody Elevated CO2 Warming, elevated CO2, altered rainfall regimes Do Not Cite, Quote or Distribute CCP3-29 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report areas into previously open changed herbivore areas in the last 20 years regime African Widespread increase in tree Warming, Fire suppression High Venter et al. (2018) savanna cover across Africa with only changing rainfall, 3 countries across continent mention of CO2 Blaum et al. (2007); Blaum et experiencing a net decline in al. (2009); Sirami and tree cover Monadjem (2012); Gray and Bond (2013); P閞on and African Biodiversity responses to Woody Medium Altwegg (2015); Smit and savanna changes in vegetation encroachment Prins (2015) structure (woody Low African semi- encroachment) causing Woody High Gray and Bond (2013) arid regions declines in functional groups encroachment (savanna) that are open area specialists. Fire suppression High Bradley et al. (2016); Hufft North Records in birds, rodents, Increase in and altered and Zelikova (2016); American termites, mammals, insects. temperature and grazing/browsing Chambers (2018) drylands � Reduced tourism experience favourable regimes Caracciolo et al. (2016); sagebrush due to woody encroachment climates Archer et al. (2017) steppes Increasing Sagebrush steppes are being temperature, Chambers et al. (2014); Mote invaded by non-native elevated CO2 and High et al. (2018) grasses changing rainfall Shrub a) Warming and encroachment,(Prosopis associated decline glandulosa, Juniper ashei in snowpack and Juniper pinchotti) is b) Less occurring in the semi-arid precipitation grasslands of the southern falling as snow great plains at a rate of ~8% and an increase in per decade the rain fraction in Woody encroachment in winter. sagebrush steppes (cold deserts) (Juniper occidentalis) at a rate of 2% per decade Do Not Cite, Quote or Distribute CCP3-30 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report Central Desertification (as decreases decreased rainfall Land use change Medium Becerril-Pina et al. (2015); Mexico in vegetation indexes). + and intensification Noyola-Medrano and increase in Mart韓ez-S韆s (2017) Chinese Widespread greening trend of temperature Ecological Medium Piao et al. (2015) drylands vegetation in China over the Warming, CO2 protection last three decades; regional increase. Baccini et al. (2017); Aleman difference 1) Rising et al. (2018) atmospheric CO2 Stevens et al. (2017); Rosan et Dry sub-humid concentration and al. (2019) nitrogen Passos et al. (2018); Rosan et African Forest expansion into mesic deposition are Fire suppression Medium mesic savannas identified as the al. (2019) savannas most likely causes Donohue et al. (2013); Stevens of the greening et al. (2017); Bernardino et al. South 8% rate of woody cover trend in China, Fire exclusion High (2020) explaining 85% American increase and 41% of the average growing- cerrado season LAI trend. 2) Negative South Expansion of forest into impacts of climate Fire exclusion High American cerrado change in north cerrado China and Inner Mongolia and the positive impact in the Qinghai- Xizang plateau Increases rainfall, elevated CO2 Elevated Co2 Elevated CO2 Australian 2% rate of woody cover High savannas increase and greening of drylands Do Not Cite, Quote or Distribute CCP3-31 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report 1 Table CCP3.2: Synthesis of adaptation measures and responses to risks in deserts and semi-arid areas. Appropriateness 2 of measures is context dependent and some adaptations will be incremental or even maladaptive in some dryland 3 contexts while being transformational in other locations. Challenge Adaptation Measures and Responses References Soil erosion Rainwater harvesting and soil conservation, grass reseeding, Eldridge and agroforestry. Beecham (2018) Overgrazing Use of different breeds of grazing animals, altered livestock Kattumuri et al. rotation systems, use of new crop varieties, development of (2015), Shikuku et al. management strategies that reduce the risk of wildfire. (2017) Modification of production and management systems that involve diversification of livestock animals and crops, integration of livestock systems with forestry and crop production, and changing the timing and locations of farm operations. Improved breeds and feeding strategies and adoption of improved breeds for households without cows (both economic & environmental gain). Clearing of natural Carbon sequestration through decreasing vegetation clearing Kattumuri et al. vegetation rates, reversal through revegetation, targeting for higher- (2017), Mbow et al. yielding crops with better climate change adapted varieties, and (2014), Suckall et al. improvement of land and water management (2014) Agroforestry role in addressing various on-farm adaptation needs besides fulfilling many roles in AFOLU-related mitigation pathways (assets and income from carbon, wood energy, improved soil fertility and enhancement of local climate conditions; it provides ecosystem services and reduces human impacts on natural forests). Implementation of co-benefits strategies including provision of incentives across multiple scales and time frames, fostering multidimensional communication networks and promoting long-term integrated impact assessment. Invasive species and Achievement of triple-wins in SSA through provision of Mirzabaev et al. woody encroachment (2019); Davies and development benefits by making payments for forest services to Nori (2008); Stafford Droughts et al. (2017); smallholder farmers, mitigation benefits by increasing carbon Cromsigt et al. (2018); Ding and storage, and adaptation benefits by creating opportunities for Eldridge (2019) livelihood diversification. Morton and Barton Cliamte change is projected to facilitate the spread of invasive (2002); Abebe et al. species. Invaive species can have profound impacts on dryland (2008); Alary et al. ecosystems functioning leading to the loss of biodiversity. (2014); Catley et al. Biomass harvesting and selective clearing; utilising intense fires (2014); Mohamed et to manage encroachment, combined browsing and fire al. (2016) management. Rewilding in open ecosystems and reintroduction of mega-herbivores (e.g. in parts of Africa) to counter negative impact of woody encroachment. Chemical removal of undesirable encroached woody species Pro-active drought risk mitigation vs reactive crisis management approaches. Promoting collective action in livestock management, optimizing livestock policies and feed subsidies. interventions in livestock markets during drought onset. Expanding sustainable irrigation and shifting to drought- resistant crops and crop varieties. Environmentally sustainable sea water desalination. Promoting behavioural changes for more efficient residential water use. Moving away from water- intensive agricultural practices in arid areas. Harvesting rainwater by local communities; empowering women and engagement in local climate adaptation planning, community Do Not Cite, Quote or Distribute CCP3-32 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report Grassland and savanna based early warning systems, IRWM, water governance For review see degradation benchmarking, and exploration of palaeo channels as freshwater Buisson et al. (2019) sources using remote sensing Rangeland degradation Prescribed fire and tree cutting, invasive plant removal, grazing (decreasing fodder management, reintroduction of grasses and forbs, restoration of quality or yield, invasion soil disturbance. by fodder poor value species/refusals) Promote herd local and regional mobility during the growing De Vries and Djit鑩e season to avoid intense grazing pressure on growing annual (1982); Hiernaux et herbaceous vegetation of rangelands near settlements, water al. (1994); Hiernaux points, market. Moderate grazing facilitates grass tillering and and Le Hou閞ou herbaceous flora diversity. Ecological restoration of grazing (2006); Reed et al. ecosystems by sowing a mixture of zone-typical dominant (2015) species and life forms of plants on severely degraded land. Clearance of invasives. Ecological restoration of arid ecosystems by sowing a mixture of zone-typical dominant species and life forms of fodder plants with partial (ribbon) treatment of pasture lands. Ecological restoration of secondary salted irrigated soils using halophytes. Poor livestock Promote seasonal-regional herd mobility to optimise the use of Turner (1993); productivity complementary fodder resources (rangelands, browses, crop Schlecht et al. (2004); (reproduction/dairy/meat) residues). Implies institutionalized communal access, Fern醤dez-Rivera et in relation with poor community agreements and infrastructures (water points, al. (2005); Bonnet seasonal nutrition livestock path, grazing reserves, access to education, health and Herault (2011); care, markets for transhumant population). Cross state boundary Hiernaux et al. (2016) mobility implies international agreements such as promoted by N'djamena meeting (Declaration 2013) Promote strategic supplementation of reproductive and young Many trials in animals by the end of dry and early wet season. research stations and Secondary effect on excretion quantity/ quality to manure on farm: for example croplands. Sangar� et al. (2002a); Sangar� et al. (2002b); Osbahr et al. (2011); Sanogo (2011) Decrease trend in Rotational corralling of livestock in field during the dry season Pieri (1989); Breman cropland soil fertility (and on cleared fallow the following year in the wet season) to et al. (2001); Gandah ensure maximum retrieval of organic matter and nutrients from et al. (2003); Manlay faeces and urine deposited. Application or mineral N and P et al. (2004); fertilisers as placed (per poquet) microdoses (50-80 kg/ha) to Abdoulaye and intensify staple crop production. Impact on soil fertility, rain use Sanders (2005); Reij efficiency, vegetation cover, organic matter production and et al. (2005); recycling. Legume association with cereals (millet-cowpea; Akponikpe (2008); Sorghum-groundnut). Adapting cultivars and cropping Bagayoko et al. techniques (calendar, fertilisation) (2011); Bationo et al. (2011); Hiernaux et al. (2009b); Sendzimir et al. (2011); Turner and Hiernaux (2015); Weston et al. (2015; Reij and Garrity (2016) Salinisation and Indigenous and scientific adaptive practices to cope with Sengupta (2002); groundwater depletion salinity. Farmers in waterlogged saline areas harness sub- Buechler and Mekala surface drainage, salt tolerant crop varieties, land-shaping (2005); Wassmann et techniques and agroforestry to adapt to salinity and al. (2009); Singh waterlogging risks. Locally adapted crops and landraces, and the (2010); Jnandabhiram traditional tree- and animal-based means to sustain livelihoods and Sailen Prasad Do Not Cite, Quote or Distribute CCP3-33 Total pages: 53 FINAL DRAFT Cross-Chapter Paper 3 IPCC WGII Sixth Assessment Report Sand and dust storms in face of salinisation. Climate change is projected to increase (2012); Manga et al. 1 the salinization of groundwaters. Current unsustainable use of (2015); Sharma and 2 groundwaters is already leading to their depletion in some Singh (2015); Gupta dryland areas. and Dagar (2016); Nikam et al. (2016); Bundela et al. (2017); Sharma and Singh (2017); Patel et al. (2020); Singh et al. (2020b); Sharma, (2016); Mirzabaev et al. (2019) Use of live windbreaks or shelterbelts, protection of the loose Ahmed et al. (2016); soil particles through the use of crop residues or plastic sheets or Al-Hemoud et al. chemical adhesives, increasing the cohesion of soil particles by (2017); Sivakumar mechanical tillage operations or soil mulching. (2005); Hiernaux et al. (2009a); Hiernaux Use of perennial plant species that have the ability to trap et al. (2016); Pierre et sediments (sand and fallen dust) and form sandy mound around al. (2018); Lamers et it, such as Haloxylon salicornicum, Cyperus conglomerates, al. (1995); Michels et Lycium shawii, and Nitraria retusa. In Sahel: promote al. (1998); Bielders et herbaceous (not woody plants) to trap sand annuals such as al. (2004),UNEP Colocynthis vulgaris, Chrozophora senegalensis, Farsetia (2016); UNEP (1992) ramosissima, perennials such as Cyperus conglomeratus, Leptadenia hastate. In Sahel: leaving at least part of the crop residues (stalks) laid down on the soil during the dry season (100kg dry matter per hectare has already significant effect on wind erosion, many trials on Millet in Niger). Trampling by grazing livestock improves the partial burying of the residues. Improve monitoring, prediction and early warning. Monitoring, prediction and early warning to mobilize emergency responses for human systems & prioritize long-term sustainable land management measures. Establishment of a Global Dust-Health Early Warning System (building on the SDS-WAS initiative). Multi-sectoral preparedness and response including public health, hospital services, air and ground transportation and communication services Do Not Cite, Quote or Distribute CCP3-34 Total pages: 53