Green Horizons:Charting Trkiyes Sustainable JourneyBCG Report on Sustainability 2024 May 2024By Ayka.
2024-06-06
20页
5星级
Hydrogen Capability NetworkSecuring Liquid Hydrogen Capability in the UKMay 2024FlyZeros LH2-powered.
2024-06-05
14页
5星级
张小可GEORG PETERSENERKIN MADUMAROVAIMAN UTEYEVABAKHROM GAFORZODAGULNARA MERGENOVAAZIZ KARIMOVWater sec.
2024-06-05
97页
5星级
IUCN Urban ToolboxINTERNATIONAL UNION FOR CONSERVATION OF NATUREA catalogue of knowledge products on.
2024-06-05
32页
5星级
PORTUGALS ENERGY AND GREEN TRANSITION Potential opportunities and partnerships for Swedish companie.
2024-06-03
15页
5星级
Circular Industry Solutions for a Global Plastics TreatyI N S I G H T R E P O R TM A Y 2 0 2 4In col.
2024-06-03
34页
5星级
L E V E L I Z E D C O S T O F E N E R G Y+W I T H S U P P O R T F R O MJ u n e 2 0 2 4Table of Conte.
2024-06-01
48页
5星级
Thematic Report on Rangelands and PastoralistsThematic Report on Rangelands and PastoralistsThe Global Land Outlook Thematic Report on Rangelands and Pastoralists was produced by a team led by the author and secretariat of the United Nations Convention to Combat Desertification(UNCCD),in collaboration with supporting and contributing partners,and in consultation with key stakeholders and experts.It was made possible through the generous financial support of the European Union.Author:Pedro Maria Herrera CalvoEditors:Shannon Mouillesseaux and Sasha AlexanderGraphic Designer:Javier AcebalContributors:Nicholas Euan Sharpe(FAO,Angola),Phemo Karen Kgomotso(UNDP,Mongolia),Karim Musalem(WWF,Paraguay),Jara Febrer Santos(GOB Menorca,Spain),Mahoussi Simone Assocle(IFAD,Tanzania,Kyrgyzstan and Tajikistan),Alexis Bonogofsky(WWF,United States),Nachilala Nkombo(WWF,Zambia),Patrick Nino Oloumane,Carlo Prvil and Wael El Zerey(University of Quebec-UQAT,Canada),Nasser Eddin Obaid(ACSAD,Arab countries),Almut Therburg(CONICET,Argentina),Mara Cristina Camardelli(National University of Salta,Argentina),Bibiana Vil and Yanina Arzamendia(VICAM/CONICET,Argentina),David Cobon(Centre for Applied Climate Sciences,Australia),Edina Scherzer(Raumberg-Gumpenstein R&D,Austria),Ana Carolina Crisostomo(WWF,Brazil),Paola Agostini(The World Bank,Central Asia),Hindou Oumarou Ibrahim(AFPAT,Chad),Yongjun Li(National Forestry and Grassland Administration,China),Huxuan Dai(Shan Shui Conservation Center,China),Moustafa S.El Hakeem(Desert Research Center,Egypt),Tsegazeab Embaye Tedla(Ministry of Agriculture,Eritrea),PENHA(Ministry of Agriculture,Ethiopia),Nicola di Niro(CRAMM-Geaco/Fondazione Popoli e Territori,Europe),Dragan Angelovski(FAO,Georgia),Anshul Ojha(Desert Resource Centre,India),Vivekanandan Perumal(Sustainable Agriculture and Environmental Voluntary Action,India),Rishi Sharma(WWF,India),Pier Paolo Roggero(Desertification Research Centre,University of Sassari,Italy),Marco Bindi,Camilla Dibari,and Giovanni Argenti(University of Florence,Italy),Mira Haddad(ICARDA,Jordan),Isaac Kofi Bimpong(IAEA,Austria),Peter Ken Otieno(RECONCILE,Kenya),Fiona Flintan(ILRI,Kenya,Tanzania,Ethiopia and IGAD Region),Zhyrgalbek Kozhomberdiev and Talantbek Toktosunov(CAMP Alatoo,Kyrgyzstan),Elvira Maratova(ILC Ecosystem Restoration Platform,Kyrgyzstan and Thailand),Gregorio Velasco Gil(FAO,Mali,Mauritania,Niger and Senegal),Elisabeth Huber-Sannwald and Natalia Martnez Tagea(Instituto Potosino de Investigacin Cientfica y Tecnolgica,A.C.,Mexico),Miguel Angel Cruz Nieto(Organizacin Vida Silvestre,Mexico),Mario Rodrigo Chvez(Comisin Nacional de reas Naturales Protegidas,Mxico),Ykhanbai Hijaba(JASIL,Mongolia and Central Asia),Stefan Graf(Bern University of Applied Sciences,BFH-HAFL),Nicole Harari and Rima Mekdaschi(WOCAT,Switzerland),Sanusi Abubakar(L-PRES,Nigeria),Mohammad Zaman(IAEA,Pakistan),Duarte Marques(Aguiarfloresta,Portugal),Quasim Al-Janabi(Ministry of Environment,Iraq),Roxana Triboi(Ion Mincu University of Architecture and Urbanism,Romania),Kulik Konstantin Nikolaevich(Federal Scientific Center of Agroecology,Complex Melioration and Protective Afforestation,Russia),Orou Djega Imorou(CILSS,Sahel),Serena Ferrari(CIRAD,Senegal),Andiswa Finca(Agricultural Research Council,South Africa),Jabier Ruiz-Mirazo,Mara Turio,Yolanda Sampedro and Mireia Llorente(Entretantos Foundation,Spain),Sharini Somasiri and Sanjaya Fernando(Rajarata University,Sri Lanka),Lucia Gerbaldo(WeCAN/FAO),Namayani Rapey Edward(Pastoral Womens Council,Tanzania),Mounir Louhaichi(ICARDA,Tunisia),Pius Loupa(COPACSO,Uganda).Special thanks are also extended to the many individuals and organisations that helped to mobilise and gather the case studies used in this report.Reviewers:Cathrine Mutambirwa(UNCCD),Olga Andreeva(UNCCD),Suyu Liu(UNCCD),Soma Chakrabarti(UNCCD),Birguy Lamizana(UNCCD),Melissa Ho(WWF),Martha Kauffman(WWF),Chris Magero(IUCN),Bora Masumbuko(IUCN),Jabier Ruiz-Mirazo(Entretantos),Jonathan Davies(Consultant),Fidaa Haddad(FAO),Aurelie Bres(FAO),Rima Mekdaschi(WOCAT),Lindsey Sloat(WRI),Mulubrhan Gebremikael(WRI),Ann Waters-Bayer(CELEP),Serena Ferrari(CIRAD),Igshaan Samuels(IYRP),Nigel Dudley(Equilibrium Research),Sobirjon Umarov(Uzbekistan),Raafat Misak(Egypt),Sarab Wajaan Ajeel(Iraq),Wang Shiqin(China),Julie Suh(CSIRO),Yriz Silva and Angelo Paulo Sales dos Santos(Brazil),Jamal Annagylyjova and Tristan Tyrrell(CBD),Cludia Vieira Lisboa(UN Tourism),Mara Degania Medina Vidal(Spain),DIGMA(Argentina),Baitshepi Edith Babusi Hill(Botswana),Sultan Veysov(Turkmenistan),Rysbek Apasov(Kyrgyzstan),Assel Berentayeva(Kazakhstan),Tayebeh Mesbahzadeh(University of Tehran),Kamal Sadik Ahmed(Somalia),Maria Fernandez-Gimenez(Colorado State University),Marie Aude Even(IFAD),David Briske(Texas A&M University),Barry Irving(University of Alberta);UNCCD Science and Policy Interface(SPI):Dolors Armenteras(University of Barcelona),Nichole N.Barger(University of Colorado,Boulder),Vera Boerger(FAO),Helene Gichenje(Commonwealth Secretariat),Elisabeth F.Huber-Sannwald(Instituto Potosino de Investigacin Cientfica y Tecnolgica),Tungalag Ulambayar(Zoological Society of London),Anah Ocampo Melgar(University of Chile),and Sara Alibakhshi(University of Helsinki).Citation:UNCCD.2024.Global Land Outlook Thematic Report on Rangelands and Pastoralism.United Nations Convention to Combat Desertification,Bonn.United Nations Convention to Combat Desertification(UNCCD)Platz der Vereinten Nationen 1 D-53113 Bonn,Germany www.unccd.intCover photo:ILRI/Stevie MannISBN on-line:978-92-95118-82-9 ISBN print:978-92-95118-83-6This publication is available for download at:https:/www.unccd.int/resources/global-land-outlook/overview 2024 UNCCD.All rights reserved.Disclaimer:The designations employed and the presentation of material in this information product do not imply the expression of any opinion whatsoever on the part of the United Nations Convention to Combat Desertification(UNCCD)concerning the legal or development status of any country,territory,city or area or of its authorities,or concerning the delimitation of its frontiers or boundaries.The mention of specific companies or products of manufacturers,whether or not these have been patented,does not imply that these have been endorsed or recommended by the UNCCD in preference to others of a similar nature that are not mentioned.The views expressed in this information product are those of the authors or contributors and do not necessarily reflect the views or policies of the UNCCD.This publication was funded by the European Union.Its contents are the sole responsibility of the UNCCD and do not necessarily reflect the views of the European Union.AcknowledgementsContentsAcknowledgements iiPreface ivForeword vExecutive Summary vi1.Overview 81.1 Aim and scope 81.2 Structure and contents 91.3 New approaches 91.4 Definitions and explanatory notes 102.Rangeland health and degradation 132.1 Rangeland characteristics 132.2 Rangeland degradation 142.3 Monitoring rangeland health 162.4 Conceptual framework for rangelands and pastoralism 173.Learning from the past,planning for the future 193.1 A historical perspective 193.2 Learning from the past 213.3 Project formulation 213.4 Rangeland interventions 214.Regional analysis and case studies 254.1 East Africa 254.2 West Africa 354.3 Middle East and North Africa 394.4 Central Asia and Mongolia 444.5 Europe 504.6 South Asia 594.7 China and Southeast Asia 624.8 South America 654.9 North America 714.10 Southern Africa and Australia 765.Global support for rangelands and pastoralism 785.1 Global and regional frameworks 785.2 Land rights and tenure security 805.3 Grassroots organisations and pastoralist voices 805.4 Cultural values and heritage 815.5 A gender responsive lens 825.6 Nature conservation 825.7 Co-creation of knowledge 825.8 Resource mobilisation 855.9 Inclusive and responsible governance 865.10 Global recognition of a transversal approach 866.Conclusion 88Endnotes 92ivGLOBAL LAND OUTLOOK Thematic Report on Rangelands and PastoralistsPrefacePastures,meadows,and rangelands are more often perceived as resource and land frontiers that have yet to be exploited and are of little value until they are transformed by human hands.The term development is often taken to mean human action,agricultural development,destruction of natural habitats,draining of wetlands,or urban development.Rangelands are often referred to as arable land,a sign that planners see them as better developed when transformed than when left in their natural state.When we destroy a forested area,we talk about deforestation.Seeing a 100-year-old tree fall rightly generates a great deal of emotion.On the other hand,the conversion of rangelands-even those that are several hundred years old-is done in silence and generates little public reaction.Rangelands are as little appreciated as their users are integrated into our societies.Marginalised,pastoralists and livestock breeders find it hard to influence development policies.They are voiceless,powerless,and generally,a minority in the political and administrative machinery.Although estimated to number half a billion souls,they are sometimes classified as indigenous peoples or as societal outsiders.Rangelands are extensive ecosystems that provide biodiversity and support rural livelihoods,yet they are threatened by land degradation,climate change,and land conversion.Their importance cannot be overstated in our collective pursuit of sustainable development and planetary stability;however,they have long been underappreciated in global environmental discourse.Therefore,I am delighted to introduce the Global Land Outlook thematic report on rangelands and pastoralists.It reflects our commitment to reduce and reverse desertification and land degradation,and build drought resilience through sustainable land management that can improve the well-being of millions of people worldwide.As part of the UNCCDs ongoing efforts to support Sustainable Development Goal(SDG)15 and Land Degradation Neutrality(LDN),this report aims to set a solid foundation for sustainable management and restoration practices in close collaboration with the pastoralists and communities that reside,manage,and depend on rangelands.It showcases the importance of respecting pastoral heritage,cultures,and traditions,and highlights their role in protecting and restoring rangeland resources for current and future generations.By recognising the intrinsic value of rangelands and the irreplaceable role of pastoralists in preserving them,we are acknowledging the interconnection between ecosystem and human health and well-being.Responsible land governance,smart and targeted investments supported by policies and measures that value and protect rangelands and their communities are vital.Healthy,well-managed rangelands help combat desertification and climate change while delivering food,water,shelter,and economic opportunities.Sustainable rangeland management practices enhance resilience and the capacity of communities and ecosystems to withstand the pressures and shocks of global change.As we witness the alarming decline of species worldwide,the preservation of rangeland biodiversity is integral to our broader nature conservation efforts.In anticipation of the International Year of Rangelands and Pastoralists(IYRP)in 2026,this report serves as a catalyst for global awareness and action.It analyses numerous case studies and good practices from around the world,drawing on the experience and lessons learned,and advocates for a new paradigm to inspire governments,donors,and other stakeholders to prioritise rangeland health in cooperation with local communities.Through these collaborative efforts and a commitment to shared responsibility,we can preserve these rich cultural landscapes for the benefit of people,nature,and the climate.Ibrahim Thiaw,UNCCD Executive SecretaryvGLOBAL LAND OUTLOOK Thematic Report on Rangelands and PastoralistsForewordIn 2022,the United Nations declared 2026 the International Year of Rangelands and Pastoralists(IYRP)and named the Food and Agriculture Organization(FAO)as the lead UN agency for its implementation.The IYRP aims to raise awareness and advocate for healthy rangelands and sustainable pastoralism,and to promote capacity building and responsible investment in favour of the pastoral livestock sector.The idea to commemorate rangelands and pastoralists was spearheaded by Mongolia in collaboration with the International Support Group(ISG).Thanks to Mongolias vision,we have an opportunity to redefine the narrative surrounding rangelands and pastoralist communities,and to collectively shape a sustainable future for our planet.As the Co-Chair of the ISG for IYRP2026,I view the GLO Thematic Report on Rangelands and Pastoralists as among the first steps towards these aims.By shedding light on the challenges we face in preserving and managing rangelands globally,and recommending ways to help alleviate and address them before it is too late,this report offers policymakers,practitioners,and communities alike a pathway to support the well-being of rangelands and pastoralist communities and cultivate a sustainable future.Pastoralism has a much lower overall environmental footprint than other forms of livestock production,as it works with nature not against it.But its share of the global market for meat and milk products is far outstripped by intensively farmed operations.Efforts are underway to reduce the environmental footprint of intensive livestock farms,but unfortunately all too often the pastoralist is also thrown into the same policy basket as the intensive farmer.The IYRP aims to unpack this basket to show that pastoralists and their rangelands are different and can be even more sustainable with the right approaches to dedicated and targeted policies and investments.The IYRP aims to raise awareness as well as encourage more knowledge generation,building on the traditional and local knowledge of pastoralists.Already well in advance of 2026,the ISG,consisting of over 300 organisations and associations,has created new scientific evidence and global maps,and established platforms for cooperation.It recently released a Science Review of Land Degradation Neutrality that complements and strengthens the findings and recommendations of the GLO report and offers positive policy options at national and international levels that could have immediate impact.Mind sets are starting to change.We must translate our shared aspirations into concrete actions stopping indiscriminate conversion of rangelands into unsuitable land uses,advocating for policies that support sustainable land management,investing in research that enhances our understanding of rangelands and pastoralism,empowering pastoralist communities to preserve their sustainable practices while also gaining tools to thrive in a changing world,and supporting all stakeholders,especially pastoralists,to implement measures that effectively thwart further degradation and preserve our land,our communities,and our cultures.May this GLO thematic report propel rangelands and pastoralists to the forefront of global consciousness and,in conjunction with the upcoming IYRP2026,serve as a catalyst for lasting change.Maryam Niamir-Fuller,Co-Chair of the International Support Group for IYRP 2026viGLOBAL LAND OUTLOOK Thematic Report on Rangelands and PastoralistsExecutive SummaryKey MessagesThe conversion and loss of rangelands is done in silence and attracts little public attention.Often marginalised or considered outsiders,many pastoralist and rangeland communities are unable to influence the policies and programmes that directly impact their food security,livelihoods,and cultural identity.They are voiceless and powerless and represent a small minority in the political and administrative machinery that governs development and investment decisions in the rangelands.Pastoralist livelihoods and cultures around the world are under threat from shortsighted policies,weak governance,and economic incentives that undermine their production systems.Pastoralists are broadly defined as extensive livestock farmers,herders,and ranchers whether indigenous or not whose way of life is closely linked to the health and productivity of rangelands.Up to 500 million people across the world practise this form of animal husbandry.Yet,in many regions,they have little recourse to address the conversion,fragmentation,and degradation of rangelands.Rangelands operate as complex social-ecological systems with critical values,processes,goods,and services.They are diverse,multifunctional,and encompass a wide variety of ecosystems(e.g.,drylands,grasslands,savannahs)that have co-evolved with human communities.Covering over 50 per cent of the Earths land surface,rangelands are comprised of grasses,herbaceous plants,and shrubs that are grazed by livestock and/or wildlife.In addition to meat,dairy,fibre,and other animal products,rangelands and their biodiversity underpin critical ecosystem services from local to global scales(e.g.,nutrient/water cycling,carbon sequestration,animal/human health).Despite the extraordinary diversity and intrinsic value of rangelands and pastoralist systems,they rarely feature in global policy discussions or national development priorities.Rangelands provide important environmental,social,and economic benefits that are often taken for granted,in part due to the lack of understanding of their extent,condition,use,value,and diversity.While there are many threats to rangeland health,one is the imbalance in the supply and demand for animal forage which leads to overgrazing,invasive species,and bush encroachment as well as the increased risk of drought and wildfires.Pastoralism and extensive livestock production systems are deeply rooted in the rangelands and often the most effective means to protect,sustainably manage,and restore rangelands.Appreciating that food and fibre production is the most common economic use of rangelands,sustainable grazing is a proven,cost-effective management approach to enhancing their health,productivity,and resilience.Traditional and regenerative grazing practices can often mimic natural processes that build soil organic matter,increase water retention,sequester carbon,conserve biodiversity,and reduce the spread of invasive species.Greater political attention and informed investments are urgently needed to safeguard and improve the health and productivity of the rangelands and their inhabitants.This report offers insights and guidance on the policy and operational frameworks and other enabling factors for attracting greater attention and investments in sustainable rangeland management projects and programmes.Illustrated with case studies and good practices from around the world,it highlights the critical role of pastoralist communities in the planning and implementation of rangeland initiatives that deliver benefits in all three dimensions of sustainable development.Key ActionsSustainability Framework:National and sub-national authorities can design and implement legal and operational frameworks that align rangeland management and pastoralist livelihoods with the Sustainable Development Goals(SDGs),fully considering the environmental,social,and economic dimensions,and support efforts to:Endorse and enact national laws and regulations that are aligned with international treaties,obligations,and commitments that support the diversity,resilience,and multiple values of extensive livestock systems and rangeland ecosystem services.Recognise and enforce legitimate land rights,respect the unique circumstances and needs of rangeland communities(e.g.,mobility,transhumance,communal governance),and nurture their participatory role in the conservation,sustainable management,and restoration of rangelands.Facilitate multistakeholder platforms and networks for research and learning,knowledge co-creation and exchange,and monitoring and evaluation and to create accessible databases and repositories that collect and disseminate information on rangelands and pastoralist systems.viiGLOBAL LAND OUTLOOK Thematic Report on Rangelands and PastoralistsEnvironmental Dimension:National and sub-national authorities can take measures to support the ecological integrity,connectivity,and functioning of rangelands through conservation,sustainable use,and restoration activities that safeguard and enhance the multiple benefits they provide to societies and economies,and support efforts to:Reduce and avoid rangeland conversion resulting from inappropriate land uses(e.g.,crop monocultures,tree plantations,afforestation)that diminish the diversity and multifunctionality of rangelands,especially on indigenous,pastoral,and communal lands.Adopt and support pastoralism-based strategies that directly address the natural and human-induced drivers of rangeland degradation,such as biodiversity loss,climate change,overgrazing,soil erosion,invasive species,drought,and wildfires.Design and implement nature conservation measures that reduce and halt biodiversity loss(above and below ground)by harnessing synergies with pastoralist practices and extensive livestock production systems that boost rangeland health,productivity,and resilience.Integrate climate change mitigation and adaptation measures into sustainable rangeland management plans and programmes(or vice versa)to increase carbon sequestration and storage while enhancing the adaptive capacity of rangelands and their communities.Social Dimension:National and sub-national authorities can take measures to build social capital in rangeland communities through participatory governance and adaptive management approaches that promote gender equality,social cohesion,and trusted institutions to foster collective action,and support efforts to:Provide capacity building,skills training,and technical support to build the human and social capital needed for collective action that safeguards rangeland health and livelihoods,with particular attention to mobility,gender-responsiveness,and social inclusion.Support rangeland and pastoralist associations and networks that celebrate and defend their cultural heritage and values,increase connectivity and social services,and ensure the provision of human resources and expertise needed for responsible and inclusive rangeland governance.Facilitate women-led,women-driven,and women-only initiatives,groups,and institutions(along with mixed gender ones)to ensure that womens voices are heard and respected and to activate their contribution to all dimensions of sustainable development in the rangelands.Establish trusted institutions and mechanisms to manage wildlife and resource conflicts,resolve territorial and land tenure disputes,reduce inequalities in access and benefit sharing,and negotiate trade-offs and leverage synergies for the benefit of rangelands,their communities,and society-at-large.Economic Dimension:National and sub-national authorities can take measures to support the economic viability of extensive livestock production and the livelihoods they support through flexible long-term investments and incentives,including context-appropriate strategies and programmes that link markets and value chains to sustainable rangeland production systems,and support efforts to:Create innovative economic and financial mechanisms that are accessible to rangeland stakeholders,incentivise good management practices,provide decent work,stimulate market participation,and increase investments in sustainable pastoralism from public and private sources while avoiding adverse consequences for rangeland communities.Develop market and value chain strategies and action plans that support economic livelihoods and income diversification and expand innovative and profitable opportunities for rangeland communities engaged in extensive livestock production.Promote adaptive investment and risk management tools,such as livestock and drought insurance,resource pooling and sharing,and community credit schemes,to better manage risks and uncertainties in a creative but economically sound manner.Conduct economic valuations of rangeland ecosystem services to better understand their contribution to people,nature,and climate,to help inform rangeland policies,planning and programmes,and to attract donor funds,private sector investments,and public sector allocations for sustainable rangeland management and restoration.8GLOBAL LAND OUTLOOK Thematic Report on Rangelands and PastoralistsRangelands play a central role in achieving Land Degradation Neutrality(LDN)and contributing to local,national,and global sustainability agendas.Rangelands operate as complex social-ecological systems with critical values,processes,goods,and services.1 Rangelands and their host ecosystems(e.g.,drylands,grasslands,savannahs)have co-evolved with human communities whose food security,livelihoods,and cultural identity directly depend on the resources and opportunities that they provide.2 The United Nations designated 2024 as the International Year of Camelids(e.g.,camels,llamas,alpacas,vicuas,guanacos),a way of life for millions of pastoralists in dryland and mountainous rangelands around the world.Subsequently,the United Nations declared 2026 the International Year of Rangelands and Pastoralists(IYRP)to raise awareness and promote increased investment in the sustainable management and restoration of rangelands,while recognising and supporting pastoralist communities and their significant contribution to sustainable development.3 The IYRP designation underscores the importance of healthy rangelands and sustainable pastoralism to achieve the Sustainable Development Goals(SDGs),specifically target 15.3 to halt desertification and reduce land degradation supported by national LDN commitments under the United Nations Convention to Combat Desertification(UNCCD).Healthy rangelands are also critical to fulfil the commitments and targets under the Convention on Biological Diversity(CBD)and the United Nations Framework Convention on Climate Change(UNFCCC).As part of the global effort to combat desertification,land degradation and drought,the UNCCDs Global Land Outlook Thematic Report on Rangelands and Pastoralists(“the report”)puts forward an integrated conceptual framework that is aligned with the LDN approach4 and offers flexible pathways to improve rangeland conservation,management,and restoration outcomes.The case studies presented in the report point to the need for greater policy support,increased investment,and partnerships at all levels and across all relevant sectors.The report focuses on the relationship between rangelands and their human communities,most notably pastoralists,but also other land users that manage rangeland resources sustainably under a purposeful and regenerative management approach.The underlying premise is that this approach can be scaled up and out to protect rangelands and their functions,5 as well as to accelerate progress towards many SDG targets,6 Global Biodiversity Framework(GBF),7 United Nations Decade on Ecosystem Restoration 20212030,8 and the Paris Agreement.1.1 Aim and scopeThe report explores the complex environmental,social,and economic dimensions that link rangelands and local communities.It describes the important role and untapped potential of pastoralism and extensive livestock management systems to contribute to a just transition,climate resilience,and more equitable rural development,recognising that many of the challenges confronting rangelands originate beyond local communities and are not under their control.Drawing on case studies submitted from around the world,the report offers new perspectives on how pastoralism can contribute to more effective rangeland governance and stewardship and examines the potential for replicability and scalability.It draws on a diversity of approaches(e.g.,territorial,ecosystem,cultural)and initiatives(e.g.,global,national,local),supported by policy,implementation,and investment frameworks,to conserve,sustainably manage,and restore rangelands.The report also reflects on lessons learned to improve the design,planning,implementation,and finance for future rangeland initiatives.The relationship between rangeland health and management practices is addressed with a Driver-Pressure-State-Impact-Response(DPSIR)perspective,analysing both positive and negative impacts as well as addressing synergies and trade-offs.It concludes that local,multi-actor,transdisciplinary,adaptive,and inclusive approaches can be effective in improving the health and productivity of rangelands and safeguarding the livelihoods and cultural values of their communities.1.Overview9GLOBAL LAND OUTLOOK Thematic Report on Rangelands and Pastoralists1.2 Structure and contentsThis first chapter provides an overview of the report,its theory of change,and key definitions and explanatory notes.The second chapter aims to characterise rangelands,pastoralism,and the challenge of environmental degradation by analysing the drivers and responses within an enhanced conceptual framework to guide strategies and actions.Drawing on case studies,scientific literature,and other knowledge sources,the third chapter offers a historical perspective and reflects on the lessons learned to improve the quality and performance of rangeland and pastoralist projects and programmes.The fourth chapter includes snapshots for 10 regions of the world which are illustrated with case studies at different scales.The fifth chapter describes existing initiatives that promote and support rangelands and pastoralists around the world.The sixth chapter includes conclusions and additional guidance to support policymakers and other stakeholders in designing and implementing policies,projects,and programmes that protect and enhance rangeland health.1.3 New approachesThe report encourages a rethink of the conceptual framework currently applied to combat desertification and degradation in rangelands through an increased focus on the management practices employed in pastoralist and extensive livestock systems.It draws attention to pathways for improved policies,planning,implementation,and monitoring,with guidance for policymakers and other stakeholders on how to improve rangeland health under a sustainability framework with its three integrated dimensions.The report hopes to catalyse action at different scales to optimise rangeland benefits through sustainable production systems and value chains.The strategic approaches presented in the report can help create the appropriate enabling environment,mobilise resources(through incentives and investments),and improve the quality and outcomes of interventions that target rangelands and their inhabitants(Figure 1).The report applies elements of adaptive management models to improve rangeland planning and interventions based on a systemic and iterative decision-making approach,meaningful stakeholder engagement,sustained finance,and long-term monitoring.This approach can be supported with transition scenarios that integrate strategic,tactical,operational,and monitoring protocols that account for trends and feedback loops.9 The report introduces a robust conceptual framework to help better integrate rangeland and pastoralist initiatives into the different levels and scales of decision making.Integrated land use planning and landscape management are relevant tools and most effective when they recognise the main features of pastoralism,such as mobility,multifunctionality,diversity,adaptability,resource pooling(reciprocity and exchange),and the non-exclusive use of different and often variable natural resources.FIGURE 1 Theory of change10 ManageConserveRestoreSustainability framework for rangelands:integrating environment,society and economy Healthy rangelands,improved livelihoods,sustainable production systems and value chains Legal,policy and institutional frameworksSecure tenure and land rights Improved multi-actor governanceRANGELANDSPASTORALISTSChallengesImpactsOutcomesInterventionsEnablingEnvironmentMeans ofImplementationConceptualFrameworkCASE STUDIESRangeland and pastoralist initiatives and projectsAssessment and technical supportSustainable Rangeland ManagementCHAPTER1/623/445Investments and financeLearning and capacity buildingParticipatory processesCo-of knowledgecreation10GLOBAL LAND OUTLOOK Thematic Report on Rangelands and Pastoralists1.4 Definitions and explanatory notesThe report focuses on land use and management practices in rangelands,acknowledging the diversity of their host ecosystems and biomes grazed and browsed by livestock and wildlife.This section introduces and defines key terms and concepts used in the report,some of which may engender differences in interpretation around the world,across disciplines,and among practitioners.11 Land use is defined as the purposes and activities(primarily grazing and browsing in rangelands)through which people interact with land in these grass-dominated terrestrial ecosystems.12 Land cover refers to the character of the elements located on the surface of the land,either biophysical(e.g.,vegetation,grasses,shrubs,trees)or artificial(e.g.,buildings,livestock shelters,energy infrastructure).Land conversion or transformation,referred to as land use change or land cover change,is a major global challenge resulting from socioeconomic transitions including agricultural expansion,urbanisation,and consumer demand,among other factors.13Land management is any process or activity by which humans allocate or transform land resources for specified uses and goals,such as to generate social,environmental,or economic benefits.14 Sustainable land management(SLM)implies the use of land resources to meet changing human needs while safeguarding their long-term health and productive potential,including the maintenance of their environmental functions.15 In the report,SLM in the rangelands is referred to as sustainable rangeland management(SRLM)which can be described as a knowledge-based process that integrates social,economic,and ecological principles into rangeland policies and practices.16 Explanatory Note:The report acknowledges that pastoralist activity always has human intelligence behind decision making and planning for the protection and use of available resources(whether it is a single herder deciding the daily itinerary or a community moving from winter to summer pastures).Accordingly,the report considers all pastoralist systems as land management systems.The decision to not allow grazing or restrict other land uses(whether temporarily or permanently)is also understood as a form of land management.Abandonment is considered a discontinuation of land management typically resulting from the loss of rangeland functions and services.17Integrated land use planning(ILUP)involves designing and implementing the most appropriate land use strategies and practices based on systematic assessments of social,economic,and environmental conditions.18 The purpose of ILUP is to map and assign a mosaic of compatible land use types for a given territory in a way that is socially just and desirable and economically viable,while safeguarding ecological functions and the provision of ecosystem services for current and future generations.ILUP is an important enabling factor for the efficient and effective implementation of SRLM and restoration activities.The capacity and flexibility of ILUP instruments can allow for the combination of sustainable pastoralism and other rangeland uses within a given landscape which can promote both diversification in pastoralist production systems and the use of adaptive land management practices19 to boost community and ecosystem resilience under rapidly changing conditions.Land degradation in the rangelands is defined as a deterioration in land condition(i.e.,reduced biological and economic productivity)typically caused by direct human interventions(e.g.,overgrazing,mining)or indirect drivers(e.g.,anthropogenic climate change,socioeconomic transitions).Land degradation can be expressed as the persistent or long-term reduction or loss of ecosystem goods and services,20 which reduce biological productivity,ecological integrity,and/or economic values.Land degradation in the rangelands is a serious concern that impacts both people and nature and contributes to climate change.21 Land degradation in arid,semi-arid,and dry sub-humid areas is known as desertification.Land degradation neutrality(LDN)is defined as“a state whereby the amount and quality of land resources necessary to support ecosystem functions and services and to enhance food security remain stable,or increase,within specified temporal and spatial scales and ecosystems”.22 LDN directly responds to SDG target 15.3 by seeking a balance between land degradation and restoration through continuous improvement in management practices,while considering trade-offs and synergies with other SDGs.The UNCCD endorsed LDN as a primary vehicle to drive the implementation of the convention and embraced LDN in the vision of its 2018-2020 Strategic Framework.23Ecosystem restoration is defined as the process of assisting the recovery of degraded,damaged,transformed,or destroyed ecosystems to reinstate their ecological processes,functions,and services.24 The United Nations is supporting the Decade on Ecosystem Restoration(2021-2030),along with GBF target 2,in an attempt to recover lost biodiversity habitat and ecosystem services,and to mitigate and adapt to climate change while enhancing food security and creating livelihood opportunities.25 The inherent synergies among these targets and commitments make rangelands an optimal ground for developing adaptive approaches that maximise the full suite of benefits for people,nature,and climate.26 Explanatory Note:While the focus of non-agricultural land restoration has been primarily on forests,the report recognises the need and potential to restore rangeland ecosystems,such as grasslands,savannahs,or shrublands.Interest in restoring these ecosystems is growing rapidly and has become a priority for the UN Decade on Ecosystem Restoration.27 The report applies the principles and 11GLOBAL LAND OUTLOOK Thematic Report on Rangelands and Pastoralistsstandards of ecosystem(or ecological)restoration,which strive to conserve or regenerate the full suite of rangeland functions and services.28 However,many afforestation projects in the rangelands have raised serious concerns and intense debate.29 The report strongly maintains that the transformation of rangelands into forests or tree plantations should be avoided unless scientifically justified by the historic,ecological,and socioeconomic characteristics of the targeted area.30 Differentiating between“rangelands”and“grasslands”can be controversial.Both terms are often used as synonyms,31 although their many nuances are subject to debate.The report defines rangelands as natural or semi-natural ecosystems grazed by livestock and/or wild animals.Their vegetative cover is comprised of grasses,forbs,bushes,and shrubs,and may include open forests and agroforestry systems.Rangelands are considered complex social-ecological systems32 whereby their natural resources provide a broad range of goods,services,and values that must be considered in baseline and functional assessments.33 Many rangelands are found in the drylands,which are characterised by water scarcity typically with an Aridity Index below 0.65.34 Other important rangelands include mountain and tundra biomes that host pastoralist systems with high-value cultural and natural heritage(e.g.,reindeer herding in the Arctic,domesticated camelids in the Andes).Grasslands are defined as ecosystems dominated by grasses or grass-like plants,35 although they can contain trees or other woody vegetation as in the case of shrublands,woody grasslands,open forests,or savannahs.36 Grasslands are ecosystems of remarkable biodiversity.37 In addition to natural grasslands determined by climate and soil types,secondary grasslands can arise as a consequence of land use change or other human activities.38 The extent and degree of ecological integrity and human intervention(e.g.,seeding,mowing,fertiliser use)influence grassland characteristics.Old-growth or ancient grasslands,encompassing rich,biodiverse grasslands,savannahs,and open woodlands,39 tend to maintain higher ecological values.40 At the other extreme,monospecific seeded grasslands indicate the transformation of vegetative cover and resemble cultivated land more than a natural ecosystem.Explanatory Note:The report utilises“grasslands”as an ecosystem concept,primarily defined by vegetation cover,while the term“rangelands”is employed as a land use and land management concept within the conceptual framework(Figure 6).Rangelands,considered by some as a cultural ecosystem,are primarily defined by their use for grazing(by livestock,semi-domesticated animals,or wildlife)or the gathering of feed,whether potential or actual.41 They often comprise a mosaic of land uses and ecosystems,such as grasslands,savannahs,shrublands,drylands,deserts,steppes,mountains,and open forests,as well as agroforestry and silvopastoral systems.42Grazing systems are livestock-based production systems that integrate grazing practices with the management of soil,water,and biodiversity resources within a specific socioeconomic context.43 Pastoralist systems are based on mobile grazing animals under nomadic,transhumant,or sedentary management systems.44 Pastoralism encompasses the extensive production of livestock,using pasture or browse as the main source of feed.45 This definition is expanded in the report to include any extensive rangeland production system that dynamically manages livestock and land resources to optimise economic,social,and environmental benefits.46 Beyond livestock production,pastoralism encompasses cultural identity,knowledge pools,traditional institutions,and landscape heritage that shape the way of life for these rangeland communities.47 Some common terms used to describe pastoral systems and their features around the world include transhumance,nomadism,and animal husbandry.48 CIAT/Juan Pablo Marin Garca12GLOBAL LAND OUTLOOK Thematic Report on Rangelands and PastoralistsExplanatory Note:The report employs“pastoralism”as a comprehensive term,encompassing the entire range of extensive livestock production systems in the rangelands,including those that use rangelands as part of agropastoral,silvopastoral,or agroforestry systems.Where pastoralism is used under a more restrictive scope,this is clearly indicated in the text.In addition,some grazing systems are not considered pastoralism(e.g.,grazed crops,intensive pasture systems)and are outside the scope of the report.Pastoralists refer to the individuals,households,and communities that practice pastoralism.Pastoralists raise sheep,goats,cattle,horses,donkeys,pigs,camels,yaks,llamas,alpacas,semi-domesticated species(e.g.,bison,caribou,reindeer),or harvest from wild species(e.g.,vicua).Some poultry systems,based on ducks or chickens,can also be considered pastoralism in certain contexts.Pastoralist systems are widely distributed,from the arctic to the tropics,often with herds of mixed species and breeds in the same production unit.Pastoralist communities tend to manage their land,water,and other natural resources in a sustainable,independent,and flexible way,often governed by rights to common resources and traditional or customary arrangements that safeguard rangeland health.Pastoralist livelihoods are diverse and subject to stressors,risks,and uncertainties due to global change impacts,including climate change and socioeconomic transitions.49 Traditionally,pastoralists have overcome these constraints,which have become increasingly more challenging,with resilience strategies and adaptive capacities.50Explanatory Note:The term“pastoralist”used in the report is often not recognised by pastoralists themselves,who may prefer to self-identify with other terms,such as herders,shepherds,ranchers,producers,farmers,or other terms customary in their respective countries and cultures.The report fully acknowledges all these identities and the diversity that underpins them but adopts the use of pastoralist as a comprehensive term to facilitate a global perspective and approach.Pastoralist systems and their management practices drive sustainable livestock production that is compatible with other land uses that respect ecological integrity and prioritise the functional health of rangelands.Pastoralist systems can merge with agricultural production systems(agroforestry and agropastoralism),51 or other systems that integrate trees into livestock production for shade and shelter(silvopastoralism)52 and for grazing in forests and woodlands(agrosilvopastoralism).53 Land governance concerns the rules,processes,and structures through which decisions are made about access to land and its use,the way those decisions are implemented and enforced,and the way in which competing interests are managed.Rangeland governance refers to the relationships between formal and informal institutions,and their policies,rules,and practices that shape human and environmental interactions on those lands.54 The responsible and inclusive governance of rangelands constitutes the foundation of many initiatives driving collective action to conserve,sustainably manage,and restore them.55 The meaningful participation of all stakeholders is a key enabling factor that can be enriched with information exchange,tenure security,polycentric institutional arrangements,and adaptive management systems.56Double Zanzano13GLOBAL LAND OUTLOOK Thematic Report on Rangelands and PastoralistsRangelands cover 80 million square kilometres,over 54 per cent of the terrestrial surface,constituting the largest land cover/use type in the world.Of this,78 per cent(62 million square kilometres)occur in the drylands,mainly in the tropical and temperate latitudes(Figure 2).Drylands are characterised by their hyper-arid to sub-humid climates,indicating different degrees of water scarcity with aridity indices ranging from 0.05 to 0.65,respectively.57 Many temperate rangelands which experience water scarcity are often considered de facto drylands.582.Rangeland health and degradation FIGURE 2 Indicative map of global rangelands according to ecoregions59 2.1 Rangeland characteristicsRangelands are highly diverse,both biologically and culturally,and occupy a range of biomes and ecosystems(Table 1).They support the livelihoods of approximately 2 billion people,60 with a multiplicity of uses and management systems that demand tailored context-specific approaches.61 Rangelands support pastoralist and extensive livestock production systems,primarily based on grazing,browsing,and pasture management,which are often the only sustainable type of land use in the rangelands.According to the Rangelands Atlas,livestock production systems in rangelands cover 67 million square kilometres or 45 per cent of the global land surface,almost half of which is situated in drylands.Rangelands generate 16 per cent of global food production and 70 per cent of feed for domesticated herbivores,most significantly in Africa and South America.62 Livestock provide food security and generate income for the majority of the 1.2 billion people living under the poverty threshold in developing countries.Rangelands provide high-quality,animal-sourced proteins that directly contribute to the nutrition and health of their inhabitants.63 While pastoralism offers significant potential for poverty reduction and more resilient livelihoods,64 indigenous peoples,pastoralists,agropastoralists,and other rangeland communities remain among the poorest and most marginalised people in the world.65Rangeland typesDeserts and xeric shrublandsFlooded grasslands and savannasMediterranean forests,woodlands,and scrubMontane grasslands and shrublandsTemperate grasslands,savannas,and shrublandsTropical and subtropical grasslands,savannas,and shrublands TundraNo1234567TotalArea km227,984,644.641,096,129.623,227,266.285,203,411.0010,104,079.6320,295,424.1911,598,465.2879,509,420.6414GLOBAL LAND OUTLOOK Thematic Report on Rangelands and PastoralistsTABLE 1Rangeland extent according to biome66Rangelands as social-ecological systemsRangelands can be managed for a multitude of economic,social,and cultural values that are supported by ecosystem health and functionality.67 This includes vital ecosystem services from local to global from provisioning and regulating to cultural and supporting services.Many scientific publications highlight the effectiveness of pastoralist practices in preserving and managing those services.68 Provisioning services,such as food,feed,forage,water,and fibre,are widely recognised,however,rangelands and their biodiversity can be managed to deliver other goods and services,such as nutrient/water cycling,carbon sequestration,animal/human health,recreation,and ecotourism.In terms of supporting services,rangelands hold exceptional biodiversity values,including habitat for numerous mammals and endangered species,representing one-third of all global biodiversity hotspots.69 Protected areas in the rangelands currently cover 9.5 million square kilometres or 12 per cent of the global rangelands.Additionally,many rangelands are managed under other effective area-based conservation measures(OECMs),an approach where long-term conservation and high-value biodiversity areas are prioritised.70 With respect to regulating services,rangelands comprise about 30 per cent of the global carbon pool,71 72 and account for most of the interannual variability in the global carbon sink.73 As stewards of the rangelands,pastoralists go beyond livestock production to safeguard critical ecosystem services,establishing a clear link between effective biodiversity conservation and pastoralism.The value of cultural services,such as identity and heritage,within rangelands is also noteworthy.They are home to 24 per cent of all languages and host numerous world heritage sites in recognition of their unique landscapes and cultures and the wealth of traditional knowledge a critical source of information to scale up SRLM and restoration practices.74 As in the past,rangelands continue to shape the culture and value systems,knowledge and world visions,and sense of purpose for pastoralists and other rangeland communities.Pastoralism and extensive livestock rearing in rangelands are widely distributed throughout the world.Currently,pastoralism is practised in more than 100 countries and supports about 200 million households with herds that total nearly a billion animals and account for about 10 per cent of the worlds meat production.75 With the limited use of external inputs,pastoralists manage the soil,water,and biodiversity to produce subsistence and value-added goods,such as dairy,meat,wool,and leather.Many of these products offer significant entry points for their participation in new markets that reward more sustainable value chains.The effective governance of rangelands requires an improved understanding of their dynamics,carrying capacities,and the future demand for their goods and services.There has been a recent shift from the unsustainable demand for the tangible or market goods produced in the rangelands,to policies and regulations that recognise and value the wider range of services they provide to people,nature,and climate.76 The challenge is to ensure that supply and demand are balanced in a sustainable manner,which includes addressing the synergies and trade-offs under transdisciplinary and multi-actor frameworks.2.2 Rangeland degradationWhile there are different understandings of rangeland degradation,77 they all point to the persistent loss and deterioration of rangeland health which is manifested in their reduced capacity to deliver ecosystem goods and services.Unsustainable land and livestock management practices,together with climate change and biodiversity loss due to land conversion,are among the direct drivers of rangeland degradation.Additional drivers which lead to rangeland degradation and fragmentation include tenure insecurity,conflicts over water and grazing boundaries,policies that incentivise the overexploitation of rangeland resources,and trends in market behaviour.78 Land degradation poses a significant threat to rangelands and their communities,taking a heavy toll on pastoralists by undermining their access to the natural resources needed to sustain their livelihoods.Rangeland degradation reduces income,productivity,and mobility which have negative implications for human and animal health,with the potential of conflict over increasingly scarce land and water resources.These impacts are differentiated across households,communities,and regions,disproportionately affecting marginalised or disenfranchised groups,such as women,youth,and indigenous communities.Rangeland degradation can also have far-reaching impacts due to hydrological disturbances,becoming a source of sand and dust storms which can increase animal mortality and reduce health and productivity in the wider landscape.The shortsighted use and management of rangelands typically result in:i.the fragmentation or loss of vegetation coverii.declining soil fertility due to soil erosion,salinisation,alkalinisation,compaction,and crusting;iii.water scarcity and moisture fluctuations;iv.the loss of biodiversity above and below ground;or v.any combination of these.79 Biome Rangeland cover(%)Deserts and xeric shrublands35%Tropical and subtropical grasslands,savannahs and shrublands26%Temperate grasslands,savannahs and shrublands13%Tundra15%Montane grasslands and shrublands6%Mediterranean forests,woodlands and scrub4%Flooded grasslands and savannahs1GLOBAL LAND OUTLOOK Thematic Report on Rangelands and PastoralistsRangeland degradation can trigger secondary consequences,like woody encroachment,invasive species,and the increased risk of drought and wildfires.The indirect drivers fuelling rangeland degradation are demographic shifts and the rapidly increasing demand for food,water,fibre,fuel,metals,and minerals.These pressures are often exacerbated by:i.weak or ineffective governance,ii.poorly implemented policies and regulations,iii.the lack of investment in rangeland communities and sustainable production models.81 These are virtually the same drivers contributing to land degradation and land use change occurring across all biomes and ecosystems of the world.The paradox is that efforts to increase food security and land productivity have converted millions of hectares of rangelands for crop production,aggravating land degradation processes and resulting in decreasing yields(Figure 3).Rangeland assessmentsThere are notable disparities in the assessments of land degradation which estimate its degree and extent globally.Land degradation is difficult to measure objectively,as it is seen as a mix of biophysical and socioeconomic factors which are often viewed subjectively.82 Estimates of rangeland degradation have changed over time,reflecting the progress made in the understanding of rangeland dynamics and indicators,assessment and monitoring tools,and management practices in the land use sector.83 Nonetheless,there are still critical gaps in the knowledge and data related to economic valuation,carbon pools,water cycle regulation,and shrub encroachment,to name a few.The first global rangeland assessment conducted in the early 1990s found that 73 per cent of the worlds rangeland area was degraded.84 This was widely contested due to the lack of field data needed to accurately verify rangeland degradation.In the last few decades,there has been a strong push to adopt a more holistic assessment approach which integrates the use of indigenous and traditional knowledge.85 More recent estimates of rangeland degradation have declined significantly,86 with some indicating that about 20 per cent of rangelands are experiencing negative trends,but experts are now concerned that these assessments may significantly underestimate the actual loss of rangeland health and productivity.87 According to the Food and Agriculture Organization of the United Nations(FAO),up to 35 per cent of grasslands are at risk of degradation,with other rangelands showing significant risk at 26-27 per cent.882.3 Monitoring rangeland health Data collection and real-time monitoring can be expensive and not easy to perform,rendering it difficult to assess rangeland health status and degradation trends.The use of Earth observation data is now common in many rangeland assessments,including numerous studies on land degradation utilising remote sensing tools and technologies along with open access data archives.89 Flagship initiatives,FIGURE 3 Feedback cycle of rangeland degradation80Demand for Food and ResourcesPolicy,Investment,and TechnologyRangeland Conversion andLoss of Pastoral LivelihoodsCropland Expansion,Soil Erosion,and Hydrological DisruptionIncreasedRate of LandDegradation Decreased ProductivityLand AbandonmentFood Insecurity16GLOBAL LAND OUTLOOK Thematic Report on Rangelands and Pastoralistssuch as the Group on Earth Observations Land Degradation Neutrality(GEO LDN)90 and the FAO System for Earth Observation Data Access,Processing and Analysis for Land Monitoring(SEPAL)project,91 have given a sharper focus on monitoring land degradation trends and highlighting rangeland health as a key global priority.Another way to assess rangeland health relies on the experience and involvement of pastoralists and other rangeland users.The Participatory Grassland and Rangeland Assessment(PRAGA)is a methodology developed by FAO and the International Union for the Conservation of Nature(IUCN)and financed by the Global Environment Facility(GEF).PRAGA aims to assess rangeland health according to the management objectives of local land users and is based on a combination of scientific,indigenous,and local knowledge.It is designed to support decision making with actionable information and data that can help guide policy and action to halt degradation and restore rangeland health and productivity(Figure 4).FIGURE 4 Nine key steps to implement the PRAGA methodologyA global framework can assist countries and communities when designing a monitoring and evaluation approach for SRLM and restoration that is specific to local circumstances.Assessments can be organised according to the key underlying factors of degradation,and integrated into a conceptual framework that addresses social-ecological processes in rangelands.Like human health,92 rangeland health is impacted by many causes and has symptoms that are particular to the context and circumstances.A comprehensive framework to assess landscape functions can be used to monitor degradation and restoration,such as the methodology designed by the United States Department of Agriculture,which involves creating indices based on simple field indicators that reflect the key attributes of rangelands(Table 2).93TABLE 2 Three attributes and 17 indicators used by the United States Department of Agriculture to assess rangeland health94ASSESSMENTPARTICIPATORYPREPARATORY010203040506070809STEPS PHASESACTIONPartnership development:local and national ownership of the processIdentifying the landscape for assessmentBaseline reviewLarge scale assessment and remote sensingParticipatory mapping of target landscapeParticipatory indicator selectionComposition and selection of assessment teamField assessmentData management post-assessment and validationBASELINEANALYSIS AND INTERPRETATION Soil/site stabilityHydrologic functionBiotic integrity1.Rills12.Functional/structural groups2.Waterflow patterns13.Dead or dying plants or plant parts3.Pedestals and/or terracettes15.Annual production4.Bare ground16.Invasive plants5.Gullies6.Wind-scoured and/or depositional areas14.Litter cover and depth7.Litter movement10.Effects of plant community composition and distribution on infiltration17.Vigor with an emphasis on reproductive capability of perennial plants8.Soil surface resistance to erosion9.Soil surface loss and degradation11.Compaction layer17GLOBAL LAND OUTLOOK Thematic Report on Rangelands and PastoralistsWhile there is not one assessment methodology that would be uniformly applicable to all situations,there are sufficient common elements to begin monitoring under a flexible global framework that is tailored to different contexts.The rangeland health framework constitutes a steppingstone in the process to build a conceptual framework that addresses the challenges and envisions solutions as demonstrated by the Driver-Pressure-State-Impact-Response(DPSIR)model95 which addresses complex challenges at the interface of society and the environment(Figure 5).962.4 Conceptual framework for rangelands and pastoralism Rangelands are associated with their actual or potential use for grazing and,thus,primarily characterised as managed lands.Raising livestock is an important,but not exclusive,activity in the rangelands which can offer a mix of social,economic,and environmental benefits.The multifunctionality of rangelands is seen as a desirable outcome which demands sound management practices and committed people implementing them.97 The report emphasises the development and operationalisation of policy,planning,and implementation mechanisms under an umbrella of sustainable management approaches.This is reflected in the conceptual framework where the elements and relationships shaping rangelands are organised in an interactive way,pointing to multifunctional approaches that link rangeland health and specific management systems(Figure 6).The framework shows how pastoralists and rangelands are intimately linked within the same social-ecological system and points to the need for a systemic approach to understanding and managing rangelands.Beyond just land users,pastoralist communities have been,and still are,considered stewards of the rangelands.98 They bear the ultimate responsibility for,and consequences of,their management practices.While the participation of other land users and stakeholders in rangeland governance is important,pastoralists must be prioritised as shareholders with the capacity to sustainably manage and restore rangelands.It is this complex network of relationships occurring in diverse political and social environments that ultimately shapes the use and management of rangelands.Addressing land governance challenges opens the scope of interventions to the whole territory and to all stakeholders involved,often seen as a prerequisite for achieving the national and global objectives addressed in the report.99 The conceptual framework,complemented with the DPSIR model,underpins the global effort to protect rangelands and contributes to the effectiveness of initiatives at national and local levels.As many rangelands share common features,multi-scale approaches and context-specific interventions will help refine a global conceptual framework.In addition to generic strategies and approaches,case studies and good practices can also help inform specific response measures,management systems,and governance approaches used by various initiatives(Chapter 4).FIGURE 5Driver-Pressure-State-Impact-Response(DPSIR)model of rangeland health and degradation statusEnabling environments Improved policy frameworksSustainable Rangeland ManagementGender responsivenessGrazing mobility,rotation and restRecognition and differentiationAgroforestry and multifunctionalityEcosystem restorationImproved governance/institutionsTechnical improvementsUpgraded investmentsResearch,data collection,monitoringCo-construction of knowledgeCapacity and social capital buildingLobby and advocacyInclusive and participatory planning and managementEquity and inclusion of youth,women and other groupsRESPONSEReduced ecosystem servicesLack of support capacityLess biodiversityReduced biomass/cover Erosion/soil lossIncreased wildfire risk Shrub encroachmentInvasive speciesLess water retentionLess infiltrationDrougths/extreme eventsLess water availabilityReduced carbon storage Increased emissions Soil occupationStewardship activities lostFertility lossProductivity lossFood insecurityLivestock bad health/shapeIncreased povertyWorse quality of life Eviction/Social conflict Lack of recognition/imageIMPACTPRESSUREHuman ActivitiesDRIVEREconomic,market and social drivers:Global demand Capital investment TechnologyPolicies,legal,and institutional frameworksand government actionBiophysical factorsResource variabilityin the rangelandsGlobal changeClimatic changeLand use changePopulation growthGlobalizationSTATERangeland Health and DegradationMISMANAGEMENTUnbalanced grazingOverexploitationReduced mobilityShort resting periodsInadequate breeds Fragmentation/fencingSedentarizationLack of governance Weakened institutionsAbandonmentGLOBAL TRENDSUrbanizationMigrationMarket rulesIncreased demand of animal productsHealth and food threatsComplexation of urban-rural interface Inequity/discriminationTRANSFORMATIONConversion to other uses(mining,crops,urban)Incompatible usesAfforestationPrivatization and tenure changesLand demand/grabbingConflicts over landGrazing bansAgriculture intensification18GLOBAL LAND OUTLOOK Thematic Report on Rangelands and PastoralistsTechnical solutions to avoid,reduce,and reverse rangeland degradation through conservation,sustainable use,and restoration practices are cost-effective,widely available,and supported by scientific evidence.Incentives in the form of secure tenure,access to markets and credit,and the provision of extension services are important forms of support for pastoralists engaged in SRLM and restoration activities.Participatory and multi-actor initiatives help ensure the inclusion of all relevant stakeholders in planning,implementation,monitoring,and evaluation.There are numerous manuals,guidelines,and training materials that offer a range of technical measures to avoid,reduce,and reverse degradation trends in rangelands(Chapter 5).100 FIGURE 6 Social-ecological conceptual framework in the context of rangeland management and restorationMichal KnitlLAND GOVERNANCEPASTORALISTSOTHERSTAKEHOLDERSRANGELANDSOTHERGRAZING LANDSSUSTAINABLERANGELANDPLANNINGMANAGEMENTPROTECTIONAND RESTORATIONLAND COVERLAND USEGRASSLAND ECOSYSTEMS Grazing landsSavannahsCerradosMiomboSteppesShrublandsWoodlandsOTHERECOSYSTEMS DrylandsWoodlandsForestsWetlandsTundraMosaicsHUMAN-MODIFIED LANDSCAPESCrops and agricultural fieldsPeri urban and urban open areasWastelandsParks and public areasLIVESTOCKPRODUCTION SYSTEMSNomadicTranshumantShepherd-based localRotational grazingAgropastoralMultifunctional systemsAgrosilvopastoralSilvopastoralRanchingWildlife herdingOTHER SYSTEMS AND ACTIVITIESHuntingBeekeepingRewildingUrban Industrial agricultureNature conservationEnergyTourismENABLING ENVIRONMENTSPOLICY&SOCIAL FRAMEWORKFULLConflict-solving,Facilitation,Baseline assessment,Decision makingClimate,biodiversity,and ecosystem servicesInstitutional and collective capacitiesEducation,values,and cultural identityTools,technologies,and technical supportLegislation,regulation,and investmentLand governance and tenure securityKnowledge and awareness raisingData,research,and monitoringINVESTMENTPublicPrivateCommunityDonorsPARTICIPATIONManaging trade-offs and interests,19GLOBAL LAND OUTLOOK Thematic Report on Rangelands and PastoralistsUsing examples from all regions of the world,the report demonstrates the untapped potential of rangeland projects and programmes to provide multiple co-benefits for people,nature,and climate(Chapter 4).Evidence suggests that successful SRLM and restoration projects and programmes have several common elements:i.informed,targeted,and sustained finance;ii.meaningful participation of all relevant stakeholders in the assessment,planning,design,implementation,monitoring and evaluation stages;iii.establishment of clear goals and measurable ecological and socioeconomic objectives;101iv.space for innovation and adaptive management;v.focus on governance,enabling environments,and supporting policies;vi.use of qualitative and quantitative data,indicators,and other information for monitoring,evaluation,and communications.Even when these elements are contained in SRLM and restoration projects and programmes,the specific challenges and complexities of rangelands and pastoralism result in an alarmingly high rate of failure.102 This is not unique to rangelands,especially considering the unequal power dynamics associated with land and natural resources that often marginalise many rural communities.As with nature conservation and rural development,rangeland users and managers must be proactive,undertake systematic analyses,and implement strategies that learn from these failures rather than seeking to mechanically replicate actions that may have been successful in very different contexts.103 The systematic analysis of rangeland projects and programmes was common during the 1990s and 2000s,104 105 106 107 but has since diminished significantly with a few notable exceptions.108 109 110 Despite recent efforts to support and implement new rangeland and pastoralist initiatives,111 there is still limited evidence on the main constraints and bottlenecks.While there is increasing public attention and scientific literature devoted to the contextual and conceptual understanding,much less has been reported on the technical aspects.This chapter addresses both the underlying concepts and the technical aspects of rangeland and pastoralist projects and programmes while providing a critical historical perspective and offering pathways of action that can enhance the success of current and future policies,projects,programmes,and investments.3.1 A historical perspectiveHistory provides an obvious first step to understand the various challenges that limit the success of rangeland and pastoralist projects and programmes.While perspectives on pastoralism,rangelands,and rural development have evolved considerably over the past 50 years,current initiatives tend to perpetuate common misconceptions.In the 1950s and 1960s,livestock and rangeland initiatives were focused primarily on technical improvements in production systems(e.g.,industrial breeds,forage production,groundwater extraction,veterinary care)with the exclusive aim of modernisation that overlooked the value of pastoralist livelihoods and management systems.In the 1970s,pastoralism began to gain increased global recognition.However,attention was still centred on how to transform pastoralist livelihoods through settlement and modernisation.For many new nation states,government priorities,much like those of their colonial predecessors,were focused on efforts to assert their authority,secure borders,and reduce conflict.Investments were directed towards improving infrastructure,technical assistance with animal health,industrial livestock production methods,and marketing as part of an overall strategy of intensification.112In the last decades of the 20th century,rangeland management gradually shifted its approach with more projects and programmes that created grazing reserves,reduced herd sizes,promoted cooperatives,and improved land governance and tenure security.In general,the scientific understanding of rangeland functioning improved,while many outdated colonial perceptions receded.This paradigm shift had important implications for SRLM and restoration which have yet to be fully realised,especially with regard to poverty,decent work,and environmental sustainability.Since the 2010s,methodologies,analytical tools,and good practices have advanced but have not matched the pace of improvements in the conceptual understanding and frameworks for action.Land and livestock managers involved in rangeland and pastoralist initiatives need practical applications that respond to these new,updated frameworks.113 While it is increasingly popular to design and promote community based SRLM and restoration projects under adaptive approaches,114 many historical flaws and challenges remain(Table 3).115 3.Learning from the past,planning for the future20GLOBAL LAND OUTLOOK Thematic Report on Rangelands and PastoralistsTABLE 3 Conceptual causes of failureConceptual defectsActions with shortcomingsCausesConsequencesInsufficient recognition of pastoralismDeveloping goals for pastoralism that are misguided,under a conventional perspectiveMisconceptions of pastoral systems and prejudices over pastoralists Project failure,abandonment Promoting changes regardless of their impact on basic needsMisunderstanding of traditional pastoralisms role in subsistence and risk prevention Impoverishment conflict,vulnerability Destocking,resizing herds,promoting“alternatives”Lack of recognition of the economic,social and cultural values of pastoralist culture Vulnerability and marginalisationDeveloping actions that focus on the role of adult menDisregard for the roles of women,youth and other groupsInequity,lack of replacementConducting poor baseline assessmentUndervaluation of traditional knowledge,insufficient knowledge availableShortcomingsUnderestimation of the complex interacting forces in pastoralist environmentsTransforming rangelands towards different usesEconomic interests,misguided policiesLoss of pastoral lands,increased stress,loss of critical assets for pastoralistsFocusing on large stationary infrastructure,slaughterhouses,water.Ignoring need for mobility and flexibility in pastoralism,maladapted water infrastructure Lack of water,uneven grazingReducing pressure,destocking,developing misled grazing plans Misguided interventions on grazing and mobility regimes Uneven grazing,land degradationFocusing on overstock herd sizes,fenced ranching,private land rightsMisled rangeland management,lack of flexibilityUneven grazing,land degradationDeveloping actions that are not flexible under changing conditionsLack of awareness of change and variability,unexpected events harming project planningIncreased risk of failureOversized technological interventionsFocusing on high-performance breeds,external inputs,feed supplementationAim for intensification of pastoralist productionCollapse of natural resourcesEncouraging settlementSedentary mindset of external developersConflict,impoverishmentFocusing on fencing,water points,centralised infrastructureInadequate investments based on non-flexible approachesLoss of mobility,economic failures Prioritising technical action.Overlooking social,economic and cultural issues and needsPoor social outcomes,hidden constraintsMisunderstanding of pastoralists decision-making and governance institutions and processes Developing participatory actions that overlook/lack key agentsNon-definition of the community involved,participants not well chosen,lack of diversity in representation of participantsInefficiency of participationNot developing specific actions to secure rightsLand rights and security of tenure overlooked and insufficiently consideredInsecurity,conflict,misuse of resourcesPromoting alternative activities for pastoralistsAttempt to change pastoralist perception or behaviour;pastoralism is weakenedConflicts,imbalanced power,abandonmentImplementing state and promoters interventions unilaterally Overlooking of existing governance institutions and local management capacitiesWeakened traditional governance institutions,conflicts,degradationEnabling centralisation,homogenisationMarkets unaware of pastoralists needs,lack of synchrony between markets and pastoralistsPoor access to markets for pastoral productsMisinterpretation of the role of commonsAllowing privatisation,land grabbing,state appropriation of common landsMisconception about the importance of common landsWeak governance,mismanagementLack of participation from the early stagesDesigning projects that lack necessary capacitiesPoor use of pastoralist experience,knowledge and skills,top-down approaches,resource constraints,cultural/language barriersMaladaptation of the projectInadequate state actionClosing borders,assigning lands to the state,limiting land and movement rightsStates consolidating their power over land,action of state weakening traditional systemsLoss of mobility,insecurity,conflict21GLOBAL LAND OUTLOOK Thematic Report on Rangelands and PastoralistsStates continue to try and control pastoral lands,especially in border or conflict areas,where pastoralists previously moved freely.At the same time,the most substantial investments are aimed at projects and programmes that convert rangelands into large-scale irrigated agriculture,tree plantations,renewable energy projects,and even protected areas.Legal frameworks,development plans,and private investments are driving these land use changes,while land grabs and the free,prior,and informed consent for investment in pastoral areas are often ignored or given only token attention.116 As a result,pastoralists and other rangeland stakeholders are often excluded,distanced from their land and cultural identity,or forced to abandon their traditional livelihoods.3.2 Learning from the pastThe report emphasises two key means to address the shortcomings of the past.The first is that pastoralism and extensive livestock production need to be fully integrated into projects and programmes to improve rangeland health.117 While pastoralism is not the only human activity on rangelands,it is often the most critical one to consider.Failure to do so can reduce the efficiency and effectiveness of rangeland initiatives that aim to boost their health and productivity,118 such as those focused on rural development,119 nature conservation,120 or ecosystem restoration.121 A conventional approach to SRLM and restoration is often inefficient and even counterproductive,such as when a project employs measures to conserve biodiversity without considering livestock production.122 Strategies that overlook the role of grazing and instead focus on other practices(e.g.,exclosures,seeding,beekeeping)are often insufficient to adequately address the degree and extent of rangeland degradation.123 124 125 It is important to recognise that pastoralism can directly and indirectly accelerate progress towards land and ecosystem restoration targets,such as by enhancing ecological connectivity through the preservation of traditional transhumance routes.The second key means to address shortcomings is to create synergies between nature/climate goals and integrated management-based approaches that seek to improve food security,livelihoods,and sustainable production in rangelands.These approaches are not only compatible but complementary as they both draw on recognised SRLM and restoration principles and prioritise the participation,rights,and knowledge of indigenous peoples and local communities.A flexible and context-specific management approach can help minimise trade-offs and maximise returns on limited investments.The potential shortcomings analysed below can help inform rural development and ecosystem management initiatives even though they do not specifically address the multifunctionality of rangelands or pastoralism.A lack of focus on rangelands or pastoralism does not mean that they should be ignored.In some cases,they serve to highlight misguided strategies that could yield more benefit through improved design and implementation.3.3 Project formulationOne means to improve the way rangelands and pastoralist initiatives are formulated is to ensure that a fit-for-purpose conceptual framework is applied at all stages of the project cycle.A fit-for-purpose conceptual framework offers a starting point to improve project and programme design through a holistic perspective on rangelands and pastoralism one that is adapted to local realities by ensuring inclusive and meaningful participation as well as the institutional arrangements that support collaboration and cooperation during all phases of the project cycle.Each element of the framework(e.g.,land uses,ecosystems,stakeholders,institutions,production systems,cultural norms)can be mapped and acknowledged within the local context to provide a comprehensive baseline assessment.126 Project design and funding proposals must increasingly recognise the role of pastoralists and their rangeland management practices.FAO and International Fund for Agricultural Development(IFAD)have developed three strategies to overcome these shortcomings and create a minimum standard for sustainable pastoralism:127 i.develop national development strategies and action plans that recognise and support pastoral systems;ii.avoid policies and investments that undermine pastoralism;iii.improve land governance and tenure security to enfranchise pastoralist communities while recognising their diversity as a valuable asset.1283.4 Rangeland interventionsIn addition to conceptual failures,the poor quality of technical interventions is another leading cause of disappointment in many rangeland initiatives.The analysis of common technical flaws has been arranged according to the project life cycle:i.conducting baseline assessments;ii.design and planning;iii.implementation;iv.monitoring and evaluation.129Envato22GLOBAL LAND OUTLOOK Thematic Report on Rangelands and Pastoralists3.4.1 Baseline assessments External drivers and pressures are frequently identified as threats to the success of a project,however insufficient knowledge of the status and dynamics of the rangelands targeted for intervention is a significant constraint.A poor baseline assessment can seriously weaken the design of rangeland and pastoralist projects which make them unlikely to be well adapted to the realities on the ground.This may be due to a lack of actionable data(e.g.,gender-disaggregated),a disregard of local knowledge when planning new initiatives,or power dynamics that lead to subjective analysis and misinformation that perpetuates biases and narrow interests(Table 4).TABLE 4 Baseline analysis-related causes of failure ThreatsOriginCausesConsequencesIncomplete baseline analysisLack of data and informationInsufficient information for decision-making,actions led by incomplete data Unpredictability of resultsGeneralisations about the pastoral development environmentInadequate scale of work,projects developing conventional actionsLack of compatibility between actions and local conditionsVagueness of key parameters:beneficiaries,project scales Inadequate targets,actions pointing to misguided targetsInefficiencyLack of risk assessmentRisks underestimated,not measured or forecasted,lack of adaptation capacity High vulnerability of projects to riskLack of inputs from similar projectsUnawareness of potential mistakes and constraints,repeated errors of other projectsAvoidable mistakes:unrealistic optionsLack of inputs from local stakeholders and pastoralistsLack of contact with,or awareness of,the reality,actions not aligned with local interestsIll-defined roles,responsibilities and processesMisunderstanding of power balanceBiased information and diagnosticsFavouring particular interests,ill-defined roles,responsibilities and processes,lack of common goalsMisidentification of stakeholdersUnbalanced outcomes;actions not aligned with common interestsIll-defined roles,responsibilities and processes3.4.2 Design and planning Another potential cause of project failure results from poor choices in the design stage which leads to a weak operational plan.Table 5 lists and elaborates upon factors which could be addressed with alternative choices at the start of the project,while others are unavoidable but still need to be considered.One example refers to partner selection.The lack of reliable partners(e.g.,local authorities,NGOs/CSOs,private sector)can undermine project success if roles and responsibilities are unclear or there is a lack of critical stakeholder consultations during the design and planning stage.Another refers to the need for clear project objectives,such as production,performance,and productivity,to guide operational plans and meet the aspirations of rangeland producers and pastoralist communities.The choice of project or programme scale is instrumental to prevent mismatches between biophysical interventions and socioeconomic goals as well as to address resilience trade-offs across scales.130 In addition,sustained finance,institution building,and developing a solid evidence base need to be fully considered in the design and planning stage.131 23GLOBAL LAND OUTLOOK Thematic Report on Rangelands and PastoralistsTABLE 5 Design and planning causes of failureThreatsOriginCausesConsequencesUnfavourable policy environmentsLegal framework incompatible with the project Projects trying to succeed under restrictive legal environments,poor legal supportWeakness,lack of recognition,abuseUnfavourable scenariosUnplanned influence of external factorsEconomic,social and environmental constraints,increasing barriers to actionsPoor resultsUnfavourable political relationshipsPoor relationship with governmentsProjects not integrated into larger programmes,isolated actionsLow impactHidden agendasIntrusion of external goals and agendasPriority given to external goals instead of project goals,actions unaligned with project goalsLack of trust and commitment Political expediencyIntrusion of implicit politics and government interestsPriority given to political goals instead of project goalsLack of trust towards states and policiesUnsatisfactory partner selectionPartners not suited for their role in the projectPartners unable to fulfil their commitments;lack of capacity,insufficient influence,poor performance Prevalence of opportunity interests;actions not properly developed by responsible partnersLack of efficiencyPoor strategic planningDiscontinuity between baseline and strategyInadequate solutions;use of conventional targets for pastoralist productionsIncapacity to reach goalsPoorly defined problemSymptoms addressed rather than causes Actions unable to introduce changesLack of correlation between target and actionsIncoherent project,inadequate actions Lack of resultsLack of flexibility in specific objectivesLow capacity of reaction facing uncertainty,Pursuit of project goals,regardless of other circumstancesProject goals become unreachable or irrelevantLack of contingency plansUnforeseen difficulties that stress implementation,lack of flexibility Lack of efficiencyLack of reactive capacityNo element(s)of responsiveness,actions insensitive to external conditionsInability to respond to changing conditionsLack of project ownershipLack of participation/consultingActions seen as not aligned with beneficiary interests or needsLow impact/interestLack of ownershipBad strategic choices and technical shortcomingsLack of development-planning skills among project personnelWeak project-building process,actions uncoordinatedLoss of synergiesFailure to involve pastoralists in the planning processInsufficient mapping and incorporation of stakeholders Unfit field actionNeglect of institution building/consolidation/updatingLack of facilitation,lack of governance and access to resourcesActions not properly deployed in the fieldOmission of goals related to justice and sustainabilityActions not addressing critical sectorsImbalance of results Faulty,unproven,or inappropriate technologyInadequate tools to reach goals,limited effectiveness of actionsGoals not fulfilled24GLOBAL LAND OUTLOOK Thematic Report on Rangelands and Pastoralists3.4.3 ImplementationShortcomings during the implementation stage can significantly reduce the expected benefits of the project or even generate unintended harmful outcomes.Implementation flaws have been detected in many rangeland and pastoralist initiatives.Even well-designed projects and programmes can fail due to a lack of capacity,skills,supervision,or commitment to execute activities in a coherent way.Other key issues associated with the implementation stage that may limit success include:i.lack of sustained financial and technical support due to short project cycles;ii.insufficient linkages with existing local institutions and attention to socioeconomic conditions;iii.forced scaling up/out of untested or immature interventions;andiv.biases towards market-based mechanisms and incentives even when they are inappropriate or undermine cultural values(Table 6).TABLE 6Implementation-related causes of failureThreatsOriginCausesConsequencesShortcomings in project managementPoor integrity and coordination between actionsUnexpected interactions Contradictory results Weak managerial skills and experience of personnelActions poorly managed,weak project implementationLow impactPoor communication on project teams and with stakeholdersLow level of coordinationReduced impactOver-management and bureaucracyTeams more focused on paperwork than actions,inefficiency Shortcomings in action implementation,burnoutWeak structural or systemic capacity of project managersWeak project implementation,underachieving actionsLow impactUnderstaffing,low capacityWeak economic capacityFew personnel to implement actions and manage the project,low capacityWork overload,underachievementLow commitment from participantsWeak participation processes,lobbying and networkingLack of support,actions underachieving goalsLack of efficiencyEnvato25GLOBAL LAND OUTLOOK Thematic Report on Rangelands and Pastoralists3.4.4 Monitoring and evaluationThe lack of capacity for evaluation and monitoring is often a challenge for many project managers and implementing agencies.Project evaluations frequently highlight deficiencies in understanding the local context as well as the capacity and flexibility of local stakeholders to implement off-the-shelf measures which can involve balancing risk taking and risk aversion.Monitoring and evaluation protocols tend to be ad hoc or have a low profile in the operational plan of many rangeland initiatives.This underscores the importance of research applications to improve information flows that increase the capacity for adaptation through contingency plans and risk management strategies.Participatory approaches to monitoring and evaluation should be explored whenever possible(Table 7).Many of the shortcomings in rangeland projects and programmes can be addressed systematically by using a checklist developed from the tables above.Most urgent is the need for a coherent conceptual framework to help guide their design,implementation,and monitoring.The next chapter provides brief insights into rangeland and pastoralist initiatives from around the world that can help strengthen that framework.The case studies point to different strategies and approaches that spotlight the diversity of rangelands and pastoralist systems.While many of these projects and programmes are underfunded and rarely acknowledged,their efforts to overcome challenges and constraints are an inspiration and a rich base of evidence to guide other SRLM and restoration initiatives.TABLE 7Monitoring and evaluation-related causes of failureThreatsOriginCausesConsequencesEvaluation shortcomingPoor monitoring systemLow feedback from the environment,actions unable to be redesigned Lack of responsiveness and reactive capacityLow feedback from working teams,actions unable to be reprogrammedLack of responsiveness and reactive capacityLack of supervision/reviewLow feedback from supervisors,actions unable to feed future projectsLack of improvement capacityEnvato26GLOBAL LAND OUTLOOK Thematic Report on Rangelands and Pastoralists4.Regional analysis and case studiesMany countries,organisations,and communities are engaged in SRLM and restoration activities that:i.recognise the critical role of rangelands and pastoralists in achieving LDNii.help create the enabling conditions and participatory governance systemsiii.provide technical and financial support.This chapter is divided into 10 sections roughly corresponding to regions of the world.Each section starts with an introduction,followed by a regional analysis supported by national and sub-national case studies.Each section concludes with an overview of rangeland degradation trends and a discussion on the key issues considered most significant in advancing the SRLM and restoration agenda.MethodologyThis chapter contains case studies submitted by diverse stakeholders(contributors are listed in the acknowledgements)who responded to a request by the UNCCD secretariat to submit their experiences related to rangeland management and pastoralism.The call for contributions was opened to all UNCCD stakeholders,including national focal points,in February 2023 along with a submission template.A total of 65 case studies from 39 countries were received as well as numerous global and regional initiatives(Chapter 5).After an initial review,each contributor was asked to provide additional references,data,and photos or to clarify specific issues.No effort was made to validate,complete,or update the information provided in the submissions.In the end,55 case studies were selected to provide a representative balance between regions,countries,and approaches.Contributors were also asked to review the final text of their case study.The case studies are presented here with due respect to the original content and style,offering insights into a wide range of design,implementation,and monitoring approaches.Statistical data and maps displayed were extracted from referenced publications and supported by scientific evidence,fully recognising that this information could be outdated or differ from official sources.While not reflecting the full status or breadth of rangeland policies or interventions in countries or regions,the case studies demonstrate a diversity of strategies and methodologies that address many of the specific drivers,pressures,impacts,and solutions highlighted throughout the report.The report refrains from evaluating their performance,measuring their success,or criticising their approaches.The references provided allow the reader to explore further details and draw their own conclusions.4.1 East Africa East Africa is characterised by expansive drylands,which occupy nearly 75 per cent of its land surface,ranging from 20 per cent in South Sudan to 99 per cent in Eritrea.Pastoralism is the predominant land use,with these communities representing a significant proportion of their populations.Pastoralism produces almost 90 per cent of the livestock and animal products consumed in the region,contributing to GDP in Ethiopia(19 per cent),Kenya(13 per cent),Uganda(8 per cent)132 and,on average,57 per cent of the agricultural GDP in the 8 member states of the Intergovernmental Authority on Development(IGAD).133 Nevertheless,poverty and forced migration in pastoralist communities are widespread and concerning.134 EAST AFRICA WEST AFRICA NORTH AFRICA/MIDDLE EAST CENTRAL ASIA/MONGOLIA EUROPE CHINA/SOUTH-EAST ASIA SOUTH AMERICA NORTH AMERICA OTHER COUNTRIESSOUTH ASIA FIGURE 7 Regional distribution of case studies27GLOBAL LAND OUTLOOK Thematic Report on Rangelands and PastoralistsPastoralist communities constitute a range of culturally and linguistically diverse groups which is reflected in their varied production systems,livestock species and breeds,and the use of natural resources and external inputs.Nonetheless,they share a common livelihood strategy whereby mobile pastoralism relies on extensive common lands,decentralised decision-making that accounts for diverse voices and interests,and often employs opportunistic strategies to cope with scarcity.135 For example,traditional tenure systems favour communal access and priority of passage to move herds between key resource areas.East African rangelands are widely acknowledged for their cultural and biodiversity values.Pastoralists and their livestock have played a large role in shaping the ecology of the rangelands through their grazing,mobility,and fire management practices.136 137 These activities influence vegetation and tree cover by controlling shrub encroachment and protecting wildlife habitat.Since pastoralism emerged as a land use system in sub-Saharan Africa more than 5,000 years ago,natural resource management and herding strategies have modified ecosystems to such an extent that,in many cases,the removal of pastoralism would be detrimental to biodiversity conservation efforts.The linkages between biodiversity and pastoralism call for an integrated conservation strategy that fully considers the needs and rights of pastoralists,138 while recognising that wildlife populations in many rangeland areas are experiencing drastic declines due to land degradation,land use and climate change.139East African rangelands are undergoing a significant shift towards a better recognition of their multiple benefits and values,unleashing demand to acquire,control,and invest in these lands.140 While this transformation is helping to reverse decades of underinvestment and marginalisation,141 governments and investors now see these rangelands as development frontiers with abundant land and resources,142 with major actors investing in the construction of ports,pipelines,roads,solar/wind farms,and monoculture plantations.These large-scale investments,which are often part of wider commercial and development strategies,can offer opportunities to reduce poverty and increase the resilience of rangeland communities.Unfortunately,many of these investments tend to disrupt traditional management practices and ignore customary land rights.Pastoralist representation in politics and governance does exist in some countries,such as Ethiopia,Kenya,and Uganda where parliamentary bodies have been established and enjoy different levels of formalisation.However,policies at the national level rarely support mobile pastoralist livelihoods,but instead promote sedentary and“modern”livestock production systems even though many civil society and non-governmental organisations have long been advocating for the interests of traditional mobile pastoralists.143Addressing rangeland challenges in East Africa requires coordinated action to design and sustain finance to implement SRLM and restoration initiatives at regional,national,and local levels(Figure 8).Rangeland productivity and economic diversification can only be addressed by strengthening critical linkages within social-ecological systems.Integrating locally adapted management practices,agricultural technologies,and extension services have the potential to simultaneously target SRLM,food security,and improved livelihoods.144 For example,ecotourism in rangeland and pastoralist areas can be a driver for economic diversification in East Africa.There have been significant efforts to formulate both regional and national policies(Table 8).FIGURE 8 Cause-effect framework of challenges in the rangelands of the IGAD region145Degraded Rangelands/Pasture ScarcityUnhealthy Rangelands and Insecure Pastoral LivelihoodsResource ConflictsLack of SupportivePolicies and LegalFrameworksWeak Statutoryand CustomaryInstitutionsInadequate Research,Extension Services,Human Resourceand KnowledgeManagementFrequent and More Intense Droughts and Other Climate Change ImpactsLow Investmentin SustainableRangelandMangementInsecure LandRightsRangelandEncroachmentand FragmentationRestrictedMobilityand UnequalAccess toResources28GLOBAL LAND OUTLOOK Thematic Report on Rangelands and Pastoralists CountryPolicy/Strategy/PlanStatusUgandaRangeland Management and Pastoralism Policy,2017DraftSudanThe Rangelands and Forages Resources Development(Rationalization)Act,2015OperationalPastoral Strategic Action Plan for Semi Desert Savanna Sudan,20142024OperationalSouth SudanNational Livestock Development PolicyOperationalMARF,Policy Framework and Strategic Plans,20122016OperationalEthiopiaPastoralist Development Policy and Strategy,2018DraftNational Strategy on Prosopis Juliflora Management,2017FinalisedThe Federal Rural Land Administratio
2024-05-31
110页
5星级
SPAINS ENERGY AND GREEN TRANSITION Potential opportunities and partnerships for Swedish companies w.
2024-05-31
16页
5星级
1IntroductionHumanity has only one Earth,our one and only home.In the face ofenvironmental challenge.
2024-05-31
32页
5星级
Fantastic Freshwater 1FANTASTIC FRESHWATER50 Landmark Species for ConservationFantastic Freshwater 2.
2024-05-31
41页
5星级
IUCN WCPA Good Practice Guidelines Series No.XThe world at nightPreserving natural darkness for heri.
2024-05-30
160页
5星级
Global Critical Minerals Outlook 2024The IEA examines the full spectrum of energy issues including o.
2024-05-28
282页
5星级
Water sector financial governance gap analysis in Central AsiaGeorg Petersen Erkin MadumarovAiman Ut.
2024-05-28
149页
5星级
Time for TransparencyDeforestation-and conversion-free supply chains2023 CDP Forests DataMay 2024DISCLOSURE INSIGHT ACTION2Time for Transparency3Time for TransparencyImportant NoticeThe contents of this report may be used by anyone providing acknowledgment is given to CDP Worldwide(CDP).This does not represent a license to repackage or resell any of the data reported to CDP or the contributing authors and presented in this report.If you intend to repackage or resell any of the contents of this report,you need to obtain express permission from CDP before doing so.CDP has prepared the data and analysis in this report based on responses to the CDP 2022 information request.No representation or warranty(express or implied)is given by CDP as to the accuracy or completeness of the information and opinions contained in this report.You should not act upon the information contained in this publication without obtaining specific professional advice.To the extent permitted by law,CDP does not accept or assume any liability,responsibility or duty of care for any consequences of you or anyone else acting,or refraining to act,in reliance on the information contained in this report or for any decision based on it.All information and views expressed herein by CDP is based on their judgment at the time of this report and are subject to change without notice due to economic,political,industry and firm-specific factors.Guest commentaries where included in this report reflect the views of their respective authors;their inclusion is not an endorsement of them.CDP,their affiliated member firms or companies,or their respective shareholders,members,partners,principals,directors,officers and/or employees,may have a position in the securities of the companies discussed herein.The securities of the companies mentioned in this document may not be eligible for sale in some states or countries,nor suitable for all types of investors;their value and the income they produce may fluctuate and/or be adversely affected by exchange rates.CDP Worldwide and CDP refer to CDP Worldwide,a registered charity number 1122330 and a company limited by guarantee,registered in England number 05013650.2024 CDP Worldwide.All rights reserved.ContentsExecutive summaryIntroduction:A time of rapid action for deforestation-and conversion-free supply chains2023 CDP forests disclosureDisclosure of progress toward DCF goalsPaths to DCF supply chainsPolicy and practices lay the groundwork for DCF disclosureRecommendations for improved DCF disclosureAnnex 1:Criteria for high-quality DCF responsesAnnex 2:Allowable certification schemes providing DCF assuranceAnnex 3:Data table 030710131924383537384Time for Transparencycompanies disclosed on at least one commodity through CDPs forests questionnaire.881made sufficiently comprehensive and high-quality disclosure using appropriate methods such as certification and monitoring systems to determine DCF status.of those companies responded to questions asking about DCF performance.445510621%Companies that produce or source agricultural or forestry commodities must eliminate deforestation and conversion of other ecosystems from their supply chains to meet near-term climate and nature targets as well as comply with emerging regulatory requirements.CDPs forests questionnaire has been tracking companies progress toward eliminating commodity-driven deforestation for over a decade.In 2023,CDP introduced new indicators developed in partnership with the Accountability Framework initiative(AFi)to facilitate clearer disclosure of performance and progress towards deforestation-and conversion-free(DCF)supply chains.Standardized DCF indicators allow companies progress toward deforestation-and conversion-free production and sourcing to be assessed in a comparable and easy to interpret way by stakeholders,including buyers,investors,financiers and civil society.The data provides those stakeholders with information they need to make informed decisions about their purchasing,financing and advocacy.This report provides a detailed examination of the responses provided in 2023 by companies to those DCF indicators.It provides a baseline view of companies current capacity to understand and control deforestation and ecosystem conversion associated with their operations and supply chains,and provides recommendations for how companies and others can support improved reporting.Executive summary1,152companies reported on their management of deforestation during the previous year.In 2023,These figures indicate that many companies are working to meet these new disclosure expectations while others are not yet able or willing to provide this information.5Time for TransparencyThis represents 7%of disclosing companies,demonstrating both that achievement and disclosure of DCF supply chains is possible but that it is still uncommon.64companies made a high-quality disclosure and reported that at least one commodity supply chain was 100forestation-or conversion-free.98 companies made a high-quality disclosure and reported that at least one commodity supply chain was less than 90F.Of those,27 companies reported that less than 20%of their volumes were DCF for at least one commodity.While these disclosures show that achieving DCF supply chains may still take time,it is encouraging to see companies willingness to publicly disclose their performance to stakeholders.This information can help stakeholders both support company improvement and make informed decisions about their own sourcing or investments.Companies that disclosed on DCF supply chains tended to have no-deforestation or no-conversion policies or commitments,and to report engaging their suppliers.Of 445 companies that responded to the DCF question for at least one commodity:69%had a DCF policy or commitment,compared to 33%of companies that did not disclose DCF progress;82%reported engaging their suppliers,compared to 49%of companies that did not disclose DCF progress;and 50%reported engaging in landscape and jurisdictional initiatives1,compared to 19%that did not disclose DCF progress.This suggests that having policies in place can lay the groundwork for disclosing on DCF performance,and that supplier or landscape engagement may support supply chain monitoring and management.It may also indicate that DCF disclosure via CDPs forests questionnaire is at this time limited largely to companies that have greater maturity on forest and supply chain action.1 187 out of 370 companies disclosing through the full-tier questionnaire20F217 high-quality disclosures of DCF progress made by186 companies20-89F90-99F100F29814166Timber productsSoyPalm oilCattle productsCocoaCoffeeRubber311596311High-quality disclosures indicating 100F supply chains:6Time for TransparencyThe most common issue with DCF disclosure was the use of certification models that do not provide sufficient DCF assurance.Of the 445 companies that disclosed on DCF progress:44%relied on certification models that do not provide sufficient assurance of deforestation-and conversion-free volumes,including mass balance chain-of-custody models;36%were missing key information,or information was not consistently disclosed;and 12%excluded significant volumes,products,activities,suppliers,or regions from their total production or sourcing,and therefore DCF volumes reported were not representative of the total commodities produced or sourced.In addition,many responses indicated a poor or incomplete understanding of the capabilities of risk assessment and monitoring tools with regard to both level of rigor regarding DCF assurance and the ecosystem types included in analyses.To better account for and communicate deforestation-and conversion-free production and sourcing,companies should:Respond comprehensively and accurately to CDP and other disclosure requests,regardless of the amount of progress that has been made.Communicate intentions to achieve deforestation-and conversion-free supply chains,both publicly via policies and commitments,and to suppliers via engagement and support.Fully understand the capabilities of different certification schemes and chain-of-custody models and how they can be applied to demonstrate DCF status.Consider impacts on all natural ecosystems,not only forests,when setting,monitoring and disclosing on DCF commitments.Adopt an informed approach to selecting and using risk assessment systems to determine if they can effectively ensure that materials produced in specified sourcing areas are free of both deforestation and ecosystem conversion.Understand and disclose on highly transformed commodities in their supply chains,especially soy embedded in animal product supply chains.Introduction:A time of rapid action for deforestation-andconversion-free supply chains17Time for Transparency8Time for TransparencyDeforestation and ecosystem conversion are the most significant impacts of agricultural commodity and forestry production on our planet.Land clearance for agriculture accounts for more than 10%of human-caused greenhouse gas emissions2 and is associated with at least a third of global biodiversity loss3.Expanding agricultural commodity and forestry production also impacts the rights of Indigenous Peoples and local communities,including land rights and access to resources.Companies that produce or source agricultural or forestry commodities are receiving ever-clearer mandates from their buyers,investors and regulators to eliminate deforestation,ecosystem conversion,and associated human rights abuses from their supply chains and to be transparent about their progress.Introduction:A time of rapid action for deforestation-andconversion-free supply chainsEmergence of near-term targets and regulationsIn recent years,sustainability goals related to supply chain impacts have transformed into the need for immediate action to meet near-term targets for climate and nature,as well as emerging regulatory demands.These include:Eliminating deforestation from key commodities by 2025 and eliminating all land use change from supply chains by 2030 to achieve emissions reduction targets in line with a 1.5C pathway,as required by the Science Based Targets initiative(SBTi).Eliminating deforestation and ecosystem conversion associated with commodity production,to meet conservation goals laid out in the Convention on Biological Diversity and implemented through the Science Based Targets Networks(SBTN)Land Targets.Requirements that companies that sell agricultural commodities into the EU market demonstrate deforestation-free origins to meet the European Unions Deforestation Regulation,which goes into effect in 2025.These targets and deadlines mean that buyers,investors and other stakeholders need information about company progress toward eliminating deforestation and ecosystem conversion from their supply chains,with little time left before key milestones are in the rearview mirror.Comprehensive disclosure is now a minimum expectation for companies that produce or source agricultural and forestry commodities,and reporting platforms and standards are now available to capture and organize that information.2 IPCC Sixth Assessment Report,20233 IBPES Global Assessment Report on Biodiversity and Ecosystem Services,2019Comprehensive disclosure is now a minimum expectation for companies that produce or source agricultural and forestry commodities.9Time for TransparencyA standardized approach to deforestation-and conversion-free disclosureIn 2023,for the first time,companies were able to disclose comprehensively on deforestation-and conversion-free(DCF)commodity volumes(see Box 1)through CDP using a set of standardized metrics developed through a collaborative process led by the Accountability Framework initiative(AFi).This same set of metrics has also been incorporated into other reporting standards such as the Global Reporting Initiative,industry association protocols such as the Consumer Goods Forums Forest Positive Coalition KPIs,and other assessment tools.These metrics provide a consistent and comprehensive way for companies to disclose performance and progress towards DCF supply chains.Using them to guide disclosure allows companies to disclose the DCF status of 100%of the agricultural or forestry commodity volumes that they produce or source,broken down by the method used to assess or verify DCF status.These indicators are complemented by indicators related to engagement with suppliers and in sourcing regions to eliminate deforestation,conversion,and human rights abuses.Disclosure using these metrics enables companies to communicate performance and progress to their buyers,investors and stakeholders in a clear and credible way.It also demonstrates leadership necessary to support and enable sector-wide change.Box 1:What are deforestation-and conversion-free commodity volumes?Agricultural and forestry commodities are considered deforestation-and conversion-free(DCF)when they can be shown to originate on production units(such as farms,ranches or forests)on which conversion from forests or other natural ecosystems to cropping systems,pastures or plantations has not occurred after a specified cutoff date.Deforestation:Loss of natural forest as a result of:(i)conversion to agriculture or other non-forest land use;(ii)conversion to a tree plantation;or(iii)severe and sustained degradation.Conversion:Loss of a natural ecosystem as a result of its replacement with agriculture or another land use,or due to a profound and sustained change in a natural ecosystems species composition,structure,or function.Source:Accountability FrameworkThese metrics provide a consistent and comprehensive way for companies to disclose performance and progress towards DCF supply chains.2023 CDP forests disclosure210Time for Transparency11Time for TransparencyCDPs data provides a yardstick to measure achievement against the consensus-based principles and guidance set out in the Accountability Framework.In 2023,the number of companies disclosing on sustainable commodity production and sourcing through CDP increased for the seventh consecutive year,providing valuable insights on company policies,practices and performance.Figure 1.Commodity disclosures by companies through CDPs forests questionnaire in 2020 to 20232023 CDP forests disclosureIn 2023,1,152 companies reported through CDP on their management of deforestation,conversion and restoration during the previous year.Of these,232 companies reported that they did not produce,source or use any of the seven high-risk commodities identified by CDP in the reporting year,and 39 companies made disclosures associated with mining projects.The remaining 881 companies disclosed on at least one of the seven high-risk commodities responsible for most agriculture-related deforestation and conversion,for a total of 1,498 commodity-specific disclosures(Figure 1).The most reported commodities in 2023 were timber products,followed by palm oil,soy,cattle products,natural rubber,cocoa,and coffee(Figure 1;Table 1).These proportions have been consistent over recent years.Table 1.Commodity disclosures and companies disclosing on each commodity CommodityTimber productsPalm oilSoyCattle productsCocoaRubberCoffee#of disclosures650304194162716453%of disclosures43.4 .3.0.8%4.7%4.3%3.5%of companies disclosing on each commodity73.84.5.0.4%8.1%7.3%6.0 2002004006008001000120014001600CoffeeCocoaRubberSoyCattle productsPalm oilTimber products202120222023881companies disclosed on at least one of the seven high-risk commodities responsible for most agriculture-related deforestation and conversion,for a total of 1,498 commodity-specific disclosures.12Time for TransparencyFigure 2.Proportion of disclosing companies by value chain stage(n=881).Reporting companies included 349 organizations based in Europe,303 from Asia,285 from the US and Canada,191 from Latin America,13 from Oceania and 11 from Africa.There was a decrease in reporting from North American companies in 2023,while disclosures from Latin America increased by almost a third on recent years.Disclosures from Oceania and Africa remained low.Companies disclosing through CDP in 2023 produced or sourced significant amounts of the global production of four key commodities(Table 2).Consumption data refers to volumes of commodity that were sourced or purchased by the company in raw or processed forms.With most disclosers coming from value chain stages further downstream,most of the coverage is accounted for under consumption.The exception to this is in palm oil,where both production(volumes produced on land owned or managed by the company)and consumption are well represented.This is largely due to a relatively high proportion of Indonesian and Malaysian palm oil refiners disclosing through CDP.Table 2.Global commodity production and volumes reported through CDPCommodityCommodity volume produced globally in 20224(millions of tons)Disclosed production volumes(millions of tons)Disclosed consumption volumes(millions of tons)Cattle products100.34.224.4Palm oil 88.85 29.149.4Soy384.52.1103.6Timber products4,415.0256.6537.80202!00%ProductionProcessingTradingManufacturingRetailingThe largest share(60%)of disclosing companies engaged in manufacturing,while fewer than 13%produced raw commodities(Figure 2),with many companies operating in more than one capacity.This data therefore presents more information about companies further downstream in the supply chain,and less inference can be drawn about those further upstream.4 Food and Agriculture Organization of the United Nations timber,crop and livestock production 2022.5 Latest available data for global palm production covering 2021.Disclosure of progress toward DCF goals3 13Time for Transparency14Time for TransparencyIn 2023,companies were requested to report the proportion of volumes in their operations and supply chains that they considered to be deforestation-and/or conversion-free(DCF)during the previous reporting year.To be able to consider commodity volumes as DCF,companies must have been able to determine that materials did not originate from production units where conversion from forests or other natural ecosystems occurred after a specified cut-off date(see Box 1).Disclosure of progress toward DCF goalsTable 3.Criteria for determining high-quality responses to questions on DCF progressCriteria for high-quality responseDetailsDCF disclosure was credible and well explainedThe company used one or more tools or methodologies to support its DCF disclosure and:described the process;and/or described how it resulted in credible/consistent characterizations of risk;and/or the methodology or outcomes were verified.DCF disclosure was comprehensiveThe company reported no significant exclusions from its disclosure.The exclusions that were accepted in this years analysis were:volumes less than 5%of the commodity total;reporting limited to own-brand products;embedded soy;or mergers or acquisitions in 2022.Certifications used to substantiate DCF status were suited to this purposeDCF claims were made based on certification schemes with robust DCF criteria and chain-of-custody models that provide physically certified volumes.Metrics for DCF disclosure and assessment of responsesCompanies were asked to disclose the percentage of reported volumes verified as deforestation-and/or conversion-free(DCF),in relation to the full volume of each commodity that the company produced or sourced in the reporting period.Companies were given the option of identifying one or more approaches to assessing or verifying DCF volumes,including:1.Volumes demonstrated as DCF based on origination from areas with no or negligible risk of deforestation or conversion.2.Volumes demonstrated as DCF based on monitoring of the location where the commodity originated.3.Volumes demonstrated as DCF through physical certification.Responses to questions about DCF production and sourcing were assessed to determine whether information was disclosed in line with guidance published by CDP and AFi,and whether the company provided appropriate evidence to substantiate the status of volumes reported as DCF.Table 3 provides a summary of the criteria used to distinguish high-quality DCF responses;the full set of criteria can be found in Annex 1.15Time for TransparencyResponse rates to questions on DCF volumesHalf of disclosing companies(445/881)responded to questions about their deforestation-and conversion-free commodity production or sourcing for at least one commodity.These DCF responses were included in 43%(638/1,498)of commodity disclosures submitted(Figure 3).Response rates for the DCF portion of the questionnaire were the highest for palm oil 55%of palm oil disclosures included DCF responses(168/304)followed by 49%of timber disclosures(318/650).Response rates to DCF questions were the lowest for rubber,with only 8 of 64 companies(13%)disclosing DCF progress(Figure 3).Figure 3.Total disclosures,disclosures to questions on progress toward deforestation-and/or conversion free supply chains,and high-quality DCF disclosures 16245304168646503189864194707153105Commodity disclosuresDCF responsesHigh-quality DCF responsesSoyPalm oilTimber productsCocoaRubberCoffeeCattle products8319111422Commodity disclosuresDCF responsesHigh-quality DCF responsesSoyPalm oilTimber productsCocoaRubberCoffeeCattle products831911142216Time for TransparencyQuality of responses to questions on DCF volumesWhile half of companies responded to questions about their DCF commodity production or sourcing,only 21%(186/881)presented information that was clear and comprehensive enough to be considered high-quality for at least one commodity(see Table 4 and Annex 1 for criteria for high-quality responses).While companies disclosing on cocoa and coffee had the lowest response rates to the DCF questions,the DCF responses that were disclosed were the most likely to be high-quality,with 58%of cocoa(11/19)and 50%of coffee(5/10)responses being high-quality(Figure 3).Palm oil saw the highest proportion of high-quality disclosures in relation to total disclosures(21%,64/304).38%of DCF responses related to palm oil were categorized as high-quality(64/168),a higher proportion than for timber,soy,and cattle disclosures.Two-thirds of responses to DCF questions were determined to have serious issues that undermined the reliability or interpretability of the information disclosed(421/638).The most common problems related to the use of certification systems;nearly 200 companies had at least one response to DCF questions rated as not high-quality because the certification models disclosed did not provide sufficient assurance of DCF status.Another common concern with the DCF responses involved companies claiming that materials were DCF due to sourcing from a broad geographic region such as the US,UK,or EU without further information about the methodologies used to make those determinations or the types of ecosystems threatened in those regions.In addition,more than a quarter of DCF responses(181/638)included missing or inconsistent information across different response fields.Finally,10%of DCF responses(63/638)had significant exclusions in the scope of disclosures relative to the companys total production or sourcing.1/3of responses to DCF questions(217/638)were considered to be high-quality DCF disclosures,representing 14%of total commodity disclosures(217/1,498).Overall17Time for Transparency20F20-89F100F90-99F13%indicated that the relevant supply chain was less than 20F.37%indicated that it was between 20-89F.19%indicated that it was between 90-99F.30%indicated that the company had achieved 100forestation-free and/or conversion-free production or sourcing for that supply chain.Of the 217 high-quality disclosures of DCF progress made by 186 companies:Degree of progress disclosedResponding companies were at many different stages of maturity in achieving DCF supply chains(Figure 4).Figure 4.Number of high-quality DCF disclosures indicating progress toward DCF supply chainsNumber of disclosures4035302520151050High-quality disclosures with 20FHigh-quality disclosures with 20-89FHigh-quality disclosures with 90-99FHigh-quality disclosures with 100FCattle productsCocoaCoffeePalm oilRubberSoyTimber products18Time for TransparencyHalf of the companies disclosing 100F supply chains(32/64)identified as retailers or manufacturers only.Fifteen producers disclosed 100F production,nine of which were for timber.Companies were encouraged to disclose the volumes they had determined to be DCF even if the percent they reported as deforestation-and/or conversion-free was low,with a focus on transparency rather than performance.Of 186 companies making a high quality DCF disclosure,98 reported that at least one commodity supply chain was less than 90F.Of those,27 companies reported that less than 20%of their volumes were DCF.Cocoa,palm oil and rubber companies were the most likely to disclose data for supply chains for which the proportion of DCF material was lower.Of high-quality palm oil disclosures,69%(44/64)indicated that supply chains were less than 90F.Far fewer cocoa and rubber companies disclosed through CDP and responded to DCF questions,but 73%of high-quality cocoa disclosures(8/11),and 67%of high-quality rubber disclosures(2/3)indicated less than 90F supply chains.Conversely,nearly 60%of high-quality soy(13/22)and timber(57/98)disclosures,and 71%of high-quality cattle disclosures(10/14),indicated supply chains that were at least 90F.While in some cases this indicates strong progress,in others it may be evidence of continued reluctance by many companies to disclose in advance of significant progress in these sectors.64companies reported at least one commodity supply chain as 100forestation-or conversion-free via a high-quality DCF disclosure.Timber productsSoyPalm oilCattle productsCocoaCoffeeRubber311596311High-quality disclosures indicating 100F supply chains:Paths to DCF supply chains419Time for Transparency20Time for TransparencyResponses indicate that few companies use only one method of DCF monitoring and assurance for all of their supply chain volumes.Rather,companies use a collection of approaches,resulting in a diverse set of contextualized or overlapping supply chain control mechanisms(see Box 2 for examples).Paths to DCF supply chainsDiverse approaches to achieve DCF volumesAs a result of these overlapping approaches to assessing DCF volumes as well as the structure of the 2023 questionnaire,there was a great deal of variability in the way this information was reported,and therefore a high level of uncertainty in the data.However,some broad patterns can be seen.Refinements being made in CDPs 2024 questionnaire are intended to support companies in disclosing more clearly(see Box 4 for more information).Of 551 commodity-level DCF responses that provided sufficiently interpretable information:At least half indicated the use of certification for at least some amount of DCF volumes;At least a third indicated DCF volumes arising from sourcing areas with no or negligible deforestation risk;and At least a quarter indicated that DCF volumes were verified through direct monitoring of production units.Disclosures on cattle had the most homogenous approach;at least 70%of DCF disclosures on cattle claimed DCF volumes based on sourcing from jurisdictions without deforestation/conversion risk.Cattle supply chains were also the least likely to use certification to determine DCF volumes,with fewer than 10%of disclosures indicating the use of certification for this purpose.70%of DCF disclosures on cattle claimed DCF volumes based on sourcing from jurisdictions without deforestation/conversion risk.At least21Time for TransparencyThis text is taken or adapted from public CDP disclosures.It is drawn from self-reported information and has not been verified by CDP or AFi.Anonymised responses in this section have been selected as examples of clear DCF disclosure,they are not an endorsement of a companys overall performance.Methodology:Certification negligible risk sourcing53%of soybeans used by the company in Europe are grown in France or Italy,and the remaining 47%come from Canada.100%of these soybeans are ProTerra Segregated certified.Soybeans used by the company in North America are grown in the United States,with a very small percentage from Canada.90.5%is Proterra Segregated certified and the remaining 9.5%is with non-GMO Project verified or organic certifications providing chain-of-custody guarantee.In December 2022,a third-party verification process was set up:with the support of an external commodities consultancy,the company put in place a new traceability process for key forest-risk commodities including soy.The purpose of this traceability process is to track,monitor and verify the volumes,origin,certification status,and deforestation and conversion risk of these key commodities provided to the company by its suppliers.Box 2Examples from disclosures using a range of approaches to DCF assessmentCompany 1:100F Soy Company 2:75F Leather Methodology:Direct monitoring negligible risk sourcingThe company maintains a list of low-,medium-and high-risk countries for leather sourcing,which includes information on deforestation and conversion risk,alongside other environmental,human rights and animal welfare criteria.The company does not source leather from high-risk countries.For leather coming from low-or medium-risk countries they use the following approaches:1.Traceability to the slaughterhouse in a low-risk subnational area and alignment with the companys Standards requirements on supply chain transparency&traceability;OR2.Traceability to the slaughterhouse(geo-referenced boundaries)and to direct and indirect farms that the slaughterhouse has purchased from(geo-referenced boundaries),and a verification of no deforestation or conversion after a cut-off date of January 1,2020;OR 3.Traceability to the slaughterhouse(geo-referenced boundaries)and audit of DCF safeguards to verify DCF compliance within their entire supply chain(direct and indirect farms).This may include a combination of traceability systems with segregated certification ensuring DCF compliance.22Time for TransparencyCompany 3:96F Palm oil Company 4:99F Timber products Methodology:Direct monitoring negligible risk sourcingThe companys cut-off date is December 31,2015.Raw materials are assessed as deforestation-free when they can be traced either to low-risk origins or have been assessed as deforestation-free either from the sky or from the ground.Assessed from the sky means that volumes have been assessed through satellite monitoring of production sites in our supply chain identified through traceability(89%of volumes in 2022).Assessed on the ground means that volumes have been assessed through on-the-ground assessments,including by High Carbon Stock Approach and High Conservation Value assessments,by our partners(eg Earthworm Foundation,Proforest,SGS)and/or through certification such as the RSPO.Only segregated volumes are accepted as deforestation-free(6%of volumes in 2022).Traceable to low-risk origin means that volumes have been traced back to regions classified as at low risk of deforestation,using tools such as Maplecroft(1%of volumes in 2022).Methodology:Certification20212022In 2021,the company achieved a significant milestone by attaining 100%availability of FSC-certified forests and other controlled sources.This means that all products delivered to customers were accompanied by FSC CoC third-party certification,ensuring compliance with the FSCs requirement of zero harvesting from areas undergoing conversion to plantations or non-forest use.In 2022,1%was sourced from Russia during a period when FSC certificates were suspended.Company policies require all our paper suppliers to comply with either the FSC Controlled Wood Standard or the FSC Forest Management Standards,assessed by third-party independent auditing and verification.23Time for TransparencyMany tools for identifying DCF sourcing areasClaims about DCF sourcing arising from jurisdictions or other spatially defined sourcing origins with negligible deforestation or conversion risk were common across many commodities.However,only 88 disclosures identified classification systems that were used to determine negligible risk of deforestation and/or conversion.Generally little information was provided about how these systems were used,but responses show that companies are getting information about deforestation and conversion risk from a wide variety of sources.Of those 88 companies:To support future high-quality and consistent disclosure,CDPs 2024 corporate questionnaire will request additional detail on criteria used for area-level DCF determinations.For the purposes of this report,a number of those disclosures were considered to be high quality,provided that they sufficiently detailed the sourcing locations in open text responses,even if they did not provide detailed disclosure on the classification system or process used to determine low or negligible risk.As scrutiny of these disclosures increases,companies will need to provide greater detail about the methodologies used and the assurance that these tools provide(see Section 6 for further recommendations).To support future high-quality and consistent disclosure,CDPs 2024 corporate questionnaire will request additional detail on criteria used for area-level DCF determinations.Companies will be expected to provide both the classification methodology used and the specific origins classed as having negligible risk.disclosed using certification for this process,primarily for timber products,with FSC and PEFC the most common.used commercial tools,with Maplecroft the most common.12 Global Forest Watch Pro 9 Preferred by Nature risk toolsused international indexes,including FSCs National Risk Assessment and Transparency Internationals Corruption Perceptions Index.29121424reported using publicly available tools.Policy and practices lay the groundwork for DCF disclosure524Time for Transparency25Time for TransparencyPolicy and practices lay the groundwork for DCF disclosureThe findings in this report show that leading companies can add and are adding DCF disclosure to an existing set of good practices to address deforestation and conversion in their supply chains(see Box 3).Companies that disclosed on progress toward achieving deforestation-and conversion-free supply chains were more likely to have policies and practices in place that communicated and advanced those goals.This suggests that strong policies and practices can support and enable companies to determine and report their DCF performance.In the coming year,the expectation would be that companies engaging in these good practices will see and disclose significant improvements in their DCF sourcing figures.These findings may also indicate that DCF disclosure via CDP forests was limited largely to companies that have greater maturity on forest and supply chain action.Therefore,the information gathered this year about DCF performance and approaches is likely not representative of all disclosing companies,but rather reflects the more advanced subset of disclosers that chose to respond to questions on DCF performance.Box 3:DCF disclosure as part of a suite of good practicesOf the 217 high-quality disclosures of DCF progress made by 186 companies:94%indicated having a traceability system in place for at least some of their supply chain.92%included using third-party certification for at least some of their volumes.The only commodities for which this number was not at least 90%were soy(68%used certification)and cattle(57%).62%indicated use of a risk classification system.26Time for TransparencyOverall,half of companies disclosed having a no-deforestation or no-conversion policy or commitment(437/881).A further quarter did not have DCF policies or commitments but had other types of forests-related policies or commitments,for example commitments to legality with regard to commodity production(211/881).Companies that disclosed on DCF progress for a given commodity were more than twice as likely to have a no-deforestation or no-conversion policy for that commodity than companies that did not disclose on DCF progress.Of companies that disclosed on DCF progress,69%had no-deforestation or no-conversion policies or commitments for the given commodity(308/445),compared to only 33%of companies that did not disclose on DCF progress(172/516).Figure 5.Response rate to DCF questions by companies with and without DCF commitments or policiesSimilarly,companies with DCF policies and commitments were far more likely to disclose their DCF progress and to disclose it in a high-quality way(Figure 5).Of companies with a DCF policy or commitment,70%responded to the DCF questions for at least one commodity(308/437),compared to 23%of companies without DCF policies or commitments(65/286).Nearly one third of companies with DCF policies or commitments had high-quality DCF responses for at least one commodity(132/437),compared to only 9%of those without forest-related policies(27/286).This trend was consistent across all commodities,indicating that strong no-deforestation and no-conversion policies are closely associated with robust transparency on supply chain progress.In addition,of 126 disclosures that contained a timebound public DCF commitment and high-quality DCF disclosure,99 of them(79%)had a cutoff date of 2020 or earlier associated with that commitment,indicating that these commitments are in line with good practice.DCF policies tend to precede DCF disclosureCattle productsCocoaCoffeePalm oilRubberSoyTimber products100pP0 %0%of companies with DCF policies/commitments that disclose on DCF volumes%of companies without forest related policies/commitments that disclose on DCF volumes27Time for TransparencyFigure 6.Supplier engagement disclosed by companies that did or did not respond to DCF questionsOverall,65%of companies(575/881)report conducting some form of supplier engagement,either with smallholders or with direct or indirect suppliers.This is highest for palm oil disclosures(65%of disclosures;198/304)and timber disclosures(62%;402/650)and lowest for rubber(31%;20/64)and cocoa(39%;28/71).Across commodities,companies that disclose on DCF progress are more likely to disclose some form of supplier engagement(Figure 6).Of the companies that disclosed DCF progress for a commodity,82%(363/445)reported supplier engagement associated with that commodity,compared to 49%among the companies that did not disclose DCF progress(252/516).Of companies with high-quality DCF Companies that disclose on DCF progress are more likely to disclose supplier engagementdisclosure,all five companies disclosing on coffee and all but one of 22 companies disclosing on soy report supplier engagement.In addition,23%of companies responding to CDPs full-tier forests questionnaire(122/542)reported engaging non-compliant suppliers.This proportion was nearly three times higher for companies that disclosed on their DCF progress(29%)than those that did not(11%).Finally,50%of full-tier companies disclosing their DCF progress reported engaging in landscape and jurisdictional initiatives,compared to 19%of companies that did not disclose DCF progress.Cattle productsCocoaCoffeePalm oilRubberSoyTimber products100pP0 %0Supplier engagement among companies with a high-quality DCF disclosureSupplier engagement among companies not making a DCF disclosureRecommendations for improved DCF disclosure628Time for Transparency29Time for TransparencyRecommendations for improved DCF disclosureFor most companies,disclosure of progress toward eliminating deforestation and ecosystem conversion from their operations and supply chains is a relatively new expectation.Responses to the 2023 CDP forests questionnaire indicate both that many companies are working to assess and communicate the extent to which they have succeeded in this goal and that there is still a need for companies to better understand the approaches and tools that can be used to provide assurance of DCF materials.Companies should respond comprehensively and accurately to CDP and other disclosure requests,regardless of the amount of progress that has been made.Indicatorsdeveloped bythe Accountability Framework initiative arenow availableacrossseveral leading reporting and assessmentplatforms,including CDP,providinga clear and consistent way for companies to disclose performance and progress towardsDCF supply chains.Company disclosures following these indicators will enable stakeholders to understand and recognize progress as companies advance towards a range of climate and nature goals.All companies that produce or source agricultural or forestry commodities should disclose this information,regardless of their position in the supply chain or level of progress towards DCF supply chains.To prepare for effective disclosures,companies should review new and existing guidance on these topics.That includes materials from the AFi and CDP as well as other relevant sources.Companies should attend CDPs introductory and questionnaire change webinars in 2024 so that they can be aware of new questionnaire content and structure and disclose effectively in the coming year.1Responses to the 2023 CDP forests questionnaire indicate both that many companies are working to assess and communicate the extent to which they have succeeded in this goal and that there is still a need for companies to better understand the approaches and tools that can be used to provide assurance of DCF materials.30Time for TransparencyBox 4:A new format for the CDP questionnaireIn 2024,CDP has combined the three existing questionnaires into one CDP Corporate Questionnaire,so that companies asked to respond across multiple environmental issues can do so in a single place.Integration of the questionnaires follows the latest science,aligns with new high-quality disclosure frameworks and standards,and includes incremental changes to the data points in previous climate change,forests and water security questionnaires.CDPs integrated corporate questionnaire structureFor 2024,the updated forests module will focus on disclosures relevant to the companys production and use of commodities,their commitments towards eradicating deforestation and natural ecosystem conversion,the DCF status of the total commodity volumes they handle,and the actions that they are taking to progress towards sustainable sourcing and restoration within and beyond their supply chains.All companies disclosing on soy,timber,palm or cattle will receive a single forests score,while coffee,rubber and cocoa will not continue not to be scored in 2024.The updated question structure will distinguish between the methods that support DCF claims and those that indicate companies are still working towards DCF.The questionnaire and scoring have been adapted to better accommodate DCF disclosures where overlapping risk,certification and monitoring methods have been used-acknowledging that overlapping methods may be needed to provide DCF assurance.For the latest information about the integrated questionnaire,please visit the CDP website.Module 1IntroductionModule 2Identification,assessment&management of dependencies,impacts,risks,&opportunitiesModule 3Disclosure of risks&opportunitiesModule 4GovernanceModule 5Business strategyModule 6Environmental performance Consolidation approachModule 7Environmental performance Climate changeModule 8Environmental performance ForestsModule 9Environmental performance Water securityModule 10Environmental performance PlasticsModule 11Environmental performance BiodiversityModule 12Environmental performance Financial servicesModule 13Environmental performance Further information&sign offCross-issue modulesCross-issue modulesSector specific moduleEnvironment issue-specific modules31Time for TransparencyCompanies should communicate intentions to achieve deforestation-and conversion-free supply chains,both publicly via policies and commitments and to suppliers via engagement and support.Corporate policies and commitments to eliminate deforestation and ecosystem conversion from commodity supply chains provide an essential baseline for action and progress toward achieving these goals.They communicate a companys intentions and approaches to buyers,investors,civil society and the public.They also support internal buy-in to take action to manage supply chains to reduce deforestation and conversion,and enable disclosure of progress.Companies should therefore set or strengthen policies and commitments in alignment with the Accountability Framework.In addition,managing supply chains to both monitor and address deforestation and conversion requires engagement with suppliers,including communication of policies and procurement criteria and support to enable suppliers to achieve compliance with the policies.Companies that disclose DCF progress are far more likely to engage their suppliers on deforestation and conversion,indicating that implementation and transparency are complementary actions.Companies should better understand the capabilities of certification programsThe most common cause of DCF responses not meeting the quality criteria set out in this report was reliance on certification models that do not provide assurance of DCF status.Only certification systems that have robust deforestation-and/or conversion-free criteria can be used to support DCF claims.In addition,only chain-of-custody models that allow products to be physically traced to certified product units such as segregated or identity-preserved models can be used to claim volumes as DCF.Other forms of certification,such as mass balance,require additional monitoring and due diligence processes to ensure products are DCF5.While CDP will provide clearer guidance on this in the 2024 questionnaire(see Annex 2),companies should familiarize themselves with the capabilities and shortcomings of the certification programs with which they work.This may entail researching the principles,criteria and chain-of-custody models of various certification schemes,or directly contacting the organizations that administer these schemes to gather more information.5 Note that the FSC Controlled Wood standard ensures that even non-certified materials are deforestation-free,providing this additional due diligence and supporting DCF claims.23Corporate policies and commitments to eliminate deforestation and ecosystem conversion from commodity supply chains communicate a companys intentions and approaches to buyers,investors,civil society and the public.32Time for TransparencyCompanies should consider additional ecosystems besides forests when monitoring and disclosing on DCF commitmentsCurrent CDP disclosures indicate that companies have commitments and policies that represent a mix of deforestation-free and DCF goals.Companies that have or are moving towards more comprehensive DCF targets should ensure that systems for monitoring compliance with those targets at both the level of the production unit and the level of the sourcing area are suitable to assess conversion of grasslands,savannahs and wetlands,in addition to forests.Certification systems identified by CDP as providing DCF assurance already include assessment of non-forest ecosystem conversion(see Annex 2).In some key commodity origins,such as many countries in South America,data and tools are already available for monitoring compliance with DCF commitments.In other regions,effective risk assessment and monitoring tools may still be in development,and companies should work to identify tools and methodologies that are available in their sourcing regions.Companies should adopt a more informed approach to selecting and using risk assessment systems and indices Risk assessment tools and indices are used for many purposes,including identifying the most significant impacts of company activities,prioritizing action where most needed,and determining appropriate levels of due diligence.The use of risk-based approaches to determine whether sourcing areas,such as jurisdictions,can be considered deforestation-and/or conversion-free requires a higher level of scrutiny than other uses of risk assessments.Companies should therefore ensure they are selecting and using risk assessment systems that can effectively ensure that materials produced in specified sourcing areas are free of both deforestation and ecosystem conversion.Some tools consider deforestation only,without including risk of conversion of other ecosystems such as grasslands or wetlands.For example,indices that show that the US is deforestation-free for cattle may not consider other ecosystem types,such as grasslands;therefore cattle sourced from the US could not be shown to be DCF.Companies using this approach to DCF assessments should therefore gather further information about the methodologies used to make those determinations and the types of ecosystems threatened in those regions before using them as the basis for determining DCF status of sourcing areas or commodity volumes.4533Time for TransparencyCompanies should understand and disclose on highly transformed commodities in their supply chains,especially soy embedded in animal products For companies that purchase or source animal products or manufactured goods,comprehensive disclosure should include commodity products that are highly transformed,including soy and palm derivatives and soy fed to animals.Beginning in 2024,the CDP corporate questionnaire will place specific focus on soy embedded within animal products to ensure impacts are transparently disclosed.New questions about embedded soy volumes will allow companies to detail the actions they are taking and progress they are making against those volumes,in addition to direct commodity sourcing.Embedded soy enters a companys supply chain indirectly as animal feed used in the production of animal products such as meat,farmed fish,dairy,eggs or other animal products that a company sources or uses as an ingredient.Companies can estimate the amount of embedded soy in their supply chain using published methodologies such as those recommended by the Consumer Goods Forum.They can also estimate potential soy origins and DCF status using trade data or supply chain mapping tools as they work to achieve further traceability.Clear disclosure of the level of traceability and DCF progress for embedded volumes will provide company stakeholders with more information about company exposure to soy-related risks.6Annexes34Time for Transparency35Time for TransparencyAnnex 1:Criteria for high-quality DCF responsesHigh-quality DCF disclosureComplete and consistent with deforestation free guidanceNot high-quality DCF disclosureIncomplete or inconsistent with guidanceDescriptionGenerally clear,comprehensive,with a plausibly robust methodology to ensure at a minimum no-deforestation.Company has used one or more tools or methodologies and:Describes the process and how it results in credible/consistent characterization of risk.Methodology/outcomes were verified.Only recycled materials were used.Unclear or not comprehensive including any of the following issues:Significant exclusions.Methodological issues.Incomplete disclosure.CoverageOnly negligible or small exclusions reported(=5%),or:Specific products.DCF volumes exclude or are limited to certain tiers of the supply chain.Certain geographies or products are excluded.Determination risk States risk tool or method used and describes process and/or results.Method or outcomes were verified,including certification.Sourcing cattle or soy from EU,US,CAN or NZ.Reports volumes as negligible risk but does not use a tool or classification system,explain method or detail outcome including listing sourcing areas determined to be of negligible risk.Only assesses risk in tropical forest countries(classing temperate/boreal forests as low risk).Determination certificationStates certification providing assurance of no deforestation or no conversion status via physical certification(see list in Annex B)or addresses limitations with certifications in one or more of the following ways:Sourcing mass balance certified product for which the mix is a known percentage,and claiming that minimum percentage as DCF.Use of additional tools such as NDPE/IRF profiles in the delivering category to verify DCF volume.Use of complimentary monitoring to assure DCF status of mass balance volumes eg monitoring mill sourcing radius.Certification does not have robust DCF criteria or volumes are not physically certified eg credit/offset or mass balance certifications and no additional traceability and verification undertaken.36Time for TransparencyHigh-quality DCF disclosureComplete and consistent with deforestation free guidanceNot high-quality DCF disclosureIncomplete or inconsistent with guidanceDetermination monitoringStates verification method or tool used and:Describes process and or results.Method or outcomes were verified.Monitoring direct suppliers and engaging in LA/JAs.Uses third party monitoring method that assures deforestation or conversion free eg FSC Controlled Wood methodology used and verified to monitor non-certified timber volumes.Reports volumes as verified but does not explain tool or process used(eg satellite monitoring,farm-level monitoring).Does not report independent verification.Data quality Percentages in please explain are consistent with the%of reported volume verified as deforestation-and/or conversion-free data disclosed and with certification and monitoring data points.Describes a clear or plausible methodology for identifying or verifying deforestation-or conversion-free status.Missing or inconsistent information,response is not clear enough to make an assessment of exclusion or methodological validity.Information provided is not consistent between fields.Information reported in question F1.5 is not consistent with certification or monitoring data provided in F6.Certification schemeAccepted chain of custody modelAssuranceBiosuisse organicIdentity preserved/segregatedDFDonau SojaSegregatedDCFEurope SojaSegregatedDCFFSCAll modelsDCFISCCIdentity preserved/segregatedDFNaturlandSegregatedDCFProTerra certificationIdentity preserved/segregatedDFRainforest Alliance Sustainable Agriculture StandardIdentity preserved/segregatedDCFRSB Global FuelsIdentity preserved/segregatedDFRSPOIdentity preserved/segregatedDFRTRSSegregatedDCFSoil Association Organic Farming&Growing(GB and Northern Ireland)SegregatedDCFSustainable Biomass ProgramSegregatedDF37Time for TransparencyAnnex 2:Allowable certification schemes providing DCF assuranceAnnex 2 summarizes the certification schemes that were judged to provide credible deforestation-or conversion-free assurance.The list has been produced using desk-based research and consultation.SoyCoffeeTimberCocoaRubberPalm oilCattle products38Time for TransparencyAnnex 3:Data tableNote:data on non-compliant supplier engagement and jurisdictional or landscape engagement were collected for only a subset of companies,and those datapoints are not included here.Number of companiesCattle productsCocoaCoffeePalm oilRubberSoyTimber productsTotal companies disclosing on at least one commodityTotal commodity-level disclosuresDisclosing on a commodity1627153304641946508811498Disclosing as producers10351471176113126Disclosing as processors481211501244167280344Disclosing as traders226641924124188232Disclosing as manufacturers79421820737116356533855Disclosing as retailers822434652045219262489Responding to DCF questions451910168870318445638Not responding to DCF questions117524313656124332516860With high-quality DCF disclosure141156432298186217With high-quality DCF disclosure reporting DCF volumes 201423142729With high-quality DCF disclosure reporting DCF volumes between 20-89414006277781With high-quality DCF disclosure reporting DCF volumes between 90-992504264041With high-quality DCF disclosure reporting DCF volumes of 100c11519316466That have no-deforestation or no-conversion policies or commitments7233241652192322437729That have other forest-related(non-DCF)policies or commitments215559926150211275That have no-deforestation or no-conversion policies or commitments and respond to DCF questions36158122557219308462That do not have no-deforestation or no-conversion policies or commitments and respond to DCF questions3322426386578That have no-deforestation or no-conversion policies or commitments and make a high-quality DCF disclosure11945421859132157Reporting supplier engagement9328241982099402575864Responding to DCF questions and reporting supplier engagement39139138357250363509With high-quality DCF disclosure and reporting supplier engagement9755412172147169Not responding to DCF questions and reporting supplier engagement541515601742152252355With non-high-quality DCF disclosures3185104548220330421With non-high-quality DCF disclosures and issues with certifications used6337421111198218With non-high-quality DCF disclosures and missing,inconsistent or unclear information18632942794161181With non-high-quality DCF disclosures and large exclusions421017395263Disclosing DCF production/consumption volume from areas with no or negligible risk of deforestation/conversion3011533593142177Disclosing DCF production/consumption volume verified through monitoring systems1261452367133154Disclosing DCF production/consumption volume physically certified310581433165227301With high-quality DCF disclosures that use certification81056331595176199With high-quality DCF disclosures that use a classification system to determine sourcing area risk962432153488111With high-quality DCF disclosures that have a traceability system in place141055932092175203AuthorsContributorsAbout CDPCDP is a global non-profit that runs the worlds environmental disclosure system for companies,cities,states,and regions.Founded in 2000 and working with over 700 financial institutions representing more than US$142 trillion in assets.CDP pioneered using capital markets and corporate procurement to motivate companies to disclose their environmental impacts,and to reduce greenhouse gas emissions,safeguard water resources and protect forests.Over 24,000 organizations around the world disclosed data through CDP in 2023,including more than 23,000 companies worth two thirds global market capitalization,and over 1,100 cities,states,and regions.Fully TCFD aligned,CDP holds the largest environmental database in the world,and CDP scores are widely used to drive investment and procurement decisions towards a zero carbon,sustainable and resilient economy.CDP is a founding member of the Science Based Targets initiative,We Mean Business Coalition,The Investor Agenda,and the Net Zero Asset Managers initiative.Visit or follow us CDP to find out more.About the Accountability Framework initiativeAbout the Accountability Framework initiativeThe Accountability Framework initiative(AFi)is a collaborative programme to protect forests,other natural ecosystems,and human rights by making ethical production and trade the new normal.To achieve this critical transformation,the AFi promotes and supports implementation of the Accountability Framework,a detailed roadmap for setting goals,taking action,and reporting progress towards ethical supply chains.The initiative is led by the AFi Coalition:a diverse group of environmental and human rights organisations from around the world that developed the Accountability Framework and work to drive positive impact in the agriculture and forestry sectors.Visit accountability-framework.org or email contactaccountability-framework.org to find out more.DISCLOSURE INSIGHT ACTIONLeah SambergLead Scientist,AFiTomasz SawickiHead of Land,CDPViera UkropcovaManager,CDPAbigail DrabickSenior Project Officer,CDPAlona RivordCommunications Lead,AFiAndre SocratesProject Manager,CDPDavid KosciulekSenior Engagement Officer,CDP CDP 2024Jeffrey MilderDirector,AFiJennifer ToesCommunications Associate,AFiMaddy BraveryCommunications,CDPNiall RobbTechnical Manager,AFiThomas MaddoxDirector of Nature,CDP
2024-05-28
39页
5星级
carbonpricingSTATE AND TRENDS OF2024WORLDBANK.ORG2 State and Trends of Carbon Pricing 2024 2024 Inte.
2024-05-27
76页
5星级
CENTER FOR DATA INNOVATION 1 Rethinking Concerns About AIs Energy Use By Daniel Castro|January 29,2024 Concerns about the energy used by digital technologies are not new.Near the peak of the dot-com boom in the 1990s,a Forbes article lamented,“Somewhere in America,a lump of coal is burned every time a book is ordered online.”1 The authors of the article,which became widely cited in subsequent years in debates about energy policy,estimated that“half of the electric grid will be powering the digital-Internet economy within the next decade.”2 However,the estimate was wrong,with errors in both its facts and methodology.3 In hindsight,there is no longer any dispute,as the International Energy Agency(IEA)estimates that todays data centers and data transmission networks“each account for about 11.5%of global electricity use.”4 This mistake was not an isolated event.Numerous headlines have appeared over the years predicting that the digital economys energy footprint will balloon out of control.5 For example,as the streaming wars kicked off in 2019with Apple,Disney,HBO,and others announcing video streaming subscription services to compete with Netflix,Amazon,and YouTubemultiple media outlets repeated claims from a French think tank that“the emissions generated by watching 30 minutes of Netflix is the same as driving almost 4 miles.”6 But again,the estimate was completely wrong(it is more like driving between 10 and 100 yards),resulting from a mix of flawed assumptions and conversion errors,which the think tank eventually corrected a year later.7 With the recent surge in interest in artificial intelligence(AI),people are once again raising questions about the energy use of an emerging technology.In this case,critics speculate that the rapid adoption of AI CENTER FOR DATA INNOVATION 2 combined with an increase in the size of deep learning models will lead to a massive increase in energy use with a potentially devastating environmental impact.8 However,as with past technologies,many of the early claims about the consumption of energy by AI have proven to be inflated and misleading.This report provides an overview of the debate,including some of the early missteps and how they have already shaped the policy conversation,and sets the record straight about AIs energy footprint and how it will likely evolve in the coming years.It recommends that policymakers address concerns about AIs energy consumption by taking the following steps:Develop energy transparency standards for AI models.Seek voluntary commitments on energy transparency for foundation models.Consider the unintended consequences of AI regulations on energy use.Use AI to decarbonize government operations.THE FACTS ABOUT AIS ENERGY USAGE AND CARBON EMISSIONS Creating accurate estimates of the energy use and carbon emissions of AI systems over their lifetimes is challenging because these calculations depend on many complex factors,including details about the chips,cooling systems,data center design,software,workload,and energy sources used for electricity generation.This problem is not unique to AI.As a group of energy researchers described the problem in an article in the Annual Review of Energy and the Environment:Creating credible estimates of electricity requirements for information technology is fraught with difficulty.The underlying data are not known with precision,the empirical data are limited,the most useful data are often proprietary,and the technology is changing so rapidly that even accurate data are quickly obsolete.9 However,several studies have attempted to quantify the current and future energy demands and carbon emissions of AI systems.Unfortunately,some of the initial estimates have fallen into the same trap as past early studies about the energy use of digital technologies and have produced misleading estimates.These studies generally consider the energy needed for an AI system over its lifetime in two stages:1)training the AI model;and 2)using the AI model to respond to specific queriesa process called“inference.”Training AI Models Researchers at the University of Massachusetts Amherst estimated in 2019 the carbon emissions of several AI models,one of the first major studies of its kind.10 The study found that BERTwhich at the time was Googles state-of-the-art large language model(LLM)emitted CENTER FOR DATA INNOVATION 3 approximately 1,438 pounds of carbon dioxide(CO2)during 79 hours of training using 64 advanced graphics processing units(GPUs),the chips commonly used for training AI models because of their superior parallel processing capabilities.To put this in perspective,a roundtrip flight from New York to San Francisco creates approximately 2,000 pounds of CO2 emissions per passenger.The researchers also estimated carbon emissions for training an AI model for neural architecture search(NAS),a technique for automatically finding one or more neural network architectures for a given taskone of the most computationally complex problems in machine learning.Specifically,they evaluated the energy usage of a NAS used to create a better English-German machine translation model.11 The researchers estimated that training the model in question generated 626,155 pounds of CO2 emissions(roughly equivalent to 300 roundtrip flights from the East Coast to the West Coast).12 Not surprisingly,given journalistic tendencies to skew toward negative coverage of tech,virtually all the headlines in the popular media focused on this latter estimate despite its narrow use case.13 Even respected scientific news outlets such as MIT Technology Review ran such headlines as“Training a single AI model can emit as much carbon as five cars in their lifetimes.”14 These articles suggested that the massive energy needed to train this particular AI model was normal despite this estimate clearly referring to an atypical example.It would be like an automotive news outlet running an article that suggested“driving a car emits as much carbon as an airplane”based only on a study that looked at the environmental impact of a flying car prototype.Moreover,both the original research paper and the subsequent news articles often noted that while the large AI model outperformed existing ones at language translation benchmarks,the improvements were only marginal.The implication was that AI researchers are making trivial performance improvements at the expense of non-trivial amounts of carbon emissions.Indeed,other AI researchers made this point explicit in a widely read paper“On the Dangers of Stochastic Parrots:Can Language Models Be Too Big?”15 They argued that it is“environmental racism”for wealthy Western nations to deploy ever-larger AI models because these AI systems will have negative impacts on poor communities in the Global South.Specifically,they wrote:Is it fair or just to ask,for example,that the residents of the Maldives(likely to be underwater by 2100)or the 800,000 people in Sudan affected by drastic floods pay the environmental price of training and deploying ever larger English language models,when similar large-scale models arent being produced for Dhivehi or Sudanese Arabic?16 Given the chargesthat training AI systems is not only dangerous for the environment but also an overt act of racismit is not surprising that many policymakers have raised questions about AIs energy consumption.However,the headline-making estimate in the 2019 study was wildly CENTER FOR DATA INNOVATION 4 incorrectjust like many prior claims about the oversized energy footprint of digital technologies.The University of Massachusetts Amherst researchers had made several false assumptions that grossly inflated their estimates both for the total energy used and the carbon emissions.In response to the 2019 study,the researchers involved in the NAS model provided a detailed summary of the energy use and carbon emissions from their work,noting why the outside researchers estimates were wrong.The actual emissions were 88 times smaller than the earlier studys estimate.17 Unfortunately,the popular media paid little attention to correcting the record or noting the new findings,and so the initial impressions have lived on.Researchers have published multiple studies in subsequent years estimating the energy needed to train many well-known AI models as well as their carbon emissions.As shown in table 1,while larger models generally require more energy usage than smaller ones do,the exact figures vary significantly across different AI models.For example,researchers estimate that training GPT-3the 175 billion parameter AI model used in the popular ChatGPT applicationcreated 552 tCO2 emissions,but comparable AI models including OPT(a 175 billion parameter AI model created by Meta)and Gopher(a 280 billion parameter AI model created by Google)have significantly smaller carbon footprints.Moreover,the efficiency of training AI models continues to improve.For example,18 months after GPT-3,Google produced GLaM,an LLM with 1.2 trillion parameters.Despite GLaM being nearly 7 times larger than GPT-3 and outperforming the other AI model,GLaM required 2.8 times less energy to train.18 Finally,the energy mix used to power the data center where developers train an AI model impacts its carbon emissions.For example,the developers of BLOOM used a French data center powered by nuclear energy,which reduced its carbon footprint.19 Despite the new research,groups critical of AI have repeatedly cited the initial incorrect study in their demands for policymakers to reduce investment in large-scale computing resources.For example,the American Civil Liberties Union(ACLU)sent a letter to the Office of Science and Technology Policy(OSTP)in October 2021 complaining about the“environmental costs”of the White Houses planned National AI Research Resource(NAIRR)and arguing that“the NAIRR should focus on offering an alternative to the data-and compute-hungry applications that are the focus of many industry and research labs.”20 Similarly,the Center for AI and Digital Policy falsely claimed in 2022 that“AI-enabled systems require exponentially rising computing power.This increase in computing power requires substantial energy consumption,generating a huge carbon footprint and upending the green effects of digitalization.”21 In each case,they made these claims despite overwhelming evidence showing they were misleading and overblown.CENTER FOR DATA INNOVATION 5 Table 1:Estimated energy demand of training various AI models Model#of Parameters Chips (model x#)Hours Energy(MWh)CO2 Emissions(metric tons)Estimate Source BERT 0.1B V100 x64 79 1.5 0.7 Strubell et al.,201922 GPT-2 1.5B TPUv3x32 168 1.7*0.7*Strubell et al.,201923 Llama 2 7B A100 x(n/a)N/A 74*31.2 Meta,202324 Llama 2 13B A100 x(n/a)N/A 147*62.4 Meta,2023 Llama 2 70B A100 x(n/a)N/A 688*291.4 Meta,2023 LaMDA 137B TPUv3x1024 1,385 451 26 Thoppilan et al.,202225 GPT-3 175B V100 x10000 355 1,287 552 Patterson et al.,202126 OPT 175B A100 x992 N/A N/A 75 Zhang et al.,202227 BLOOM 176B A100 x384 2,820*433 24.7 Luccioni et al.,202228 Gopher 280B TPUv3x4096 920 1,151*380 Rae et al,202229 PaLM 540B TPUv4x6144 TPUv4x3072 1200 326 3,436*271.4 Chowdery et al.,202230 GLaM 1,162B TPUv4x(n/a)N/A 456 40 Patterson et al.,202231 GPT-4 1,800B A100 x25000 2280 N/A N/A Walker,202332*Inferred based on available data,see appendix for details Using AI Models Despite the attention from policymakers and the media on the energy costs of training AI models,multiple studies have concluded that most of the CENTER FOR DATA INNOVATION 6 energy costs associated with AI systems come from using AI modelsa process known as“inference”(because the model is inferring results based on a given input).For example,Amazon Web Services estimates that 90 percent of the cost of an AI model comes from inference.33 Similarly,a study from Schneider Electric estimates that 80 percent of the AI workload in data centers in 2023 is from inference and 20 percent is for training.34 Finally,a study by researchers at Meta notes that the exact breakdown between training versus inference varies across use cases.For LLMs,they estimate that inference is associated with 65 percent of the carbon footprint,but for recommendation models where parameters must be updated frequently based on new data,they estimate an even split between training and inference.35 Multiple factors impact the amount of energy used during inference,including the type of task and the AI model.As shown in table 2,the energy requirements for inference can vary significantly by task.For example,using an AI model to classify text is generally computationally less intensive(and thus uses less energy)than using AI to generate an image.36 Different AI models also have different energy costs,and within specific models(e.g.,Llama 2 7B versus Llama 2 70B),a larger number of parameters generally requires more energy for inference.Table 2:Average energy use per 1,000 queries by task37 Task kWh Text classification 0.002 Image classification 0.007 Object detection 0.038 Text generation 0.047 Summarization 0.049 Image generation 2.907 Given that training a particular AI model incurs a one-time cost,whereas using an AI model continues to consume energy over time,it makes sense that most of the energy used for AI will eventually come from inference.It also means that the energy requirements for running AI models will have a significant impact on the overall energy use for AI systems.While most critics have focused on the energy used to train AI models,some people have expressed concern about the energy used during inference.38 For example,writing in the October 2023 edition of the journal Joule,one researcher estimated that interacting with an LLM requires approximately CENTER FOR DATA INNOVATION 7 10 times as much energy as conducting a typical web search query,and extrapolated from that estimate to conclude that“the worst-case scenario suggests Googles AI alone could consume as much electricity as a country such as Ireland(29.3 TWh per year).”39 There are many reasons to doubt that such a“worst-case”scenario is on the near horizon.In 2022,Googles total global energy consumption across the entire company was 21.8 TWh.40 For the worst-case prediction to be true,Googles energy use for AI alone would have to more than exceed its current total global energy use.It is true that the companys energy consumption has grown over time,particularly from its data centers,as its business has grown.For example,Googles data centers used about 3 TWh more electricity in 2022 than the year before.41 But while its overall energy usage has grown,for the three years between 2019 and 2021,the proportion of energy it used for machine learning remained constantbetween 10 to 15 percent of its total energy consumptionwith approximately 60 percent of that used for inference.42 One explanation for the relatively constant proportion of energy used for inference is the improvements seen in AI models and hardware.Indeed,as shown in table 3,both performance and efficiency tend to improve over time.The table shows that over a few years,the accuracy of computer vision AI models improved significantly.In addition,the energy requirements for inference across these models generally decreased with the release of a newer chip.As noted in one recent study of the energy used for inference in AI models,“when a SOTA state-of-the-art model is released it usually has a huge number of FLOPs floating point operations,and therefore consumes a large amount of energy,but in a couple of years there is a model with similar accuracy but with much lower number of FLOPs.”43 In other words,the newest AI models may not be particularly efficient by design because researchers are focusing on performance improvements,but over time,researchers will address efficiency.Table 3:Energy consumption for inference of deep neural networks used for computer vision for two different GPUs44 Year Top-1 Accuracy(ImageNet)P100,released June 2016(Joules)V100,released June 2017(Joules)AlexNet 2012 56.52 0.033 0.023 GoogLeNet 2014 69.77 0.077 0.055 Vgg16 2014 71.59 0.542 0.373 ResNet50 2015 75.30 0.179 0.132 CENTER FOR DATA INNOVATION 8 WHAT AI ENERGY FORECASTS GET WRONG One reason forecasts about future energy demands from AI are so high is they use inaccurate or misleading measurements,as described previously.Another reason is the forecasts ignore the practical economic and technical realities that come with widespread commercialization of AI.The Energy Use of AI Is Limited by Economic Considerations Many of the high-end estimates for AI energy use are impractical because of the costs involved.Buying more chips,building more data centers,and powering those data centers is expensive.For example,as even the author of the prediction that Googles AI alone might consume 29.3 TWh annually admitted,reaching this level would require a$100 billion investment in chips along with billions more in operating costs for the data center and electricity.45 Even large tech companies would find it unsustainable to pay for such massive amounts of computing.Businesses are profit-seeking enterprises and computing costs money;therefore,they are not going to offer services for long that cost more to operate than they receive in revenue.Either the energy costs for using AI will come down or how companies deploy AI will be limited by cost factors.The Rate of Performance Improvements in AI Will Decline Over Time AI models have improved significantly in the past few years.For example,OpenAIs LLM model,GPT-4,released in March 2023,can pass many popular exams designed for humans,such as the SAT,GRE,LSAT,and AP tests for a variety of subjects.46 These results are a substantial improvement over its earlier model released the prior year.While AI still cannot perform many tasks as well as humans,such as abstract reasoning,now that some AI models perform so highly on many benchmarks,there is substantially less opportunity for improvement in certain domains.As a result,many developers will likely focus more on optimizing their AI models rather than squeezing out ever-smaller improvements in accuracy because they will not receive a return on investment for building and operating larger models.Future Innovations Will Improve AIs Energy Efficiency The history of computing is one of continuous innovation,and these innovations extend to energy efficiency.For example,over the past decade,demands on global data centers have increased substantially even as the energy intensity of data centers has decreased by approximately 20 percent annually.47 Between 2010 and 2018,there was a 550 percent increase in compute instances and a 2,400 percent increase in storage capacity in global data centers,but only a 6 percent increase in global data center energy use.48 These energy efficiency gains came from improvements in hardware,virtualization,and data center design,and they are part of the reason that cloud computing has been able to scale.CENTER FOR DATA INNOVATION 9 Similar trends are already appearing in AI.As one recent paper notes,“Many studies report that the size of neural networks is growing exponentially.However,this does not necessarily imply that the cost is also growing exponentially,as more weights could be implemented with the same amount of energy,mostly due to hardware specialization but especially as the energy consumption per unit of compute is decreasing.”49 Improvements in hardware and software will likely keep the pace of energy growth from AI in check.Chipmakers continue to create more efficient GPUs for AI.For example,Nvidias recent transition from one generation of GPUs to another resulted not only in significantly faster processing but also nearly doubled energy efficiency.50 Likewise,researchers continue to experiment with techniques,such as pruning,quantization,and distillation,to create more compact AI models that are faster and more energy efficient with minimal loss of accuracy.51 These types of advancements are one reason the proportion of energy Google uses for AI has remained constant in recent years despite machine learning growing to account for 70 to 80 percent of the compute used at the company.52 Indeed,as one researcher succinctly put it,“AIs energy consumption is not skyrocketing,contrary to commonly expressed fears.”53 AIs Energy Footprint Ignores Substitution Effects Discussing the energy usage trends of AI systems can be misleading without considering the substitution effects of the technology.Many digital technologies help decarbonize the economy by substituting moving bits for moving atoms.For example,sending an email replaces mailing a letter,streaming a movie replaces renting a DVD,and participating in a video conference replaces traveling to an in-person meeting.AI will have a similar impact over time,both by further digitalizing many activities(such as by improving the quality of video calls)and by using AI to complete tasks more efficiently than using human labor.One study in 2023 estimates the carbon footprint of using AI versus using a human for writing a page of text or creating an illustration.After considering the carbon emissions for different AI models(ChatGPT,BLOOM,Midjourney,and DALLE-2)and comparing workers in the United States and India,the researchers found“AI writing a page of text emits 130 to 1,500 times less CO2e than a human doing so”and“AI creating an image emits 310 to 2,900 times less.”54 Of course,since using AI does not eliminate humansthey still exist and eat,breathe,etc.using AI does not eliminate these carbon emissions.But AI does eliminate the carbon emissions from the devices humans use for these tasks,such as laptop or desktop computers.As shown in table 4,these savings can be substantial;however,there are limits to generalizing these findings.For example,by making it easier to produce text and images,the volume of activity may increase.Nevertheless,these findings show how,holding all else equal,using AI to substitute for human labor can reduce carbon emissions in certain cases.CENTER FOR DATA INNOVATION 10 Table 4:Carbon footprint(grams CO2e)for using a human(in the United States)versus AI(BLOOM/Midjourney)for certain tasks AI Laptop Desktop Human Writing a page of text 0.95 27 72 1,400 Creating an image 1.90 100 280 5,500 HOW AIS ENERGY USE FITS INTO THE BIGGER PICTURE The debate about AIs energy use is part of a larger debate about how to address global climate change.Within that context,there are important factors policymakers should keep in mind.AI Will Play an Important Role in Addressing Climate Change There are many opportunities to use AI to reduce carbon emissions,support clean energy technologies,and address climate change.These opportunities span multiple industries,including the transportation,agriculture,and energy sectors.For example,AI is crucial for integrating renewable energy sources such as wind and solar into the electric grid by using data points to forecast supply and demand.Likewise,utilities are using AI for predictive maintenance of energy assets,managing and controlling grids,and setting dynamic pricingall critical elements for an efficient electric grid.55 AI can also help make sense of complex climate data from sensors and satellites,such as changing sea levels,surface temperatures,and rainfall,to create better forecasts and address risks of climate change.For example,AI can detect methane emissions from satellite data,allowing regulators to more effectively monitor industry.56 Similarly,farmers can use AI for precision agriculture,reducing their use of fertilizer and water and their associated environmental costs.57 Already businesses,governments,and consumers are using AI to operate more efficiently.AI is a key part of creating smart cities that use AI to operate efficient buildings,roads,waterways,and more.58 For example,in California,the government is using AI to monitor over a thousand cameras to detect and respond to wildfires quickly,reducing carbon emissions that come from these fires.59 And AI can enable firms to optimize industrial processes,reduce waste,and use energy more efficiently,thereby reducing their carbon intensity.60 For example,logistics providers use AI to optimize delivery routes,thereby reducing fuel consumption of their fleets.61 And in the consumer space,tools such as the Nest smart thermometers saved customers 113 billion KWh between 2011 and 2022 and more efficient driving from Google Maps has reduced carbon emissions by 1.2 million metric tons.62 CENTER FOR DATA INNOVATION 11 As AI matures,policymakers should continue to look to the technology as a key tool for addressing climate change.There Is No Unique Market Failure for AIs Energy Use Solving the global climate challenge will require transitioning to clean energy technologies that have a price and performance on par with dirty ones.63 In the interim,any activity that uses energy has an environmental impact,and AIs use of energy is no different.However,there are no unique market failures associated with AIs use of energy that would lead to greater environmental impact than alternative uses would.A kilowatt-hour used for AI is no different than a kilowatt-hour used for watching television,microwaving popcorn,powering lights,or any other activity.Indeed,as noted previously,in many cases,AI applications will be used as a substitute for less energy-efficient activities and to address climate change.In both cases,those who consume energy must pay for it,and rational actors will generally seek to minimize these costs.While such costs may not include negative externalities associated with energy use,that problem is not unique to AI and cannot be addressed for AI alone.Large Tech Companies Have Made Bold Net-Zero Commitments Large tech companies are at the forefront of AI,and these are the same companies that have made some of the boldest commitments among corporations to reducing their carbon footprints.Consider the following:Google(now Alphabet)became carbon neutral in 2007,the first major company to do so.64 A decade later,it became the first major company to purchase enough renewable energy to match its electricity consumption.65 And,in 2020,it purchased carbon offsets to eliminate the companys entire carbon legacy.66 The company continues to press forward and has committed to operating all its data centers and campuses on carbon-free energy by 2030.67 Amazon co-founded The Climate Pledge in 2019,whereby companies commit to net-zero carbon emissions by 2040,10 years ahead of the Paris Agreement.68 And,in 2022,Amazon reported that 90 percent of the electricity it consumed came from renewable sources,and it was on track to reach 100 percent by 2025,five years ahead of its goal of 2030.69 Microsoft has committed to be carbon negative by 2030 and eliminate its companys entire carbon legacy by 2050.70 It has also committed to using 100 percent renewable energy by 2025.71 Facebook(now Meta)reached zero carbon emissions in its direct operations(i.e.,data centers and offices)in 2020 and has pledged to reach net-zero emissions across its entire value chain by 2030.72 CENTER FOR DATA INNOVATION 12 These companies have remained publicly committed to these pledges even as they lead in the development and deployment of AI.Indeed,their net-zero pledges are one of the reasons these companies must carefully consider the efficiency of the AI models they train and deploy.HOW POLICYMAKERS SHOULD ADDRESS AIS ENERGY USE Witnessing the rapid advancements in AI,policymakers around the world are considering whether and how they should regulate the technology,including its energy usage.For example,UNESCOs“Recommendation on the Ethics of AI”which was adopted by 193 member states in November 2021states,“Member States and business enterprises should assess the direct and indirect environmental impact throughout the AI system life cycle,including,but not limited to,its carbon footprint,energy consumption and reduce the environmental impact of AI systems and data infrastructures.”73 The impact of AI on energy and the environment should be part of the policy debate,but policymakers should also be careful not to overreact,especially given the prevalence of misleading narratives falsely depicting AIs energy consumption as out of control.There are reasonable steps policymakers can take to ensure AI is part of the solution,not part of the problem,when it comes to the environment.To that end,policymakers should do the following:Develop Energy Transparency Standards for AI Models It is usually easier to manage things that can be measured,and energy use of AI models is no different.While many AI developers have begun publishing model cardsshort documents that accompany the release of an AI model that detail information about its performance,limitations,and other relevant informationthese do not always contain information about the energy used to train or use them.74 When they do include information about the environmental impact,they tend to focus on the carbon emissions from training rather than the energy needs for inference.75 This focus on the energy used for training is partially out of necessity,as the developers of a model cannot control the hardware others might run their model on in the future or specific use cases.But the amount of energy used for training does not necessarily impact the energy that will be required for inference.Therefore,choosing models based on the amount of energy used to train them rather than the life cycle energy costs could lead to less-efficient outcomes.To address this problem,policymakers should support the development of energy transparency standards for AI models,both for training and inference.In the United States,for example,the National Institute of Standards and Technology should work with the Department of Energy to develop a recommended best practice for assessing the training and inference energy costs.For example,this standard might include a set of benchmark tests and hardware to give comparable energy performance CENTER FOR DATA INNOVATION 13 metrics across different models.The United States should also work with the G7 and the Organization for Economic Cooperation and Development(OECD)to ensure broad adoption of these energy transparency standards to avoid different disclosure practices in different jurisdictions,especially since some countries might make such disclosures mandatory.Seek Voluntary Commitments on Energy Transparency for Foundation Models While developing transparency standards for AI models will help,it will also be important for leading AI companies to adopt these standards and disclose this information publicly.The White House has proactively sought out and obtained voluntary commitments from most of the leading U.S.-based AI companies to promote“safe,secure,and transparent development and use of generative AI(foundation)model technology.”76 While these commitments included important pledges from companies to engage in extensive testing to detect vulnerabilities and promises to avoid discrimination and bias in their models,they did not include any commitments around energy.To address that shortcoming,the White House should continue its dialogue with these companies to seek a voluntary commitment to publicly disclose the energy required to train and operate these foundation models,as well as the associated carbon emissions,especially for cloud-based AI service providers.Making this information publicly available will give users of foundation models the option to take the environmental footprint of AI into consideration when deciding which AI services to use.Consider Unintended Consequences of AI Regulations on Energy Use Many policymakers have called on developers to ensure their AI models minimize bias,avoid hate speech,limit disclosure of private information,and align to other,often worthwhile,goals.In many cases,developers are actively working to create models and build safeguards to address these concerns because they have strong market incentives to do so.However,policymakers rarely consider that their demands can raise the energy requirements to train and use AI models.For example,debiasing techniques for LLMs frequently add more energy costs in the training and fine-tuning stages.77 Similarly,implementing safeguards to check that LLMs do not return harmful output,such as offensive speech,can result in additional computing costs during inference.78 Thus,many of the proposed mandates for AI models could come at the expense of energy efficiency goals.The converse is also true:Mandates for energy-efficient AI models could create trade-offs that result in AI models that are less fair and more biased than they otherwise might be.The point is not that policymakers should never regulate any AI system,but rather that they should avoid rushing to regulate until they fully understand the implications of their decisions.For example,the EUs AI Act initially included no requirements around energy efficiency.However,in response CENTER FOR DATA INNOVATION 14 to some of the misleading claims about AIs environmental impact,the European Parliaments proposed revisions to the legislation included substantial additions around energy,such as directing AI developers to integrate“state-of-the art methods and relevant applicable standards to reduce the energy use,resource use and waste,as well as to increase their energy efficiency and the overall efficiency of the system.”79 These requirements were in tension with other obligations in the AI Act to eliminate bias from AI models.While the AI Act now only includes more reasonable energy transparency requirements,the proposal from the European Parliament shows the potential for bad facts to lead to bad policy.Use AI to Decarbonize Government Operations AI offers important opportunities to improve the quality and efficiency of many government services,and adopting AI broadly across government agencies at every level should be a key priority for policymakers.In addition,AI can help the public sector reduce carbon emissions through more efficient digital services,smart cities and buildings,intelligent transportation systems,and other AI-enabled efficiencies.At the 2022 United Nations Climate Change Conference of the Parties(COP27),the United States launched the Net-Zero Government Initiative,which commits national governments to reaching net-zero carbon emissions for their operations by 2050.80 To accelerate the use of AI across government agencies toward this goal,the president should sign an executive order directing the Technology Modernization Funda relatively new funding system for federal government IT projectsto include environmental impact as one of the core priority investment areas for projects to fund.In addition,the United States should invite and share best practices for using AI in government from the other countries that are part of the Net-Zero Government Initiative.CONCLUSION Policymakers need accurate information about the energy implications of AI.Unfortunately,groups that oppose AI,whether from honest misunderstanding of the evidence or intentional cherry-picking of the facts,continue to push the narrative that AIs energy footprint is growing out of control.In December 2023more than two years after the record was correcteda columnist writing in The Guardian repeated the original false and misleading statistic about AIs energy impact that has generated so much concern.The article stated:A study in 2019,for example,estimated the carbon footprint of training a single early large language model(LLM)such as GPT-2 at about 300,000kg of CO2 emissionsthe equivalent of 125 round-trip flights between New York and Beijing.Since then,models have become exponentially bigger and their training footprints will therefore be proportionately larger.81 CENTER FOR DATA INNOVATION 15 Just as the early predictions about the energy footprints of e-commerce and video streaming ultimately proved to be exaggerated,so too will those estimates about AI likely be wrong.But given the enormous opportunities to use AI to benefit the economy and societyincluding transitioning to a low-carbon futureit is imperative that policymakers and the media do a better job of vetting the claims they entertain about AIs environmental impact.CENTER FOR DATA INNOVATION 16 APPENDIX Conversions and inferences shown for Table 1.All figures from source cited in table,unless otherwise noted.BERTBERT 0.65 metric tons CO2e=1,438 lbs GPTGPT-2 2 1.7 MWh=32 GPUs x 168 hours x 289W*x 1.1 PUE*Using measurements from LaMDA.See Thoppilan et al.,2022.0.7 metric tons CO2e=1.7 MWh x 0.429 CO2e/KWh*Using measurement of U.S.average data center net CO2e/KWh from Patterson et al.,2021.Llama 2Llama 2 74 MWh=184,320 GPU hours x 400W 147 MWh=368,640 GPU hours x 400W 688 MWh=1,720,320 GPU hours x 400W BLOOMBLOOM 2,820 hours=1,082,990 million GPU hours/384 GPUs GopherGopher 1,151 MWh=4,096 GPUs x 920 hours x 283W x 1.08 PUE PaLMPaLM 3,436 MWh=(6,144 GPUs x 1,200 hours) (3,072 GPUs x 336 hours)x 378.5W x 1.08 PUE CENTER FOR DATA INNOVATION 17 REFERENCES 1.Peter W.Huber and Mark Mills,“Dig more coal the PCs are coming,”Forbes,May 31,1999,https:/ Koomey et al.,“Sorry,Wrong Number:The Use and Misuse of Numerical Facts in Analysis and Media Reporting of Energy Issues,”Annual Review of Energy and the Environment 27,no.1(November 1,2002):11958,https:/doi.org/10.1146/annurev.energy.27.122001.083458.4.Vida Rozite,“Data Centres and Data Transmission Networks,”IEA,July 11,2023,https:/www.iea.org/energy-system/buildings/data-centres-and-data-transmission-networks.5.Colin Cunliff,“Beyond the Energy Techlash:The Real Climate Impacts of Information Technology,”Information Technology and Innovation Foundation,July 6,2020,https:/itif.org/publications/2020/07/06/beyond-energy-techlash-real-climate-impacts-information-technology/.6.George Kamiya,“Factcheck:What is the carbon footprint of streaming video on Netflix?”Carbon Brief,February 25,2020,https:/www.carbonbrief.org/factcheck-what-is-the-carbon-footprint-of-streaming-video-on-netflix/.7.Ibid.8.Alex De Vries,“The Growing Energy Footprint of Artificial Intelligence,”Joule 7,no.10(October 1,2023):219194,https:/doi.org/10.1016/j.joule.2023.09.004.9.Koomey et al.,“Sorry,Wrong Number:The Use and Misuse of Numerical Facts in Analysis and Media Reporting of Energy Issues.”10.Emma Strubell,Ananya Ganesh,and Andrew McCallum,“Energy and Policy Considerations for Deep Learning in NLP,”arXiv.org,June 5,2019,https:/arxiv.org/abs/1906.02243.11.David R.So,Chen Liang,and Quoc V.Le,“The Evolved Transformer,”arXiv.org,January 30,2019,https:/arxiv.org/abs/1901.11117.12.Strubell,Ganesh,and McCallum,“Energy and Policy Considerations for Deep Learning in NLP.”13.Doug Allen and Daniel Castro,“Why So Sad?A Look at the Change in Tone of Technology Reporting From 1986 to 2013”(Information Technology and Innovation Foundation,February 2017),https:/itif.org/publications/2017/02/22/why-so-sad-look-change-tone-technology-reporting-1986-2013/.14.Karen Hao,“Training a Single AI Model Can Emit as Much Carbon as Five Cars in Their Lifetimes,”MIT Technology Review,December 7,2020,https:/ M.Bender et al.,“On the Dangers of Stochastic Parrots,”FAccT 21:Proceedings of the 2021 ACM Conference on Fairness,Accountability,and Transparency,March 2021,61023,https:/doi.org/10.1145/3442188.3445922.CENTER FOR DATA INNOVATION 18 16.Ibid.17.David A.Patterson et al.,“Carbon Emissions and Large Neural Network Training,”arXiv.org,April 21,2021,https:/doi.org/10.48550/arxiv.2104.10350.18.David S.Patterson et al.,“The Carbon Footprint of Machine Learning Training Will Plateau,Then Shrink,”IEEE Computer 55,no.7(July 1,2022):1828,https:/doi.org/10.1109/mc.2022.3148714.19.Melissa Heikkil,“Were Getting a Better Idea of AIs True Carbon Footprint,”MIT Technology Review,November 15,2022,https:/ for Information on an Implementation Plan for a National Artificial Intelligence Research Resource,”American Civil Liberties Union,October 1,2021,https:/www.aclu.org/wp-content/uploads/document/ACLU_Response_to_NAIRR_RFI.pdf.21.“Response to the RFI request on the National Artificial Intelligence Research and Development Strategic Plan,”Center for AI and Digital Policy,March 4,2022,https:/www.caidp.org/app/download/8378181763/CAIDP-Statement-OSTP-03042022.pdf.22.Strubell,Ganesh,and McCallum,“Energy and Policy Considerations for Deep Learning in NLP.”23.Ibid.24.“Llama 2 Model Details,”Facebook Research,October 14,2023,https:/ Thoppilan et al.,“LAMDA:Language Models for Dialog Applications,”arXiv.Org,January 20,2022,https:/doi.org/10.48550/arxiv.2201.08239.26.Patterson et al.,“Carbon Emissions and Large Neural Network Training.”27.Susan Zhang et al.,“OPT:Open Pre-Trained Transformer Language Models,”arXiv.Org,May 2,2022,https:/doi.org/10.48550/arxiv.2205.01068.28.Alexandra Sasha Luccioni,Sylvain Viguier,and Anne-Laure Ligozat,“Estimating the Carbon Footprint of BLOOM,a 176B Parameter Language Model,”arXiv.Org,November 3,2022,https:/doi.org/10.48550/arxiv.2211.02001.29.Jack W.Rae et al.,“Scaling Language Models:Methods,Analysis&Insights from Training Gopher,”arXiv.Org,December 8,2021,https:/doi.org/10.48550/arxiv.2112.11446.30.Aakanksha Chowdhery et al.,“PaLM:Scaling Language Modeling with Pathways,”arXiv.Org,April 5,2022,https:/doi.org/10.48550/arxiv.2204.02311.31.Patterson et al.,“The Carbon Footprint of Machine Learning Training Will Plateau,Then Shrink.”32.Stephen M.Walker,“Everything We Know About GPT-4,”Klu,September 1,2023,https:/klu.ai/blog/gpt-4-llm.33.Benoit de Chateauvieux et al.,“Optimize AI/ML workloads for sustainability:Part 3,deployment and monitoring,”AWS Architecture Blog,April 27,2022,https:/ FOR DATA INNOVATION 19 34.Victor Avelar et al.,“The AI Disruption:Challenges and Guidance for Data Center Design,”Schneider Electric,White Paper 110,Version 2.1.,n.d.,accessed December 1,2023,https:/download.schneider- Wu et al.,“Sustainable AI:Environmental Implications,Challenges and Opportunities,”arXiv.Org,October 30,2021,https:/doi.org/10.48550/arxiv.2111.00364.36.Alexandra Sasha Luccioni,Yacine Jernite,and Emma Strubell,“Power Hungry Processing:Watts Driving the Cost of AI Deployment?,”arXiv,November 28,2023,https:/doi.org/10.48550/arxiv.2311.16863.37.Ibid.38.Ibid.39.De Vries,“The Growing Energy Footprint of Artificial Intelligence.”40.“Google Environmental Report:2023,”Google,July 2023,https:/ et al.,“The Carbon Footprint of Machine Learning Training Will Plateau,Then Shrink.”43.Radosvet Desislavov,Fernando Martnez-Plumed,and Jos Hernndez-Orallo,“Trends in AI Inference Energy Consumption:Beyond the Performance-vs-Parameter Laws of Deep Learning,”Sustainable Computing:Informatics and Systems 38(April 1,2023):100857,https:/doi.org/10.1016/j.suscom.2023.100857.44.Ibid.45.De Vries,“The Growing Energy Footprint of Artificial Intelligence.”46.“GPT-4,”OpenAI,March 14,2023,https:/ Masanet et al.,“Recalibrating Global Data Center Energy-Use Estimates,”Science 367,no.6481(February 28,2020):98486,https:/doi.org/10.1126/science.aba3758.48.Ibid.49.Desislavov,Radosvet,Fernando Martnez-Plumed,and Jos Hernndez-Orallo.Trends in AI Inference Energy Consumption:Beyond the Performance-vs-parameter Laws of Deep Learning.Sustainable Computing:Informatics and Systems 38,(2023):100857,accessed September 22,2023.https:/doi.org/10.1016/j.suscom.2023.100857.50.“Power Efficiency,”Nvidia,n.d.,accessed December 15,2023,https:/ Hoefler et al.,“Sparsity in Deep Learning:Pruning and Growth for Efficient Inference and Training in Neural Networks,”arXiv.Org,January 31,2021,https:/doi.org/10.48550/arxiv.2102.00554.52.Patterson et al.,“The Carbon Footprint of Machine Learning Training Will Plateau,Then Shrink.”53.Ibid.CENTER FOR DATA INNOVATION 20 54.Bill Tomlinson et al.,“The Carbon Emissions of Writing and Illustrating Are Lower for AI than for Humans,”arXiv(Cornell University),March 8,2023,https:/doi.org/10.48550/arxiv.2303.06219.55.Vida Rozite,Jack Miller,and Sungjin Oh,“Why AI and energy are the new power couple,”IEA,November 2,2023,https:/www.iea.org/commentaries/why-ai-and-energy-are-the-new-power-couple.56.Vt Rika et al.,“Semantic Segmentation of Methane Plumes with Hyperspectral Machine Learning Models,”Scientific Reports 13,no.1(November 17,2023),https:/doi.org/10.1038/s41598-023-44918-6.57.David Rolnick et al.,“Tackling Climate Change with Machine Learning,”ACM Computing Surveys 55,no.2(February 7,2022):196,https:/doi.org/10.1145/3485128.58.Lin Chen et al.,“Artificial Intelligence-Based Solutions for Climate Change:A Review,”Environmental Chemistry Letters 21,no.5(June 13,2023):252557,https:/doi.org/10.1007/s10311-023-01617-y.59.Daniel Trotta,“California Turns to AI to Help Spot Wildfires,”Reuters,August 11,2023,https:/ Li et al.,“The Impact of Artificial Intelligence on Firms Energy and Resource Efficiency:Empirical Evidence from China,”Resources Policy 82(May 1,2023):103507,https:/doi.org/10.1016/j.resourpol.2023.103507.61.Chen et al.,“Artificial Intelligence-Based Solutions for Climate Change:A Review.”62.“Google Environmental Report:2023,”Google,July 2023,https:/ Gaster,Robert D.Atkinson,and Ed Rightor,“Beyond Force:A Realist Pathway through the Green Transition,”ITIF,August 29,2023,https:/itif.org/publications/2023/07/10/beyond-force-a-realist-pathway-through-the-green-transition/.64.Urs Hlzle,“Carbon neutrality by end of 2007,”Google,June 19,2007,https:/blog.google/outreach-initiatives/sustainability/carbon-neutrality-by-end-of-2007/.65.Urs Hlzle,“Meeting our match:Buying 100 percent renewable energy,”Google,April 4,2018,https:/blog.google/outreach-initiatives/environment/meeting-our-match-buying-100-percent-renewable-energy/.66.Sundar Pichai,“Our third decade of climate action:Realizing a carbon-free future,”Google,September 14,2020,https:/blog.google/outreach-initiatives/sustainability/our-third-decade-climate-action-realizing-carbon-free-future/.67.“Net-zero carbon,”Google,n.d.,https:/sustainability.google/operating-sustainably/net-zero-carbon/(accessed January 8,2024).68.“Driving Climate Solutions,”Amazon,n.d.,accessed January 8,2024,https:/ FOR DATA INNOVATION 21 70.“The 2023 Microsoft Impact Summary,”Microsoft,October 2023,https:/ Smith,“Microsoft will be carbon negative by 2030,”Microsoft,January 16,2020,https:/ January 8,2024,https:/ on the Ethics of Artificial Intelligence,”UNESCO,2022,https:/unesdoc.unesco.org/ark:/48223/pf0000381137.74.Margaret Mitchell et al.,“Model Cards for Model Reporting,”arXiv.Org,January 29,2019,https:/doi.org/10.1145/3287560.3287596.75.“About Model Cards,”Hugging Face,n.d.,accessed January 8,2024,https:/huggingface.co/spaces/huggingface/Model_Cards_Writing_Tool.76.“Voluntary AI Commitments,”White House,September 2023,https:/www.whitehouse.gov/wp-content/uploads/2023/09/Voluntary-AI-Commitments-September-2023.pdf.77.Marius Hessenthaler et al.,“Bridging Fairness and Environmental Sustainability in Natural Language Processing,”arXiv.Org,November 8,2022,https:/doi.org/10.48550/arxiv.2211.04256.78.Traian Rebedea et al.,“NeMo Guardrails:A Toolkit for Controllable and Safe LLM Applications with Programmable Rails,”arXiv.Org,October 16,2023,https:/doi.org/10.48550/arxiv.2310.10501.79.“AI Mandates,”EU Artificial Intelligence Act,June 20,2023,https:/artificialintelligenceact.eu/wp-content/uploads/2023/08/AI-Mandates-20-June-2023.pdf.80.“CEQ Launches Global Net-Zero Government Initiative,Announces 18 Countries Joining U.S.to Slash Emissions from Government Operations,”White House,November 17,2022,https:/www.whitehouse.gov/ceq/news-updates/2022/11/17/ceq-launches-global-net-zero-government-initiative-announces-18-countries-joining-u-s-to-slash-emissions-from-government-operations/.81.John Naughton,“Why AI Is a Disaster for the Climate,”The Observer,December 23,2023,https:/ FOR DATA INNOVATION 22 ABOUT THE AUTHOR Daniel Castro is the director of the Center for Data Innovation and vice president of the Information Technology and Innovation Foundation.He has a B.S.in foreign service from Georgetown University and an M.S.in information security technology and management from Carnegie Mellon University.ABOUT THE CENTER FOR DATA INNOVATION The Center for Data Innovation studies the intersection of data,technology,and public policy.With staff in Washington,London,and Brussels,the Center formulates and promotes pragmatic public policies designed to maximize the benefits of data-driven innovation in the public and private sectors.It educates policymakers and the public about the opportunities and challenges associated with data,as well as technology trends such as open data,artificial intelligence,and the Internet of Things.The Center is part of the Information Technology and Innovation Foundation(ITIF),a nonprofit,nonpartisan think tank.Contact:infodatainnovation.org datainnovation.org
2024-05-27
22页
5星级
Gas Market Report,Q2-2024The IEA examines the full spectrum of energy issues including oil,gas and c.
2024-05-21
59页
5星级
Reducing low-carbon hydrogen investment and operating costsA prerequisite for mass adoptionThe low-c.
2024-05-20
27页
5星级
INTERNATIONAL CO-OPERATION TO ACCELERATE GREEN HYDROGEN DEPLOYMENTA brief from the IRENA Collaborati.
2024-05-20
30页
5星级
罗兰贝格：预见2026：中国行业趋势报告（90页）.pdf
智源研究院：2026十大AI技术趋势报告（34页）.pdf
中国互联网协会：智能体应用发展报告（2025）（124页）.pdf
三个皮匠报告：2025银发经济生态：中国与全球实践白皮书（150页）.pdf
三个皮匠报告：2025中国商业航天市场洞察报告-中国商业航天新格局全景洞察（25页）.pdf
国声智库：全球AI创造力发展报告2025（77页）.pdf
中国电子技术标准化研究院：2025知识图谱与大模型融合实践案例集（354页）.pdf
三个皮匠报告：2025中国情绪消费市场洞察报告（24页）.pdf
中国银行：2026中国高净值人群财富管理白皮书（66页）.pdf
亿欧智库：2025全球人工智能技术应用洞察报告（43页）.pdf
0731-84720580
商务合作：really158d
友链申请 (QQ)：1737380874
最新报告
中英对照
全文搜索
报告精选
PDF上传翻译
多格式文档互转
入驻&报告售卖
会员权益
机构报告
券商研报
财报库
专题合集
英文报告
数据图表
会议报告
其他资源
研报
新质生产力
DeepSeek
低空经济
大模型
AI Agent
AI Infra
具身智能
自动驾驶
宠物
银发经济
人形机器人
企业出海
算力
微短剧
薪酬
白皮书
创新药
行业分析
个股研究
年报财报
IPO招股书
会议纪要
宏观策略
政策法规
其他
人工智能
信息科技
互联网
消费经济
汽车交通
电商零售
传媒娱乐
医疗健康
投资金融
能源环境
地产建筑
传统产业
英文报告
其它
行业聚焦
芯片产业
热点概念
全球咨询智库
人工智能
500强
新质生产力
会议峰会
新能源汽车
企业年报
互联网
公司研究
行业综观
消费教育
科技通信
医药健康
人力资源
投资金融
汽车产业
物流地产
电子商务
传统产业
传媒营销
其它
十五五规划系列报告合集（共48套打包）
2026低空经济/低空产业报告合集（共47套打包）
AI、科技与通信
广告、传媒与营销
消费、零售与支付
HR、文化与旅游
金融、保险与投资
能源、环境与工业
医疗制药与大健康
物流、地产与建筑
其他行业
AI ▪ 科技 ▪ 通信
数字化
金融财经
智能制造
电商传媒
地产建筑
医疗医学
能源化工
其他行业
收藏
下载
2026-02-02
AI查数
行业数据
政策法规
商业模式
产业链
竞争格局
市场规模
产业概述
其它
2026年
AI读财报
年报
一季报
半年报
三季报
IPO招股书
社会责任报告
A股
IPO申报
港股
美股&全球
新三板
微信扫码登录
手机快捷登录
账号登录
