Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road 9782759829378

Citation preview

Editor-in-Chief Guo Huadong International Research Center of Big Data for Sustainable Development Goals Aerospace Information Research Institute, Chinese Academy of Sciences China

Map Content Approval Number: GS 京 (2022) 0849 号 ISBN: 978-7-03-071143-4 Science Press ISBN(print): 978-2-7598-2936-1  ISBN (e Book): 978-2-7598-2937-8 EDP Sciences © Science Press and EDP Sciences All rights relative to translation, adaptation and reproduction by any means whatsoever are reserved, worldwide. In accordance with the terms of paragraphs 2 and 3 of Article 41 of the French Act dated March 11, 1957, “copies or reproductions reserved strictly for private use and not intended for collective use” and, on the other hand, analyses and short quotations for example or illustrative purposes, are allowed. Otherwise, “any representation or reproduction—whether in full or in part—without the consent of the author or of his successors or assigns, is unlawful” (Article 40, paragraph 1). Any representation or reproduction, by any means whatsoever, will therefore be deemed an infringement of copyright punishable under Articles 425 and following of the French Penal Code.

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road Editorial Board Editor-in-Chief:

Guo Huadong

Associate Editors-in-Chief (in alphabetical order of last name): Chen Fang

Chen Yu

Dong Jinwei

Han Qunli

Huang Chunlin

Huang Lei

Jia Gensuo

Jia Li

Li Xiaosong

Liang Dong

Liu Jie

Lu Shanlong

Sun Zhongchang

Wang Futao

Wu Bingfang

Yu Rencheng

Zuo Lijun Editorial Board (in alphabetical order of last name): Bian Jinhu

Cai Guoyin

Cao Min

Cao Wenting

Chen Fulong

Chen Min

Chen Shaofeng

Chen Yaxi

Dong Yingying

Gao Feng

Gao Xianjun

He Guojin

Hu Guangcheng

Hu Yonghong

Huang Kan

Huang Ni

Huang Wenjiang

Jia Mingming

Ju Weimin

Li Ainong

Li Bin

Li Chaopeng

Li Jingxi

Li Junsheng

Li Yaojun

Liu Bin

Liu Ge

Liu Liu

Liu Ronggao

Liu Wenjun

Liu Yang

Long Tengfei

Lu Linlin

Luo Lei

Luo Lihui

Ma Juncai

Ma Xuanlong

Ma Zhenzhen

Mao Dehua

Mi Xiangcheng

Peng Dailiang

Qiang Wenli

Shangguan Donghui

Shen Qian

Shi Jinlian

Song Kaishan

Su Hua

Sun Chengjun

Sun Liqun

Tang Yunwei

Tian Fuyou

Tian Nan

Wang Huafeng

Wang Juanle

Wang Lei

Wang Li

Wang Litao

Wang Lizhe

Wang Meng

Wang Shenlei

Wang Xingdong

Wang Yuanyuan

Wang Zongming

Wei Yanan

Xie Yihan

Yan Dongmei

Yang Ruixia

Yang Yuanwei

Yao Yue

You Jinjun

Yu Xiubo

Zeng Hongwei

Zhang Huaguo

Zhang Lu

Zhang Meimei

Zhang Xiaomei

Zhang Zhaoming

Zhao Na

Zheng Yaomin

Zhu Jinfeng

Zhu Weiwei

Zhuang Yanli

Zhou Yan

Preface

Preface

I

n 2015, the United Nations adopted the 2030 Agenda for Sustainable Development, which includes 17 sustainable

development goals (SDGs) to be achieved by 2030. The SDGs are about achieving economic, social and environmental sustainability on a global scale. Since the launch of the 2030 Agenda, China has worked to promote the SDGs while embracing a new philosophy of innovative, coordinated, green, open and shared development, achieving impressive results in eradicating absolute poverty, addressing climate change, improving ecological environment, promoting public health service and ensuring food security. Steady progress has been made in achieving high-quality development. At the same time, China has actively engaged in and promoted international development cooperation, and has provided reliable public goods for the realization of SDGs across the world. The experience of the past six years, however, has shown that there remain a number of major challenges to scientifically evaluating the implementation of the 2030 Agenda, the most serious ones being the lack of data, the incompleteness of the indicator system, and the gap in capacity of having and using data as a result of development disparity. As China’s national

i

ii

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

scientific institute, the Chinese Academy of Sciences (CAS) has long been devoted to promoting SDGs through big data. In recent years, CAS has been working with universities, research institutes and enterprises at home and abroad to explore the application of combined new technologies such as cloud computing, artificial intelligence, space technology and network communication technology to improve the evaluation system for SDGs, develop public data products and inform decisionmaking. Chinese President Xi Jinping announced on September 22, 2020, at the 75th session of the United Nations General Assembly, that China will establish an International Research Center of Big Data for Sustainable Development Goals (CBAS), to provide new impetus for the implementation of the 2030 Agenda. Subsequently, CBAS was officially launched in Beijing on September 6, 2021. President Xi Jinping sent a congratulatory letter, and United Nations Secretary-General António Guterres delivered a video message to congratulate on the establishment of the Center. I believe that CBAS will use big data to give support to the sustainable development of China and the world. In recent years, CAS has, based on its advantages, done demonstration studies on monitoring and evaluation of indicators for the goals of Zero Hunger, Clean Water and Sanitation, Sustainable Cities and Communities, Climate Action, Life below Water and Life on Land, and issued annual reports on big earth

Preface

data in support of the sustainable development goals. The 2021 report continues to focus on the practical scenarios for these SDGs’ realization, and presents research results including single indicator progress evaluation and integrated multi-indicator evaluation. These results provide stronger scientific basis for understanding the dynamic trends of SDG indicators and analyzing the problems hindering sustainable development, and they can inform decision on SDG realization in different scales and regions. 2021 marks the 50th anniversary of the restoration of the People’s Republic of China’s lawful seat in the United Nations. This CAS report is part of China’s sustained contribution in the form of science and technology to the implementation of the 2030 Agenda. CAS will further strengthen the collaborations with international counterparts to address new challenges to sustainable development through science, technology and innovation.

Hou Jianguo President, Chinese Academy of Sciences

iii

Foreword

T

he COVID-19 pandemic has brought unprecedented challenges to the implementation of the 2030 Agenda for

Sustainable Development across the world, to a large extent affecting existing achievements and resulting in stagnation or even regression. Recognizing the important role scientific and technological innovation can play in promoting economic and social development, the United Nations established the Technology Facilitation Mechanism (TFM) for SDGs in 2015. In the Sustainable Development Goals Report 2020, the United Nations Secretary-General António Guterres called for a coordinated and comprehensive international response and recovery effort based on sound data and science guided by the SDGs. More effective ways need to be explored to address the data challenge facing SDGs. Thanks to the development of science and technology, the global data volume is growing exponentially. Advances in computing and data technologies have made realtime processing and analysis of big data a reality, while new types of data combined with traditional data, such as statistical and survey data, can create more detailed, timely and highquality information. Big Earth Data technology, through its extensive use and further innovation, can be an effective way to address the data divide and the lack of information and tools

vi

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

available for sustainable development. The International Research Center of Big Data for Sustainable Development Goals, building upon the strength of CAS, will use big data to support the SDGs. A full-fledged center will have the potential to carry out functions that include building an SDG big data technology service system capable of storage, calculation, analysis, and service, conducting scientific research on monitoring and evaluating SDG indicators, developing and operating SDG science satellites, constructing a think tank on science and technology for sustainable development, and promoting personnel training and capacity building using big data for SDGs. In recent years, CAS has conducted case studies that use Big Earth Data technology to monitor and evaluate indicators for six SDGs—Zero Hunger, Clean Water and Sanitation, Sustainable Cities and Communities, Climate Action, Life below Water, and Life on Land. CAS issued reports on “Big Earth Data in Support of the Sustainable Development Goals” for two consecutive years during the 74th and 75th Sessions of the United Nations General Assembly, highlighting the important value and role of Big Earth Data technology in addressing challenges for sustainable development. Focusing on six SDGs, Big Earth in Support of the Sustainable Development Goals (2021): The Belt and Road presents 42 typical cases on four scales—local, national, regional and global—detailing the results of research, monitoring, and evaluating SDG indicators. Furthermore, the report demonstrates

Foreword

methods for monitoring, evaluating, and analyzing the interactions among multiple SDGs, thus laying a good foundation for future coordinated pursuits of multiple SDGs in different scenarios. The findings of the report can provide new analytical tools for a better understanding and more accurate identification of issues related to SDGs. They are also of great practical value to promoting SDGs through science, technology, and innovation. This report could not have been completed without the guidance given by the Ministry of Foreign Affairs, and the valuable feedback and suggestions from leaders and experts of the National Development and Reform Commission, the Ministry of Natural Resources, the Ministry of Ecology and Environment, the Ministry of Housing and UrbanRural Development, the Ministry of Transport, the Ministry of Water Resources, the Ministry of Agriculture and Rural Affairs, the Ministry of Emergency Management, the National Bureau of Statistics, and the National Forestry and Grassland Administration. Finally, our utmost appreciation goes to all the scientists on the team for their hard work.

Director General of the International Research Center of Big Data for Sustainable Development Goals Member of the UN 10-Member Group to support the TFM for SDGs (2018-2021)

vii

Executive Summary

This year, 2021, is the first year of the Decade of Action, an initiative launched by the United Nations to accelerate the implementation of the SDGs. There are still severe challenges in realizing the 2030 Agenda for Sustainable Development, and the COVID-19 pandemic has seriously impacted efforts to achieve the SDGs. Science and technology are important levers for meeting these challenges, and advancing and implementing the 2030 Agenda for Sustainable Development. In his congratulatory letter for the inauguration of the International Research Center of Big Data for Sustainable Development Goals (CBAS), Chinese President Xi Jinping noted: “Scientific and technological innovation and application of big data will help the international community overcome difficulties and implement the United Nations 2030 Agenda globally.” Furthermore, the United Nations Secretary-General António Guterres said in a video message for the inauguration of CBAS: “Building on the momentum that is being generated by the United Nations Technology Facilitation Mechanism, for innovation, solutions, and better results where policy meets science… we can mobilize scientific and technological communities to help achieve the Sustainable Development Goals.” This report showcases the innovative practice of applying Big Earth Data to the monitoring and evaluating indicators for six SDGs, i.e., SDG 2 Zero Hunger, SDG 6 Clean Water and Sanitation, SDG 11 Sustainable Cities and Communities, SDG 13 Climate Action, SDG 14 Life below Water, SDG 15 Life on Land, and the analysis of the interactions among multiple SDG indicators.

Regarding SDG 2, Zero Hunger, the case studies in this report focused on SDG indicators 2.3.1 (production per unit of labor), 2.4.1 (proportion of productive and sustainable agricultural areas) and 2.5.1 (number of plant and animal genetic resources for food and agriculture secured in medium or long term conservation facilities), addressing both cropping systems and livestock systems. Integrative analyses of SDG targets 2.1, 2.2 and 2.4 were also carried out. Four models were proposed, including two classification models for cultivated land, a pest and disease monitoring model, and a livestock productivity

x

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

evaluation model. Datasets included cultivated land distribution products for the Zambezi River basin and the Mediterranean, global livestock productivity, the effective utilization rate of grassland, nutrition and food production in key regions of Asia and Europe, and a collection of microbial resources and their utilization. Support for decision-making was provided for the development of cropping systems and livestock systems in different regions.

For Clean Water and Sanitation (SDG 6), the report focused on four targets: 6.3 (improving water quality), 6.4 (improving wateruse efficiency), 6.5 (integrating water resources management) and 6.6 (protecting and restoring water-related ecosystems). Case studies produced a distribution dataset of black and odorous water bodies in the representative countries of Europe and Asia, a global dataset of water clarity in large lakes and in other lake water bodies in Africa, and a distribution dataset of global wetland protection priority areas. The case studies analyzed the global water ecology and water environment conditions, changes in global crop water-use efficiency, and the conditions of integrated water resources management and water stress in Lancang-Mekong countriesa, based on Big Earth Data technology. The results in this chapter are a useful supplement to the United Nations SDG database system. The conclusions have important reference value for understanding the progress in achieving SDG 6.

In terms of SDG 11, Sustainable Cities and Communities, the report analyzed and evaluated three targets: 11.1 (urban housing), 11.3 (urban land use efficiency), and 11.4 (natural and cultural heritage). The case studies developed and improved a number of methods and indices by Big Earth Data technology, including a scene-oriented deep learning semantic segmentation model, comprehensive spatial evaluation of indicators for urbanization, and measurement of indicators for cultural heritage sustainable development.

a　Lancang-Mekong countries include China, Myanmar, Laos, Thailand, Cambodia, and Vietnam.

Executive Summary

The case studies also produced a series of datasets, for example, a dataset of shanty towns and urban built-up areas in 12 major cities of some regions involved in the Belt and Road Initiative, a comprehensive assessment dataset for urbanization development in World Cultural Heritage Sites in the Belt and Road Agreement countries and neighboring countries, datasets on the degree of human intervention in natural heritage and the world’s first cultural heritage (antiquities/ancient buildings) surface interference measurement dataset between 2015 and 2020. The results from the case studies in this chapter can be a significant contribution to the United Nations Sustainable Development Goals database system, and have important exemplary significance for objectively assessing the global implementation of SDG 11.

SDG 13, Climate Action, was examined through three targets: SDG targets 13.1 (resisting climate-related disasters), 13.2 (climate change measures) and 13.3 (climate change adaptation and early warning). The case studies took advantage of Big Earth Data to produce disaster datasets for target 13.1, including high-temperature heatwaves in the Eastern Hemisphere, global forest and grass fire range, the freeze-thaw vulnerability in High Mountain Asia, and floods in Pakistan. For 13.2, carbon budget datasets were produced, including gas flares from global oil-producing areas, global soil respiration, global net ecosystem productivity (NEP) and its driving factors. Multiple cycle change datasets were also generated, including the change in glacier mass balance in Eurasia, the melting period of polar ice sheets, global ocean heat content based on integrated observations of satellite and Argo buoys, and vegetation cover in Northwest China and Central Asia. Based on these datasets with spatiotemporal characteristics, it was found that in terms of SDG 13.1, the number and frequency of high-temperature heatwaves in the Eastern Hemisphere have increased significantly in the past 10 years; from 2015 to 2020, the global forest and grassland burned areas were similar but increased significantly in South America; by the end of the 21st century, the medium- and high-risk areas of freezethaw vulnerability in High Mountain Asia will rise from 20% to 26% (RCP 4.5) or 32% (RCP 8.5); and after 2015, flood disasters in Pakistan increased significantly. In terms of SDG 13.2, gas flare emissions from major oil-producing areas in the world increased from 378 million tons in 2010 to 422 million tons in 2019; since 2000, the total amount of

xi

xii

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

global soil respiration has increased significantly; the global terrestrial ecosystem carbon sequestration capacity has also been significantly enhanced, and land cover change and climate change are the main driving factors. For SDG 13.3, glaciers in Asia and Europe have experienced significant and accelerating mass loss in the past two decades, and the pressure on water resources in the basin is rising; from 1989 to 2020, the average melting duration of the Arctic Greenland ice sheet increased by 10 days, while the Antarctic ice sheet decreased by 9 days; in the past 30 years, the global ocean heat content has been rising and intensifying; and over the past 20 years, the ecological status of most areas in Central Asia and Northwest China has changed for the better, and the vegetation has shown a significant greening trend, especially in some areas in China. The above data and analysis can provide a scientific basis for coping with climate-change-related disasters and long-term impacts, as well as the implementation of carbon neutrality strategies.

The chapter on SDG 14, Life below Water, focused on targets 14.1 (reduce marine pollution), 14.2 (protect marine and coastal ecosystems), and 14.7 (increase economic benefits from the sustainable use of marine resources for island states and the least developed countries), supported by Big Earth Data technology. Five case studies developed datasets on the spatial distribution of microplastics in seawater around the Antarctic Ocean, the water inundation monitoring of Bangladesh from 2016 to 2020, the spatial distribution of mangroves and coastal aquaculture ponds at 10 m, and the spatial distribution of resources in the coastal zone of Mozambique from 1990 to 2020. Analysis showed that the distribution of floating microplastics in the Antarctic Ocean was characterized by high local abundance, and the subtropical front of the Antarctic circulation had the potential to prevent the transport of floating plastic debris from low to high latitudes. The most severe flooding in Bangladesh was between the second half of July and the first half of August each year, and the peak flooding area was around 20,000 km2. The total area of mangroves in 2020 was 1.44×105 km2, with the largest portion 38.99% in Asia and 19.38% in Africa. The total area of coastal aquaculture ponds in the world is 37,200 km2, of which 86% are located in Asia, with an area of about 32,000 km2.

Executive Summary

Regarding SDG 15 Life on Land, the case studies focused on four indicators: 15.1.1 (forest area as a proportion of total land area), 15.2.1 (progress toward sustainable forest management), 15.3.1 (proportion of land that is degraded over the total land area), and 15.4.2 (mountain green cover index, MGCI). The cases studies established indicator evaluation models and methods supported by Big Earth Data, and carried out the exemplary analysis in the whole area of “the Belt and Road” or representative areas. A global-scale 30 m forest coverage dataset was produced for 2020; the characteristics of global forest change patches from 2000 to 2020 were discovered; the land degradation maps of Mongolia in 2015 and 2020 were produced; an ecological vulnerability index was designed for the Bangladesh-China-India-Myanmar Economic Corridor; and highresolution mountain green cover index datasets were generated for 2015 and 2020. These products provide strong support for the dynamic monitoring and evaluation of SDG 15 indicators.

Regarding the interactions among SDG indicators, this report SDGs

focused on the methods and practice of mining and integrating spatial information using Big Earth Data. The case studies evaluated the synergy and trade-off relationships among SDG targets in the context of their correlations, and simulated multiple indicators’ interactions

in future environmental, economic, and social scenarios in the context of their temporal variations. Through integrated studies of multiple indicators, this report found potential issues and future research directions. This will contribute to the evaluation of the effects of current policies and measures, and guide the dynamic planning of policies. The integrated analyses of SDG indicators can speed up the implementation process of regional sustainable development.

xiii

Contents

i

Preface

v

Foreword

ix

Executive Summary

Chapter 1

Introduction / 1

Background / 8 Main Contributions / 9 Chapter 2 SDG 2 Zero Hunger

Case Studies / 11 2.1  Assessment of global livestock productivity and China’s contribution / 11 2.2  Dynamic changes in grassland above-ground biomass in representative East African countries / 19 2.3  Analysis of cropland area change in the 21st century in the Zambezi River basin / 27 2.4  Cropland dynamic change in the Mediterranean region from 2010 to 2020 / 34 2.5  Monitoring desert locusts in Asia and Africa / 40

Contents

2.6  Global data tracking of the CBD benefit-sharing of microbial resources / 46 2.7  Food security and agricultural sustainability in major regions of Asia, Europe and Africa / 51 Summary / 62

Background / 66 Main Contributions / 67 Chapter 3 SDG 6 Clean Water and Sanitation

Case Studies / 69 3.1  Changes in black and odorous water bodies in representative cities of Europe and Asia / 69 3.2  Monitoring and evaluating the dynamic changes in global lake water clarity / 81 3.3  Change in water transparency in Africa from 1985 to 2020 / 88 3.4  Assessment of changes in global crop water-use efficiency / 96 3.5  Comparative analysis of the implementation of IWRM and water stress in Lancang-Mekong countries / 101 3.6  Mapping of global wetland conservation priority areas / 110 Summary / 117

xv

xvi

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

Background / 120 Main Contributions / 121 Chapter 4 SDG 11 Sustainable Cities and Communities

Case Studies / 123 4.1  Monitoring shanty towns in major cities of some regions involved in the Belt and Road Initiative / 123 4.2  Land cover variability and its driving factors in global big cities (2021) / 130 4.3  Analysis of land use efficiency for cities of the countries involved in the Belt and Road Initiative with different urban sizes / 135 4.4  Comprehensive assessment of urbanization at World Cultural Heritage Sites in the countries involved in the Belt and Road Initiative and the neighboring countvies / 145 4.5  Measures and findings of interference at World Cultural Heritage Sites for SDGs / 150 4.6  Monitoring and evaluation of protection indicators of global World Natural Heritage Sites / 159 Summary / 165

Contents

Background / 170 Main Contributions / 171 Chapter 5 SDG 13 Climate Action

Case Studies / 174 5.1  Variation in the range of heatwave influence in the Eastern Hemisphere / 174 5.2  Global burned area distribution and changes / 180 5.3  Vulnerability prediction of freeze-thaw disasters in High-mountain Asia / 184 5.4  Flood changes and analysis of disaster reduction in Pakistan from 2010 to 2020 / 190 5.5  Global CO2 emissions and spatiotemporal variations from gas flaring in oil production fields / 195 5.6  Temporal and spatial changes in global soil respiration and its response to climate change / 200 5.7  Impacts of land cover change on global net ecosystem productivity / 207 5.8  Simulation of spatiotemporal characteristics of Eurasian glacier retreat and evaluation of its effects on water resources / 211 5.9  Monitoring and evaluation the freeze-thaw changes of polar ice sheet / 217 5.10  Global ocean heat content change / 227

xvii

xviii

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

5.11  Response of terrestrial ecosystems to climate change in Northwest China and five countries in Central Asia / 231 Summary / 239

Background / 244 Main Contributions / 246 Chapter 6 SDG 14 Life Below Water

Case Studies / 247 6.1  Spatial distribution characteristics of microplastics in representative areas of Antarctica / 247 6.2  Coastal flood monitoring products for Bangladesh from 2016 to 2020 / 256 6.3  Global mangrove forests spatial distribution monitoring in 2020 / 263 6.4  Spatial distribution of global coastal aquaculture ponds / 268 6.5  Remote sensing assessment of coastal resources in Mozambique for marine spatial planning / 274 Summary / 278

Contents

Background / 282 Main Contributions / 283 Chapter 7 SDG 15 Life on Land

Case Studies / 284 7.1  Global/regional forest cover (2020) / 284 7.2  Event-based analysis of global forest change in the 21st century / 289 7.3  Dynamic monitoring and control measures of land degradation and restoration in Mongolia (2015-2020) / 297 7.4  Ecological vulnerability assessment along the Bangladesh-ChinaIndia-Myanmar Economic Corridor / 303 7.5  High-resolution global monitoring of the mountain green cover index / 313 Summary / 318

Background / 322 SDGs

Chapter 8 Interactions Among SDG Indicators

Main Contributions / 325 Case Studies / 326 8.1  Trade-offs of the food-water-air quality nexus over the breadbasket of India / 326

xix

xx

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

8.2  Simulation and analysis of land use evolution in regions involved in the Belt and Road Initiative under the constraints of multiple SDGs / 332 Summary / 339

Chapter 9

Summary and Prospects / 343

References / 348

Acronyms / 361

Contents

xxi

Chapter 1

Introduction

2

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

In a push to achieve all 17 SDGs by 2030, the United Nations formally launched in January 2020 the Decade of Action, calling for accelerating sustainable solutions to the world’s greatest challenges. However, the COVID-19 pandemic has had a serious impact on the global implementation of the 2030 Agenda for Sustainable Development. It has increased the vulnerability of the global food system, with the number of people facing hunger in 2020 increasing to about 118 million, or 18% higher than in 2019, and food security emergencies at the highest level in five years. Over the past century, global water use has grown at more than twice the rate of population growth, and the United Nations estimates that global freshwater resources will decline by 40% by 2030, making a water crisis highly likely. Prior to the pandemic, cities already saw growing numbers of slum-dwellers, more polluted air, minimal public open space, and limited public transportation. The pandemic has further exposed and aggravated such vulnerabilities. The concentration of major greenhouse gases in the atmosphere continues to increase, with 2015-2020 being the warmest six years on record. Climate change has made the achievement of many SDGs less likely. The ocean constantly faces threats such as pollution, warming and acidification, which are disrupting the marine ecosystem. Deforestation and forest degradation, the continued loss of biodiversity, and the degradation of ecosystems are having far-reaching impacts on human wellbeing and survival. The global target of halting biodiversity loss by 2020 was not met (United Nations, 2021a, 2021b). The Sustainable Development Goals Report 2021 of the United Nations points out the need for concerted efforts to support a recovery guided by the 2030 Agenda for Sustainable Development, and the acquisition and the availability of data is one of the key factors in achieving a better recovery. Data that support monitoring and evaluation of SDGs have increased significantly over the years, but major gaps remain in terms of geographic coverage and timeliness of data. The Global SDG Indicators Database reveals that more than 80% of countries have data for only a few SDGs, and for most SDGs, data timeliness is a serious problem (United Nations, 2021c). These data gaps hinder the real-time monitoring of progress toward the goals and the assessment of regional disparities. Data innovation is the key to closing the gaps and accelerating the realization of SDGs, and an important area of such innovation is the fusion of geospatial information and statistical

Chapter 1  Introduction

information. Earth observation data collected by satellites, unmanned aerial vehicles and ground sensors can supplement official statistics and survey data, and fuse with traditional data to create high-quality information that is more timely and spatially representative. This type of Earth observation data with spatial attributes, referred to as Big Earth Data, has strong spatiotemporal and physical correlations and good controllability of data generation methods and sources, in addition to the general properties of big data: massive, multi-source, heterogeneous, multitemporal, multi-scale and nonstationary (Guo, 2017; Guo et al., 2016). Big Earth Data can help us understand the complex interactions and evolutionary processes between Earth’s natural systems and human social systems, thus contributing to the realization of the SDGs. Big Earth Data science includes these main technological systems: (1) ubiquitous sensing of Big Earth Data, (2) credible Big Earth Data sharing, (3) multiple Big Earth Data fusion, (4) Big Earth Data digital twin and complex process simulation, and (5) intelligent cognition of Big Earth Data (Figure 1.1). Using Big Earth Data to support SDG monitoring and evaluation has the following unique advantages. First, monitoring results are more transparent and repeatable due to data from diverse sources verifying each other. Second, information on spatial differences and dynamic changes is linked to SDG indicators, enabling decision-makers to use the former to detect and address the imbalances and weak links in the latter to identify changing trends and policy effects.

Figure 1.1  Technological system of Big Earth Data science

CAS uses Big Earth Data to support SDGs and has established platforms focused on the field. The Big Earth Data Science Engineering Program (CASEarth), the Big Earth Data Sharing Service Platform and the Big Earth Data Cloud Service Infrastructure provide data, online calculation and visualization for monitoring and evaluating SDG indicators (Figure 1.2). As of December 31,

3

4

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

2021, CASEarth has shared a total of about 11 PB of data, and updates 3 PB of data every year. It has more than 410,000 unique IP users in 174 countries and regions with 66 billion data visits.

Figure 1.2  Diagram of Big Earth Data supporting SDGs

Of the 17 goals, SDG 2 Zero Hunger, SDG 6 Clean Water and Sanitation, SDG 11 Sustainable Cities and Communities, SDG 13 Climate Action, SDG 14 Life below Water and SDG 15 Life on Land are closely related to Earth’s environment and resources. This report, Big Earth Data in Support of the Sustainable Development Goals: The Belt and Road, in its 2021 edition, presents results of studies carried out on selected indicators under six goals where Big Earth Data can play an important role in their monitoring and evaluation. In its 2019 and 2020 editions, the reports presented new methods, new products and decision support cases using Big Earth Data to monitor and evaluate progress on the SDGs. The 2021 report focuses on updates and extensions, new methodologies and indicators, the tracking and evaluation of SDG implementation, the study of interactions among multiple SDGs, and coordinated development in light of those interactions. It presents 42 case studies on 22 targets relevant to countries and regions involved in the Belt and Road Initiative. The report showcases the results of research, monitoring, and evaluation of SDGs and their indicators at four scales— local, national, regional, and global—totalling 37 data products, 19 methods and models, and 32 decision-support recommendations.

Chapter 1  Introduction

5

81

Background

91

Main Contributions

111 Case Studies 621 Summary

Chapter 第二章 2

Zero Hunger SDG 2 零饥饿

8

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

Background More than six years after the implementation of the 2030 Agenda for Sustainable Development, progress toward the goal of Zero Hunger has been slow and there is still a huge gap between the goal and its realization. The United Nations Secretary-General António Guterres has warned that the world is facing its biggest food crisis in 50 years. Conflicts, climate variability and extremes, economic slowdowns, and recessions have all increased dramatically in frequency and severity over the past decade (FAO et al., 2021). The COVID-19 pandemic could add 83 million to 132 million more undernourished people globally to the current 690 million. Africa, in particular, has seen per capita food production fall to its lowest level since 2014 due to the double impact of desert locusts and COVID-19. In addition, due to the simultaneous impact of conflict, extreme weather, and economic recession, the food security situation in the region has deteriorated over the past five years, and the number of people affected by hunger is at risk of doubling. Transforming agri-food systems to make them more efficient, inclusive, resilient and sustainable is key to achieving Zero Hunger (FAO, 2021). This includes action based on environmental measures such as the “5 Fs” (food, feed, fiber, forestry and fuel), as well as urban gardening and flower cultivation. In The State of World Food and Nutrition 2021, the Food and Agriculture Organization (FAO) proposed six potential ways to transform food systems in the dimensions of food supply security, access, utilization, stability, mobility, and sustainability, and listed technology, data, and innovation as accelerating factors (FAO et al., 2021). Big Earth Data has the capability for dynamic, rapid, macro-level monitoring, which can provide a basis for regional assessments of food production and environmental change, informing the overall cognition of large-scale progress and a detailed grasp of regional differences. Effectively combining Big Earth Data with statistical data can greatly improve the evaluation of SDG indicators that currently have no method or data, and provide new impetus for the monitoring of SDGs. In the reports of the 2019 and 2020, we focused on SDG targets 2.3 and 2.4, mainly on planting in agricultural production systems, and carried out a series of studies on farmland area distribution, planting pattern monitoring, desert locust disaster assessment, farmland productivity monitoring, production potential assessment, and food security early warning. In the chapter on SDG 2 this year, animal husbandry will be included in the research based on further monitoring of cultivated land use in key regions. In this way, the contributions of Big Earth Data to monitoring SDG target 2.5 will be expanded. Finally, a comprehensive analysis of SDG 2 in developing countries will be presented.

Chapter 2  SDG 2 Zero Hunger

Main Contributions Case studies were carried out regarding cropping systems and livestock systems, focusing on SDG indicators 2.3.1 (production per labor unit by agricultural/livestock/forestry firm size), 2.4.1 (proportion of agricultural areas under productive and sustainable agriculture), and 2.5.1 (number of plant and animal genetic resources for food and agriculture secured in medium or long term conservation facilities), which reflect food production security. An integrative analysis of SDG targets 2.1, 2.2 and 2.4 was also carried out (Table 2.1). Four models were proposed, including two classification models for cultivated land, a pest and disease monitoring model, and a livestock productivity evaluation model. Datasets include cultivated land distribution products for the Zambezi River basin and the Mediterranean, global livestock productivity, the effective utilization rate of grassland, nutrition and food production in key regions of Asia and Europe, and a collection of microbial resources and their utilization. Support for decision-making was developed for cropping systems and animal husbandry in different regions.

Table 2.1  Cases and their main contributions Indicators/Targets

SDG 2.3.1 Production per labor unit by agricultural/ livestock/ forestry firm size

Tiers

Cases

Contributions

Assessment of global livestock productivity and China’s contribution

Data product: Data on changes in global livestock productivity and effective utilization rate of grassland productivity Method and model: Livestock productivity assessment model integrating feed consumption, remote sensing monitoring and livestock production Decision support: Provides decision support for improving global livestock productivity

Dynamic changes in grassland aboveground biomass in representative East African countries

Decision support: Provides decision support for the development of animal husbandry in representative regions in East Africa through case analysis

Tier Ⅱ

9

10

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

(continued) Indicators/Targets

SDG 2.4.1 Proportion of agricultural areas under productive and sustainable agriculture

SDG 2.5.1 Number of plant and animal genetic resources for food and agriculture secured in medium or long term conservation facilities SDG 2.1 Food is nutritious and safe; SDG 2.2 Eliminate malnutrition in all its forms; SDG 2.4 Sustainable food production systems

Tiers

Tier Ⅱ

Tier Ⅰ

Tier Ⅰ

Cases

Contributions

Analysis of cropland area change in the 21st century in the Zambezi River basin

Data product: Cropland distribution datasets in the Zambezi River basin in 2000 and 2020 Method and model: Classification model of cropland in the Zambezi River basin Decision support: Provides a basis for policies for Zero Hunger in the Zambezi River basin

Cropland dynamic change in the Mediterranean region from 2010 to 2020

Method and model: Remote sensing model of long-term, time-series cultivated land in cases lacking ground samples Decision support: Provides information on the dynamic change in the food supply situation and causal analysis in the Mediterranean region

Monitoring desert locusts in Asia and Africa

Method and model: Pest monitoring model Decision support: Supports multinational joint prevention and control efforts to safeguard agricultural and animal husbandry production and regional stability in the invaded country

Global data tracking of the Convention on Biological Diversity (CBD) benefits-sharing of microbial resources

Data Product: Collection data of microbial resources and their utilization in China from 2001 to 2019 Decision support: The big data platform of microbial genetic resources provides decision support for the benefit-sharing and utilization of microbial resources in China

Food security and agricultural sustainability in major regions of Asia, Europe and Africa

Data Product: Dataset on nutritional status and food production in major regions of Asia, Europe and Africa Decision support: Provides decision-making support for eliminating hunger, achieving food security, improving nutrition and promoting sustainable agricultural development in major regions of Asia, Europe and Africa

Note: Tier Ⅰ: The indicators are divided into three level. Indicator is conceptually clear, has an internationally established methodology and standards are available, and data are regularly produced by countries for at least 50 per cent of countries and of the population in every region where the indicator is relevant.Tier Ⅱ: Indicator is conceptually clear, has an internationally established methodology and standards are available, but data are not regularly produced by countries. Tier Ⅲ: No internationally established methodology or standards are yet available for the indicator, but methodology/standards are being (or will be) developed or tested. (As of the 51st session of the United Nations Statistical Commission, the global indicator framework does not contain any Tier Ⅲ indicators)

Chapter 2  SDG 2 Zero Hunger

Case Studies 2.1  Assessment of global livestock productivity and China’s contribution Target SDG 2.3: By 2030, double the agricultural productivity and incomes of smallscale food producers, in particular women, indigenous peoples, family farmers, pastoralists and fishers, including through secure and equal access to land, other productive resources and inputs, knowledge, financial services, markets and opportunities for value addition and non-farm employment.

Background Food security has become an important issue worldwide under the effects of climate change, socio-economic development, and population growth. In addition, most developing countries are undergoing a transition in food systems, which has led to increasing demands for animal-source food in these regions. Livestock production has contributed the most to the increasing food supply in recent years. The increasing trade of feed and forage grass was also driven by the expansion of livestock production. As the scarcity of water and land resources intensifies globally, the promotion of global livestock productivity will be a great challenge. The existing assessments of “Zero Hunger” (SDG 2) are mainly focused on direct goals, and lack the evaluation of livestock productivity and its contribution to food security. As the calorie intake per capita increases globally, protein and vitamin deficiency will be a major source of hidden hunger. The proportion of animal-source food in the nutrition supply will differ with the variation of resource endowment and demands of consumption. Continuous tracking of SDG 2.3 is a prerequisite to ensuring the realization of the target. Therefore, based on FAO databases, remote sensing data, and existing research, we carried out a global assessment of livestock productivity, explored its spatiotemporal changes and regional differences, and focused on the evolution of China’s livestock productivity and its role.

11

12

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

Data ●●

Global livestock and poultry production data, feed consumption data from FAO statistics at the national scale (1990-2018).

●●

Distribution coefficients of grain and other feed products in different livestock categories were derived from the literature (1990-2018) (Kastner et al., 2014).

●●

Global grassland productivity data using MOD17A3HGF, 500 m spatial resolution (2000, 2010, 2018).

Method (1) Analysis of the concentration trend of livestock production: analyze the changes in livestock production, output structure, and per capita supply, and display the centralization characteristics of livestock production using a Lorenz curve. (2) Measurement of livestock production efficiency: the feed for poultry production in most countries consists of cereals, beans and their by-products, and the production efficiency of poultry is measured by the ratio of feed energy consumption to meat protein. The production system and parameters of cattle and sheep are complex, but grass has a high proportion in all production systems (Herrero et al., 2013); thus the effective utilization coefficient of grassland productivity is used to measure the beef and mutton production efficiency, calculated by the ratio of beef and mutton production to total hay production in each country.

Results and analysis Animal-source food plays an increasingly prominent role in global nutrition supply. From 1990 to 2018, per capita annual protein supply has increased from 25.7 kg to 29.6 kg globally, of which per capita annual animal-source protein supply has risen from 7.7 kg to 10.0 kg, and the proportion of animal-source protein of the total has increased from 30.2% to 33.9%. The protein supply from animal sources in China has increased from 3.8 kg to 11.5 kg per capita annually, from below the global average to above the global average. The production of livestock meat increased from 179 million tons to 343 million tons globally from 1990 to 2018. Poultry meat increased the most by 2.2 times, followed by pork, increasing by 73%, while beef and mutton increased by 30% and 64%, respectively. Per capita annual meat supply has increased from 34.2 kg to 46.1 kg globally, an increase of 34.8%. China’s meat supply has been lifted from 28 million tons to 86 million tons, with an increase of nearly two times from 1990 to 2018, accounting for 25.2% of the world’s total production in 2018, and has become the

Chapter 2  SDG 2 Zero Hunger

world’s largest livestock producer. The production of global livestock meat was concentrated in a few countries; the output of the top ten countries took up 63% of the total production [Figure 2.1(a)]. The number of countries

Figure 2.1  Concentration of global meat production (a) and temporal and spatial changes in per capita meat production of each country (b)

13

14

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

with per capita meat production higher than the global average increased from 39 to 64 from 1990 to 2018. However, the disequilibrium of per capita meat production showed an increasing trend. From 1990 to 2018, China’s per capita annual meat supply increased from 24.1 kg to 60.6 kg, and transformed from below the global average to above the global average, but still showing a large disparity compared with the highest output. The structure of global meat production also showed significant change (Figure 2.2). The proportion of poultry meat in total increased from 22.8% to 37.5%, while the share of beef and mutton decreased from 36.2% to 25.5% from 1990 to 2018. China’s meat output structure was quite different from the global average: the proportion of pork accounted for the largest proportion of total meat production, but it has shown a downward trend in recent years, from 78.7% to 62.3%; the share of beef and mutton production in China has increased from 7.9% to 12.7%. However, this value is still lower than the global average, which means there is great growth potential.

Figure 2.2  Changes in meat production structure in major countries from 1990 to 2018

The changing trend of global pork and poultry production efficiency is shown in Figure 2.3. The feed use efficiency of pork and poultry showed an increasing trend from 1990 to 2018 as a whole, the protein output of per unit feed calorie has risen from 6.4 kg/kcal and 17.6 kg/kcal to 8.3 kg/

Chapter 2  SDG 2 Zero Hunger

kcal and 22.7 kg/kcal respectively, increasing by 30% and 29%, respectively, but there was large variation spatially. Although the feed use efficiency of pork and poultry production in China was relatively high, the feed productivity changed little due to the great change of production mode from backyard farming to industrial systems. The efficiency of other major production countries, such as the United States, Spain, Brazil and Russia, showed an increasing trend. While most countries in Southeast Asia, Central Africa, and Latin America showed a downward trend. The results of global grassland productivity and effective utilization coefficient from 2000 to 2018 are shown in Figure 2.4. The degree of effective utilization coefficient of global grassland productivity showed an increasing trend generally, but there were great differences among countries. The effective utilization degree of grassland in China increased by 30% from 2000 to 2018, higher than that of the United States, Brazil and Russia, but lower than that of Germany and Japan in 2018. The effective utilization coefficient of grassland in northern Africa, Mongolia and Central Asia showed a downward trend.

15

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

Figure 2.3  Change in global pork and poultry production efficiency (1990-2018)

16

Figure 2.4  Change in global grassland productivity and effective utilization coefficient (2000-2018)

Chapter 2  SDG 2 Zero Hunger

17

18

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

Highlights  A globally consistent and spatially comparable assessment of livestock productivity progress was developed based on an FAO database and Moderate Resolution Imaging Spectroradiometer (MODIS) data products. The spatiotemporal change and regional variation were explored.  China’s livestock productivity increased rapidly and doubled its output from 1990-2018, making the largest contribution to global livestock expansion. However, China also needs to promote its production efficiency to fulfill the goal of doubling production.

Outlook The assessment method used in this case study is operable on the global scale, and the results are globally comparable. However, the livestock production system varied among regions due to the resource endowment and production scale, as well as the unavailability of relevant data. This result does not accurately represent the progress of SDG target 2.3 in specific countries and regions, but it does recognize global progress. The results of this case study are also consistent with the conclusions of relevant research (Herrero et al., 2013). However, the accuracy of the assessment needs to be improved in the future. Livestock production requires more water and land resources and causes more environmental emissions. Thus, the improvement of livestock productivity faces greater challenges than crop production. For countries with scarce resources such as China, the key to the improvement of livestock productivity is efficiency. The production capacity should be improved by increasing the proportion of roughage such as grassland and straw for cattle and sheep production. It should be noted that although the overall productivity of global livestock has increased remarkably from 2000 to 2018. There is still large heterogeneity among different livestock categories and regions. In particular, China’s productivity of poultry production has been at a high level; thus, the potential for increasing livestock production will mainly come from beef and mutton production. However, the effective utilization coefficient of grassland productivity in China is still lower than that of German and Japan, and has the potential to increase. Other developing countries also have great potential for improving the productivity of livestock production. Therefore, the future goal should focus on narrowing the regional gap. Although many countries and regions have made improvements in a short time scale, the continuous development of positive measures is particularly important in order to improve regional efficiency in the future.

Chapter 2  SDG 2 Zero Hunger

2.2  Dynamic changes in grassland above-ground biomass in representative East African countries Target SDG 2.3: By 2030, double the agricultural productivity and incomes of smallscale food producers, in particular women, indigenous peoples, family farmers, pastoralists and fishers, including through secure and equal access to land, other productive resources and inputs, knowledge, financial services, markets and opportunities for value addition and non-farm employment.

Background Grassland resources are vital to the survival of human and animal populations, and their sustainable use is important for achieving the goal of Zero Hunger. The vast grasslands of East Africa provide the natural conditions for the development of local livestock farming, but the interference of frequent drought-related events and human activity has led to a series of problems such as overgrazing, large-scale reclamation, increased reduction of grassland productivity, reduced biodiversity, and serious desertification. Thus, the sustainable development of grassland resources and livestock farming is facing great challenges. In addition, Kenya in East Africa, where global tourism is more developed for grassland national parks, has been affected by the sudden outbreak of COVID-19, which has had a major impact on the local tourism industry. Ethiopia in East Africa has been greatly affected by the 2019-2020 locust disaster. Both of these events have affected the livelihoods of local residents and pastoralists, and in serious cases, even triggered inter-regional political unrest. Grassland above-ground biomass is a key indicator for evaluating the health of grassland ecosystems and the sustainable exploitation of grassland resources, and is also the basis for a comprehensive analysis of grass-livestock balance, whose continued stability and dynamics are crucial to the development of livestock farming, regional livestock imports, and food security in East Africa. Therefore, tracking the dynamic changes of grassland aboveground biomass in representative East African countries is important for scientifically assessing the healthy development of local livestock farming and for diagnosing the achievement of the SDG 2 Zero Hunger in the region. The 2030 Agenda for Sustainable Development defines SDG indicator 2.3.1 as the volume of production per labor unit by classes of farming/pastoral/forestry enterprise size. With the development of Earth observation technology, numerous datasets of grassland above-ground biomass products have been accumulated. This case study analyses the spatial and temporal trends

19

20

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

of grassland above-ground biomass in Kenya and Ethiopia and the impact of drought events, locusts and COVID-19 in recent years.

Data ●●

Monthly above-ground biomass data of grasslands in Ethiopia and Kenya from the sub-project “Monitoring and Assessment of Agricultural Land Resources in the Belt and Road” of the “Earth Big Data Science Project”, 1 km spatial resolution (2001-2020).

●●

African Grassland Classification Map data in 2020 from the sub-project “Monitoring and Assessment of Agricultural Land Resources along the Belt and Road” of the “Earth Big Data Science Project”, 100 m spatial resolution (2020).

●●

Monthly TerraClimate dataset products, including palmer drought severity index (PDSI) drought data and monthly precipitation data; from the University of Iowa, monthly temporal resolution, 4 km spatial resolution (2001-2020).

●●

ECMWF reanalysis dataset including day-by-day meteorological and radiometric data (20012020).

Method The Manner-Kendall (M-K) trend test method was used to analyze the spatial and temporal variability of grassland biomass from 2001 to 2020 and the logical relationships between the changes in grassland biomass of different grassland types based on the monthly data of grassland above-ground biomass of Ethiopia and Kenya from 2001 to 2020 combined with the grassland classification map data. Using the Carnegie-Ames-Stanford Approach (CASA) model, the theoretical grassland above-ground biomass was first calculated based on the reanalysis data and remote sensing data, and the disturbance of grassland above-ground biomass in Ethiopia and Kenya was obtained by subtracting it from the grassland above-ground biomass, and the spatial and temporal variation of the disturbance of grassland above-ground biomass was also analyzed based on the M-K trend test method. Taking Ethiopia as an example, the PDSI from 2001 to 2018 was used in combination with grassland classification map data, overlaid with Ethiopia’s protected area boundary file, to give a qualitative analysis of the impact of drought on the whole country of Ethiopia, different areas inside and outside the protected area, through statistical analysis of monthly grassland aboveground biomass. For the locust event impact time period and representative impact areas from June to August 2019, grassland above-ground biomass data covering this time period were used and overlaid with grassland classification map data to analyze changes in grassland above-ground

Chapter 2  SDG 2 Zero Hunger

biomass in representative locust impact areas in Ethiopia. Taking Kenya as an example, and using the year 2020 as the reference year for the emergence of COVID-19, the changes in annual precipitation within the protected area in Kenya in recent years were counted, and years with annual precipitation close to or higher than that of 2020 were selected, i.e., the impact of inter-annual drought factors on grassland above-ground biomass was not considered; the monthly changes in grassland above-ground biomass in that year were compared with that of 2020, and the impact of the emergence of COVID-19 on grassland aboveground biomass within the protected area in Kenya was analyzed between years.

Results and analysis The results of remote sensing monitoring showed that in the last 20 years, grassland aboveground biomass in Ethiopia increased slightly, with significant increases (11.5%) mainly in the western and central-eastern savanna and grassland areas (Figure 2.5). Meanwhile, grassland aboveground biomass in Kenya showed a slight decrease, with significant decreases (15.2%) mainly in the eastern and southern savanna and grassland areas (Figure 2.6). The overall grassland aboveground biomass trends for different grassland types were woody savanna > savanna > enclosed shrubland > grassland > open shrubland. As a result of several factors, the amount of disturbance of grassland above-ground biomass in Ethiopia tended to decrease overall, with significant decreases (10.3%) mainly in the western and central-eastern grassland areas (Figure 2.7), while in Kenya the amount of disturbance of grassland above-ground biomass tended to increase overall, with significant increases (10.5%) mainly in the eastern and southeastern grasslands (Figure 2.8).

Figure 2.5  Spatial and temporal variation in grassland above-ground biomass in Ethiopia

21

22

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

Figure 2.6  Spatial and temporal variation in grassland above-ground biomass in Kenya

Figure 2.7  Spatial and temporal variation in the disturbance of grassland above-ground biomass in Ethiopia

The Ethiopian PDSI indicates that frequent drought events occurred from May to September 2004, September 2005 to September 2006, September 2007 to April 2008, May 2009 to September 2010, November 2010 to May 2011 and May 2016 to April 2017. The occurrence of drought events, whether within or outside protected areas, or on the national scale as a whole, led to a

Chapter 2  SDG 2 Zero Hunger

significant decline in grassland above-ground biomass (Figure 2.9); particularly within protected areas, close to 80% of the area was affected by drought. In addition, representative areas affected by locusts in Ethiopia showed a clear downward trend in grassland above-ground biomass from May to September 2019, which coincided with the time period of locust disaster from June to August 2019, indicating that the occurrence of locust events further led to a decline in grassland above-ground biomass (Figure 2.10).

Figure 2.8  Spatial and temporal variation in the disturbance of grassland above-ground biomass in Kenya

As shown in Figure 2.11, the occurrence of COVID-19 in 2020 severely affected tourism in Kenya’s national parks. Without considering the effects of inter-annual drought, precipitation in Kenyan national parks and reserves was significantly higher in 2018 (852 mm) than in 2020 (783 mm). A comparison of grassland above-ground biomass in 2018 and 2020 shows that grassland above-ground biomass in Kenyan national parks and reserves in 2020 was significantly higher in the dry season, rainy season, and throughout the year than that in the same period in 2018, with annual-scale grassland extant biomass being 32.6% higher. This result suggests that the decline in tourism caused by COVID-19 has simultaneously contributed to the ecological recovery of African national parks.

23

Figure 2.9  Overlay of grassland above-ground biomass and PDSI drought indices for Ethiopia and protected grasslands

24 Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

Chapter 2  SDG 2 Zero Hunger

Figure 2.10  Grassland above-ground biomass in a representative locust-affected area of Ethiopia, from May to September 2019

Figure 2.11  Monthly changes in above-ground biomass and precipitation in grasslands in Kenya National Park Reserves in 2018 and 2020

25

26

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

Highlights  Grassland above-ground biomass and disturbance in Ethiopia, a representative East African country, both showed a slight increase in the last 20 years, while those in Kenya showed a slight decrease.  Drought and locust events are important reasons for the decline in grassland above-ground biomass in representative East African countries.  The decline in tourism caused by the COVID-19 pandemic has contributed to the ecological recovery of African national parks.

Outlook The analysis of the dynamics of grassland above-ground biomass and the factors influencing it can help improve the understanding of the balance between grass and livestock, the health of grassland ecosystems and the sustainable exploitation of grassland resources. This case study presents a comprehensive analysis of spatial and temporal trends in grassland above-ground biomass in Ethiopia and Kenya in East Africa, and the effects of drought, epidemic and locust on grassland above-ground biomass. Grassland above-ground biomass and its disturbance both show a slight increase in the last 20 years in Ethiopia and a slight decrease in Kenya. The occurrence of drought and locust events is the main cause of the decline in grassland above-ground biomass, posing a major threat to the health of grassland husbandry. Enhancing adaptive capacity to climate change and emergencies is an urgent issue to be addressed in the region’s grassland livestock sector. The decline in tourism caused by COVID-19 has contributed to the ecological recovery of African national parks. The future plan is to use the collection of more comprehensive international data sources to analyze spatial and temporal changes in grassland above-ground biomass across Africa, using the analysis of representative East African countries in this case study as a benchmark.

Chapter 2  SDG 2 Zero Hunger

2.3  Analysis of cropland area change in the 21st century in the Zambezi River basin Target SDG 2.4: By 2030, ensure sustainable food production systems and implement resilient agricultural practices that increase productivity and production, that help maintain ecosystems, that strengthen capacity for adaptation to climate change, extreme weather, drought, flooding and other disasters, and that progressively improve land and soil quality.

Background Zero Hunger is the second goal of the 2030 Agenda for Sustainable Development (Swaminathan, 2014). In Africa, arable land expansion is becoming the main way to increase food production due to complex land systems and food yields increasing slowly. In some regions, the primitive slashand-burn rotation/expansion approach is still used (He et al., 2012). The achievement of SDG targets in Africa is the most challenging, especially as the food security situation in some African countries is not encouraging. The Zambezi River basin is located in southern Africa and is the fourth largest river in Africa. The Zambezi River basin has a population of 32 million people, of which 70%-80% depend on agriculture and fishing for their livelihoods. Most of the eight riparian countries in the basin are in a state of severe poverty, with poverty rates of 71.4% in Malawi, 62.9% in Mozambique, 61.0% in Zambia, 49.1% in Tanzania, 30.1% in Angola, 22.6% in Namibia, 21.4% in Zimbabwe, and 18.2% in Botswana (World Bank, 2019). Most of the countries in the basin are facing serious food shortages. Zimbabwe, Mozambique, and Angola have food import dependency levels of 52%, 31%, and 55%, respectively. The food insecure population in Zambia is 64.1% of the total population, in Angola it is as high as 60.4%, and in Zimbabwe 48.9% of the population is in a state of food insecurity (Baquedano et al., 2020). This basin is a key area for achieving the global goal of Zero Hunger (SDG 2) by 2030. This case study therefore uses the Zambezi River basin as a representative area to analyze cropland area change with remote sensing data to achieve SDG 2 in Africa. In most areas of the Zambezi River basin, the ownership and distribution of land by tribal leaders are in conflict with modern state-led land tenure and management systems, resulting in dual land management systems in many countries; the fragmentation of arable land due to rapid land reforms implemented by the state also limits the efficiency of food production. The Intergovernmental Panel on Climate Change (IPCC) reported that the Zambezi River basin is showing a significant

27

28

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

warm-dry trend, exacerbating water stress in agricultural production in the basin (Dai, 2010). The proportion of irrigation in the basin is very low, and rainfed agriculture is dominant, so fallow and wasteland are evident, with high inter-annual variability in actual arable land area. Since regional variation in cropland has a strong influence on the food security situation, interannual variation in arable land in the basin is important data for a direct response to the regional food security situation and for assessing food supply. However, current studies on the changes, expansion trends and utilization patterns of cropland in the Zambezi River basin have mostly remained in a perceptual context, with insufficient understanding of the spatial and temporal patterns and driving mechanisms of cropland and utilization patterns in the basin. However, the Zambezi River basin is characterized by a diverse climate, complex topography, small and fragmented cropland patches, and rapid rotation of cropland due to slash-and-burn farming practices, making it extremely challenging to obtain accurate and detailed cropland distribution. In recent years, the remote sensing big data cloud platform represented by Google Earth Engine (Gorelick et al., 2017) and the new remote sensing data organization method represented by the Data Cube (Lewis et al., 2017) have integrated data and computational resources to reveal the fluctuating undulations of cultivated land. They provide new technical means to reveal the driving mechanisms behind the fluctuations of cropland and offer a new way to enhance the support of remote sensing for food security. Therefore, this case study aims to use cloud computing and remote sensing to map the spatial distribution of arable land in the region and the dynamic changes in arable land in the 21st century, which can provide basic background data for achieving regional sustainable development goals.

Data ●●

Satellite data: Landsat 5/7/8 multispectral data at 30m and Sentinel-2 multispectral data at 10 m (2000, 2005, 2010, 2015, 2020) for the Zambezi River basin..

●●

Cropland data: the CropWatch team from the Aerospace Information Research Institute of CAS provided cropland cover in the Zambezi River basin, 30 m resolution (2015, 2020).

●●

Population data: World Bank global population data (2000-2020).

Method By Google Earth Engine (Gorelick et al., 2017), a random forest approach was used to complete cropland mapping in the Zambezi River basin at 30 m resolution at five-year intervals from 2000 to 2010. Data collection and preparation: All available Landsat datasets from 2000 to 2020 were

Chapter 2  SDG 2 Zero Hunger

collected, containing Landsat-5, ETM+ and Landsat-8 surface reflectance datasets. After topography correction, radiation correction, de-clouding and shadowing, and consistency processing for the same bands from different sensors (Roy et al., 2016), an annual median synthesis was used to form a remote sensing dataset covering the entire Zambezi River basin. Classification feature generation: The classification used 12 features, including four raw bands (blue, green, red, and near-infrared) and eight indices (NBR, NDVI, NDSI, NDMI, GCVI, EVI, SAVI and LSWI) based on the raw waveform information. Preparation of the classification samples: The model was trained using ground sample points collected in the Zambezi watershed in 2015 and 2019, and then applied to the other three years (2000, 2005 and 2010). In the classification, the land cover was divided in this case study into six categories: arable, grassland, shrub, woodland, urban and water surface. The samples were divided into training samples (70%) and validation samples (30%). Classifier selection and parameter setting: The random forest classifier has a strong non-linear fit and has the advantage of not being easily overfitted; therefore, the case study selected random forest as the classifier with the number of trees set to 500 (Murillo-Sandoval et al., 2021), because after the number of trees is greater than 350, the accuracy no longer improves. The number of trees was increased to 500 in order to increase the robustness of the random forest. Accuracy validation: Three categories of producer accuracy, user accuracy and overall accuracy were selected to carry out the accuracy evaluation.

Results and analysis Based on Big Earth Data and Google Earth Engine, the spatial distribution of cropland at fiveyear intervals from 2000 to 2020 was obtained [Figure 2.12(a)-(e)], and verification shows that the overall accuracy is around 0.85. The area of cultivated land every five years from 2000 to 2020 is shown in Figure 2.12(f). During the 2000-2020 period, cropland showed a rising trend, then falling and rising. The area of cropland in 2000 was the smallest, and the area of arable land in 2020 was the largest. Results show that the cultivated arable area in the Zambezi River basin was 19.81 million ha in 2000 and 20.22 million ha in 2020, with an increase of 410,000 ha. The average cultivated area was 19.98 million ha, with a maximum fluctuation range of 410,000 ha and a fluctuation rate of about 2.1% in the cultivated area in the 21st century. According to the results of the linear fit, cultivated land had a rate of increase of 47,000 ha per year (slope of the linear fit) between 2000 and 2020. The area of cultivated arable land in the Zambezi River basin that shrank from 2000 to 2020 occurred mainly in Zimbabwe, Malawi and Zambia (Figure 2.13); while the area of cultivated arable land that increased was concentrated in Zambia, Zimbabwe and Botswana (Figure 2.14).

29

30

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

Figure 2.12  Spatial distributions of cultivated arable land in the Zambezi River basin from 2000 to 2020

Significant population growth was likely an important reason for the increase in cultivated arable land in this region from 2000 to 2020. Among the countries in the region, Angola had the highest natural population growth rate of 3.37% from 2000 to 2019, followed by Zambia and Mozambique with 2.73%, and Zimbabwe with the lowest natural growth rate of 1.05%. Due to the low yield in the region, a large amount of arable land needs to be reclaimed to feed a larger population. The expansion of arable land is evident in Zambia and Mozambique. Zimbabwe’s population in 2019

Chapter 2  SDG 2 Zero Hunger

Figure 2.13  Cropland shrinking area in the Zambezi River basin from 2000 to 2020

Figure 2.14  Cropland expansion area in the Zambezi River basin from 2000 to 2020

31

32

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

was 1.23 times larger than that in 2000, and the expansion of arable land was evident, while at the same time, the decline in arable land was most pronounced in Zimbabwe, with a possible primitive land-use pattern of cultivation, abandonment, migration and reclamation. At the same time, drought and floods may be major drivers of the dynamics of arable land in the basin, with 2015 being a year of low precipitation and the smallest area of arable land since 2000. The cropping pattern in the basin is dominated by rainfed agriculture, with less than 5% irrigated agriculture, and the frequent floods brought about by climate change in recent years have had a significant negative impact on food security in the basin. Therefore, in the Zambezi River basin, strengthening the capacity to adapt to climate change is key to improving the stability of food production in the region.

Highlights  Cropland was mapped at 30 m resolution in the Zambezi River basin every five years from 2000 to 2020.  Cropland in the Zambezi River basin has increased by 410,000 ha in the 21st century.  The increased area in the Zambezi River basin from 2000 to 2020 was concentrated in Zambia, Zimbabwe and Botswana.

Outlook To analyze the dynamics of cultivated areas in the 21st century, we mapped the distribution of cultivated areas in the Zambezi River basin at five-year intervals from 2000 to 2020 with the support of Big Earth Data. The cropland area in the Zambezi River basin increased by 410,000 ha compared with 2000. In Mozambique, the end of the civil war in 1994 led to a period of peaceful development. In recent years, agricultural development stimulus plans have been introduced to attract investment and develop agricultural corridors and irrigated agriculture, such as the Beira Agricultural Growth Corridor and the Pungewe River Basin Agricultural Plan (Droogers and Terink, 2014). But due to climate change and political instability, agricultural cultivation has fluctuated from year to year. The radical land reforms implemented in Zimbabwe since 2000 have led to a shift from largescale estate farming to small-scale smallholder farming (Peters, 2009), a significant decline in irrigated areas and an increase in the abandonment of arable land (Hentze et al., 2017), gradually transforming Zimbabwe from the breadbasket of southern Africa into a food-importing country.

Chapter 2  SDG 2 Zero Hunger

Conversely, thanks to good resource endowment conditions and stable agricultural policies, Zambia has shown a rapid expansion of arable land, but the dual ownership of land has increased the cost of agricultural development. The cropland in the basin is predominantly rainfed, with less than 5% irrigated, and widespread droughts caused by climate change pose a major threat to the stability of rainfed agricultural production in the area. The potentially arable land in the basin is huge, with only about 14% of the potentially arable land already developed, and there is great potential for arable land expansion in the basin. At the same time, strengthening the adaptive capacity to climate change is an urgent issue that needs to be addressed in the development of rainfed agriculture in the region.

33

34

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

2.4  Cropland dynamic change in the Mediterranean region from 2010 to 2020 Target SDG 2.4 By 2030, ensure sustainable food production systems and implement resilient agricultural practices that increase productivity and production, that help maintain ecosystems, that strengthen capacity for adaptation to climate change, extreme weather, drought, flooding and other disasters, and that progressively improve land and soil quality.

Background The Mediterranean region spans Asia, Europe and Africa, and contains 19 countries and regions, acting as a bridge between regions and an important part of the Silk Road Economic Belt and the Maritime Silk Road. The region is one of the birthplaces of human civilization, giving rise to ancient Egyptian, Babylonian, Roman, and Greek civilizations. The region is also one of the birthplaces of global religions such as Islam and Christianity. However, the region also has complex human-land and human-water conflicts, with arid areas accounting for 85.98% of the region, including 48.76% of the extremely arid areas (Zeng et al., 2021). Regional cropland is mainly concentrated in Europe part and Turkey, while North Africa and West Asia, where the population is rapidly growing, have relatively insufficient arable land resources. The arid and water-scarce North Africa and West Asia are the regions with the most rapid population growth. The shortage of water resources has led to North Africa and West Asia in the Mediterranean region becoming a challenge for achieving food self-sufficiency. The per capita crop production of all countries and regions in North Africa and West Asia in the Mediterranean region is below 400 kg and the crop production is not self-sufficient. Since crop production is closely related to the acreage of cultivated cropland, it is important to assess the changes in the acreage of cultivated cropland in the region from 2010 to 2020 to assess the achievement of the Zero Hunger goal of the SDGs in the region.

Data ●●

Remote sensing data: Landsat multispectral data for the Mediterranean region, 30 m resolution (2010, 2020).

●●

Food production data: FAO total food production data (2010-2019).

Chapter 2  SDG 2 Zero Hunger

●●

Population data: World Bank global population data (2010-2020).

Method By Google Earth Engine (Gorelick et al., 2017), a machine learning approach was proposed and designed to identify cultivated cropland of the Mediterranean region in 2010 and 2020. Remote sensing data collection and processing: all available Landsat datasets for 2010 and 2020 in the Mediterranean region were collected, including Landsat-5, ETM+ and Landsat-8 surface reflectance datasets. The imagery was processed for topography correction, radiation correction, and de-clouding and shadowing. The consistency of the surface reflectance data from different sensors was achieved by the method from Roy et al. (2016). Then, a median composite method was employed to generate the annual Landsat data for 2010 and 2020 covering the whole Mediterranean region. Generation of classification features: There were 20 features used to identify cultivated cropland in the Mediterranean region, including the reflectance of blue, green, red, near-infrared, shortwave infrared and shortwave infrared 2 bands in the annual Landsat data; brightness, greenness, and wetness generated by Tasseled Cap transformation; and the indices of NBR, NDVI, NDSI, NDMI, GCVI, EVI, SAVI and LSWI were generated by the reflectance bands. The elevation, slope and aspect were generated from SRTM DEM data. Collection of reference samples for classification: In this case study, the land cover of the Mediterranean region was divided into six categories: cultivated cropland, grassland, shrub, woodland, urban and water surface. Based on the above categories, the Copernicus CORINE land cover data from 2010 to 2019 in the European part of the Mediterranean region, and MODIS land cover and land use data from 2010 to 2019 in other regions, were used to generate the reference samples for classification as the following rules: the pixels with the same categories in CORINE and MODIS from 2010 to 2019 were regarded as the reference data for cultivated cropland classification. The samples were divided into training samples (70%) and validation samples (30%). Classifier selection and parameter setting: Many studies indicated that the random forest classifier has a powerful feature mining capability; therefore, the random forest classifier was selected for model training in this case study, according to the error change and the number of trees was set to 500 (Murillo-Sandoval et al., 2021). Accuracy assessment: The producer accuracy, user accuracy and overall accuracy were selected for the accuracy assessment of cultivated cropland. Analysis of the dynamic change of cultivated cropland: The method of zonal statistics was adopted in this case study to analyze the change in cultivated land area from 2010 to 2020.

35

36

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

Results and analysis Based on the Big Earth Data method mentioned above, the cultivated cropland areas in the Mediterranean region with 30 m spatial resolution in 2010 (Figure 2.15) and 2020 (Figure 2.16) were identified in this case study, and the accuracy verification showed that the overall accuracy was around 0.8. The remote sensing monitoring showed that cultivated cropland areas in the Mediterranean region were 181 million ha in 2010 and 178 million ha in 2020, shrinking by 3 million ha. The reduction mainly occurred in France, Morocco, and Algeria (Figure 2.17), while the increase in cultivated cropland in the Mediterranean region between 2010 and 2020 was mainly concentrated in Egypt, Tunisia, Libya, Jordan and other regions in West Asia and North Africa (Figure 2.18).

Figure 2.15  Cropland distribution in 2010 in the Mediterranean region

Significant population growth was an important reason for the increase in cultivated arable land in West Asia and North Africa from 2010 to 2020. These regions were characterized by a dry climate and a high proportion of irrigated cropland. The annual population growth rate in these regions ranged from 1.06% to 2.52% from 2010 to 2020, which was the fastest population growth rate in the entire Mediterranean region. In order to feed the new increased population, it is necessary to improve crop production by cultivating arable land. The annual population growth rate reached 2.32% in Syria; as the country with the most rapidly growing population

Chapter 2  SDG 2 Zero Hunger

Figure 2.16  Cropland spatial distribution in 2020 in the Mediterranean region

Figure 2.17  Cropland reduction areas from 2010 to 2020 in the Mediterranean region

37

38

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

Figure 2.18  Cropland expansion areas from 2010 to 2020 in the Mediterranean region

in the Mediterranean region, cultivated cropland in this country increased significantly by 32% from 2010 to 2020. The above-mentioned countries are facing a very serious water shortage due to the limitation of water resources and the land degradation pattern of cultivation-degradationabandonment-migration-recultivation widely observed in this region (Zeng et al., 2021). The rapid expansion of cultivated cropland increased the amount of water consumption, which inevitably exacerbated the water shortage in the region. Studies have shown that the population of the West Asia and North Africa region will double between 2015 and 2080 (Waha et al., 2017; Mohamed and Squires, 2018), and with further population growth, food production in the region will face greater water supply pressure in the future. Climate change was an important reason for the reduction of cultivated cropland in France and Morocco from 2010 to 2020. Morocco’s agriculture is dominated by rainfed agriculture, and the crops in Morocco in 2020 suffered a serious drought during the growing season, resulting in the shrinking of cultivated cropland planting area, declining crop yields, and reducing crop production by about 47% compared with that in 2019. In 2020, France also suffered a severe drought during the wheat growing season, resulting in the decline of cultivated cropland area and crop yield. French cereal exports declined by 25% compared with that in 2020. Therefore, in the Mediterranean region, strengthening the adaptive capacity to climate change is urgent for improving the stability of food production in the region.

Chapter 2  SDG 2 Zero Hunger

Highlights  Spatial distribution of cultivated cropland was mapped at 30 m resolution in the Mediterranean region in 2010 and 2020.  Pressure from population growth is the driving force behind the growth of cultivated cropland areas in North Africa and West Asia.  Climate change, especially drought, is an important cause of the reduction of cultivated cropland in countries such as Morocco.

Outlook This case study used a Big Earth Data method to comprehensively assess the dynamic change of the cultivated cropland acreage in the Mediterranean region from 2010 to 2020. There was 181 million ha of cultivated cropland in 2010 and 178 million in 2020, a reduction of 3 million ha. Population growth and climate change were important factors in this decline. Widespread droughts caused by climate change pose a major threat to the stability of rainfed agricultural production in the region, and strengthening the adaptive capacity to climate change is an urgent issue to address for the development of rainfed agriculture in the region. The rapid expansion of cultivated cropland in West Asia and North Africa due to rapid population growth is deeply affected by water shortages, and in the future, sustainable water resource use and the expansion of cultivated cropland for food production are challenges requiring balanced solutions in the region.

39

40

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

2.5  Monitoring desert locusts in Asia and Africa Target SDG 2.4: By 2030, ensure sustainable food production systems and implement resilient agricultural practices that increase productivity and production, that help maintain ecosystems, that strengthen capacity for adaptation to climate change, extreme weather, drought, flooding and other disasters, and that progressively improve land and soil quality.

Background Food security is the cornerstone of human survival, social stability and global sustainable development. In the context of climate change, the scope and prevalence of pests have expanded and increased. Locusts are a major migratory pest, and the desert locust is one of the most destructive locusts in the world. It has the characteristics of large food intake, strong reproductive ability and long flight distance. Monitoring desert locusts is of great strategic significance for reducing the application of chemical pesticides, ensuring food security and ecological security, and establishing a sustainable food production system. Since 2018, the abnormal climate has caused desert locusts to continue to multiply on the edge of the desert in the southern Arabian Peninsula, and gradually sweep the Horn of Africa and countries in Southwest Asia. In 2020, desert locusts continued to multiply and spread in the Horn of Africa, the southern Arabian Peninsula, and the coasts of the Red Sea. The desert locust infestations in Somalia, Ethiopia and Kenya are still serious, and many areas have became new breeding areas of desert locusts. As of 2021, although the insect swarms have declined, they are still active in the Horn of Africa, and some spread to northeastern Tanzania. The disaster of desert locusts in East Africa still needs worldwide attention. FAO has issued an early warning to the world, hoping that countries will be highly alert to locust plagues and adopt multi-country joint prevention and control measures to prevent pests from invading countries and creating serious food crises. Traditional visual hand-check single-point monitoring methods and limited-site meteorological prediction methods can only obtain information on the occurrence and development of pests at “points” and cannot meet the needs of large-area “surface” monitoring for timely prevention and control of pests. Remote sensing can efficiently and objectively monitor the occurrence and development of pests on a large scale in time and space. The rapid development of Earth observation technology in recent years has provided effective technical means for large-scale monitoring of locusts, and is highly effective in large-scale rapid guidance for pests. Scientific prevention and control and

Chapter 2  SDG 2 Zero Hunger

ensuring food security are of great significance. In addition, continuously updated encrypted meteorological station data and area meteorological parameter products formed by the coupling of remote sensing and meteorological data provide a richer source of information for the dynamic monitoring of locust occurrences. This case study uses multi-source data, combined with land use/cover data, meteorological data, ground survey data and the biological characteristics of migratory pests, their development and diffusion processes, and environmental factors to build a pest monitoring model. Big data analysis and processing on the Digital Earth Science Platform were carried out for the temporal and spatial distribution of desert locust reproduction and migration in the Horn of Africa and countries in Southwest Asia and locust plague monitoring in key countries. In this case study, intercontinental locust plague monitoring was carried out and the relevant results were provided to FAO to support the joint prevention and control in multiple countries and ensure the safety of agricultural production and regional stability in invaded countries.

Data ●●

Remote sensing data: MODIS data in Asia and Africa since 2000 (resolution: 500 m, https:// ladsweb.modaps.eosdis.nasa.gov/search), Landsat data (resolution: 30 m, https://earthexplorer. usgs.gov), Sentinel data (resolution: 10 m, https://scihub.copernicus.eu), planet data (resolution: 3 m) in a representative area of key countries, worldview data (resolution: 0.5 m), SMAP soil moisture data from 2010 to present (resolution: 0.25°, https://earthdata.nasa.gov/), greenness data (http://iridl.ldeo.columbia.edu/maproom/Food_Security/Locusts/Regional/ greenness.html) and rainfall data from GSMap (https://sharaku.eorc.jaxa.jp/GSMaP) in Asian and African regions since 2000.

●●

Meteorological data: long-term complete meteorological data of international meteorological stations from 2000 to present, tropical cyclone data and numerical weather prediction products for the Indian Ocean and Arabian Sea region from 2018 to present, meteorological numerical forecast products (https://www.nmc.cn/publish/typhoon/total cyclone.htm), and ECMWF climate assimilation data for the Asia-Africa region (https://www.ecmwf.int/en/forecasts/ datasets).

●●

Basic geographic information: global land use data (resolution: 10 m and 30 m)(http://www. geodata.cn), DEM, main crop planting areas in Asia and Africa (wheat, rice, corn, etc.)(https:// ipad.fas.usda.gov/ogamaps/cropcalendar.aspx), and administrative division data(https://www. tianditu.cn/).

●●

Other data: ground survey data released by FAO (https://locust-hub-hqfao.hub.arcgis.com), crop planting calendar (http://www.ipad.fas.usda.gov/ogamaps/cropcalendar.aspx.).

41

42

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

Method Quantitative remote sensing extraction and time-series analysis were carried out on elements (source, host, environment, etc.) closely related to the reproduction, development and migration of desert locusts. An indicator system was established for monitoring desert locusts via remote sensing, and a habitat suitability model was constructed. GIS spatial analysis, geo-statistics, spatiotemporal data fusion, global land use data, ground survey data and other multi-source data were employed to quantitatively extract locust breeding areas at a large scale. Ground survey data, crop planting calendars, regional distribution, global land use data and other multisource information were integrated with the locust spreading dynamic model to monitor locust migration paths. Then, combined with the plague monitoring model, the vegetation growth curve was analyzed for each key damaged country in the past 20 years, and the information on the locust damage was extracted to delineate the spatial range and area of the locust plague. Finally, fine-scale remote sensing monitoring was carried out for the hotspot countries and regions with locust damage, including damaged vegetation types (cropland, grassland and shrub), the spatial distribution of damage and total damage area.

Results and analysis At the end of 2020, desert locusts in East Africa and Southwest Asia were mainly distributed along the coast of the Red Sea (the eastern coast of Sudan and Eritrea, Saudi Arabia and the western coast of Yemen), the central Arabian Peninsula and the Horn of Africa (eastern Ethiopia, northern Somalia, and northern Kenya). In addition, there were scattered distributions in western Yemen, southwestern Iran, southeastern Kenya and the southeastern coast of Africa. In January 2021, affected by Cyclone Gati, locusts in northern Somalia continued to lay eggs, reproduce, and mature, and the locust swarms invaded Ethiopia, southern Somalia and Kenya. At the same time, they continued to invade northeastern Tanzania to the south. Locust swarms in northwestern Somalia and central Ethiopia invaded Djibouti and Eritrea to the north, and locust swarms in western Yemen spread northward along the coast of the Red Sea to the western coast of Saudi Arabia. In February 2021, the locust swarms in the Horn of Africa moved westward to Lake Turkana in northwestern Kenya, and the locust swarms along the coast of the western Red Sea of Saudi Arabia moved eastward to the central desert area. Locust swarms in southern Kenya invaded northern Tanzania to the south, and locust swarms in western and central Saudi Arabia continued to migrate eastward to the border with Kuwait. In March 2021, as ground control operations proceeded, the number of desert locusts in Ethiopia and Somalia continued to decrease. Locust swarms in Eritrea spread north along the coast of the Red Sea to the eastern coast of Sudan. Locust

Chapter 2  SDG 2 Zero Hunger

swarms in central Saudi Arabia invaded Kuwait with strong easterly winds and crossed the Persian Gulf to invade southwestern Iran. In April 2021, with continued ground control operations, the desert locust population in various countries in the Horn of Africa continued to decline. Locusts along the central and western coasts of Saudi Arabia continued to spread to Jordan and Syria with southerly winds, and reached the Euphrates River valley on the border between Iraq and Syria. The locusts in Jordan spread further to the western and central regions with southerly winds, entered western Syria to the north, and crossed the Anti-Lebanon Mountains into Lebanon. Due to control operations, the scale and number of locusts in the three countries were relatively small. In May 2021, the desert locusts in the Horn of Africa continued to lay eggs, hatch and form locust bands. The locusts in the spring breeding area in central Saudi Arabia continued to form immature adults and migrated toward southern Yemen. At the same time, some scattered adults appeared in central and southern Yemen. Locusts in most areas began to reproduce in the spring, but the ground control operations significantly reduced the number and scale of desert locusts compared with the same period in the previous year. With the spawning and maturity of the locust swarms in Iran, the locust swarms spread eastward to Pakistan from June to July 2021. The control operations and dry weather conditions helped gradually reduce the locust bands in the inland areas of northern Saudi Arabia, and some locusts migrated south to Yemen for summer breeding. At the same time, the Horn of Africa was affected by rainfall, and locusts in eastern Ethiopia and northern Somalia continued to lay eggs, reproduce and mature. From August to September 2021, locusts spread to northeastern Ethiopia for summer breeding (Figure 2.19). June to September 2021 is an important growing season or harvesting season for food crops in various countries. The continued rampage of desert locusts will cause serious threats to the agricultural production and national livelihoods of Asian and African countries. Since June 2020, Somalia, Ethiopia and Kenya in the Horn of Africa have been severely affected by locusts. In March 2021, Somalia’s cumulative newly damaged vegetation area was approximately 3.57 million ha. The hazard areas were mainly located in deserts on the border with Ethiopia. Desert locusts have reproduced for many generations in this area, especially in the northern Mudug and Togdale states (Figure 2.20). In April 2021, the cumulative newly damaged vegetation area in Ethiopia was approximately 6.97 million ha. The areas along the Great Rift Valley and its north and south ends were the most affected. The damaged vegetation area was also relatively large on the border between eastern Ethiopia and Somalia, and at the junction of southern Ethiopia with Somalia and Kenya. In May 2021, the cumulative newly damaged vegetation area in Kenya was approximately 3.39 million ha. The affected areas were mainly located in the breeding areas in the northeast, northwest and widespread areas in the central region. Rift Valley Province and Eastern Province have suffered a large amount of damage.

43

44

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

Figure 2.19  Core breeding areas and migratory paths of desert locusts (from Jan. to Sep. 2021)

Figure 2.20  Desert locust plague monitoring in Somalia, Ethiopia and Kenya from June 2020 to May 2021

Chapter 2  SDG 2 Zero Hunger

Highlights  We analyzed the spatial distribution of desert locust core breeding area and migration paths in Asia and Africa from January to September 2021.  We found that Asia and Africa were affected by locusts from June 2020 to May 2021. Somalia, Ethiopia, and Kenya were damaged seriously.  Desert locust dynamic monitoring in Asia and Africa could provide support for pest management.

Outlook In terms of technological innovation, this case study used international shared remote sensing data to conduct systematic research on the extraction of large-scale desert locust breeding areas, long-term quantitative monitoring of locust migration paths, and quantitative monitoring of locust plagues through big data analysis and processing in the Digital Earth Science Platform. The desert locust plague in Asia and Africa was monitored by remote sensing to gather updates on damage dynamics. The research results could contribute to the protection of agricultural production and food security, and provide important information support for locust plague emergency response. In terms of application and promotion, FAO and the Global Biodiversity Information Facility (GBIF) adopted the monitoring results on the core breeding areas and migratory paths of desert locusts in Asia and Africa from 2020 to 2021, as well as the plague monitoring results in key countries (Somalia, Ethiopia and Kenya). The products of this case study provide information support for multi-country joint prevention and control of pests to ensure agricultural production.

45

46

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

2.6  Global data tracking of the CBD benefit-sharing of microbial resources Target SDG 2.5: By 2020, the genetic diversity of seeds, cultivated crops, bred and domesticated animals, and the related wild species was maintained through establishing well-managed and diversified seed and plant banks at the national, regional and international levels; benefits from the use of genetic resources and related traditional knowledge were obtained according to internationally agreed principles and shared fairly and equally.

Background SDG target 2.5 calls for building a biological resource bank to protect biodiversity and share the benefits from the utilization of genetic resources and the related traditional knowledge fairly and equally. Microbial resources are the cornerstone of sustainable economic and social development and an important guarantee of national ecological security. However, environmental damage and other factors are threatening the survival of biological and genetic resources. Lacking sound access and benefit-sharing systems, illegal collection of these resources has led to serious losses. In order to prevent the abuse of biological resources, CBD in 1992 made it clear for the first time that biological and genetic resources are under the jurisdiction of national sovereignty, and it stipulated that obtaining biological and genetic resources must be based on the prior informed consent of the providing country and that under mutually agreed-upon conditions, the benefits generated by using the biological and genetic resources should be shared fairly and equally with the resource provider. For the “access and benefit-sharing” principle to be practical, the parties to the Convention reached the Nagoya Protocol on Access to Genetic Resources and the Fair and Equitable Sharing of Benefits Arising from Their Utilization to the Convention on Biological Diversity (CBD) in 2010. Microbial resources are tracked during gathering, collection, transnational transfer, academic and commercial application and benefit-sharing in order to reinforce the management of access to biological resources and benefit-sharing and to protect and utilize biodiversity. This process could provide technical support and information support for China’s adherence to the convention and give experience and reference for the transnational transfer and benefit-sharing of global biological and genetic resources. This case study used a big data platform of biological and genetic resources to gather data on 520,000 strains of microorganisms from 141 partners in 51 countries. This case study integrated

Chapter 2  SDG 2 Zero Hunger

data on microbial resources, genomics, literature, patents, and so on in the global microbial field since 1953. The datasets were continuously updated to develop data mining tools, monitor and track information on the utilization of microbial resources, and form a knowledge base with more than 4 billion data items. Data were associated between the “Analyzer of Bio-Resources Citations” and a strain information platform, and papers and patents on the generation of microbial genetic resources were explored and analyzed to support benefit-sharing in accordance with “fair and equitable sharing and utilization” stipulated by CBD.

Data ●●

Data bank of the World Data Centre for Microorganisms (WDCM): Global Microbiological Culture Collection Center which collects microbial data.

●●

Analyzer of Bio-Resources Citations: data on publications of microbial genetic resources in China (2001-2020).

●●

PubMed: data on journals in the biomedical information retrieval system developed by the National Center for Biotechnology Information (NCBI) in the United States (2001-2020).

●●

WIPO IP Statistics Data Center: data on Chinese patents of microbiological collections (20012019).

Method Through the Global Microbiological Culture Collection Center and users in science and business circles, microbial resources were obtained and utilized within the benefit-sharing framework of the Nagoya Protocol to establish an information platform. With the developed bio-resource citation analysis being used, the global usage of microbial resources on the monitoring platform for the transnational transfer of microbial resources was explored and counted, providing technical support and information support for China’s adherence to the convention and experience with the transnational transfer and benefit-sharing. The status quo of global benefit from Chinese microbial resources was revealed by exploring and analyzing papers in English journals published in China and elsewhere from 2000 to 2020. The publications were on microbial genetic resources of three major international collection institutions: China General Microbiological Culture Collection Center (CGMCC), China Center for Type Culture Collection (CCTCC) and Guangdong Microbial Culture Collection Center (GDMCC).

47

48

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

Results and analysis According to the statistics of WDCM, as of June 2021, a total of more than 3.3 million strains of various microorganisms had been collected by 805 microbiological culture collection institutions in 78 countries and regions. By 2020, 116,875 strains of microorganisms had been collected in CGMCC, CCTCC and GDMCC. As shown in Figure 2.21, in terms of the collection of patented strains, China showed a trend of steady growth, with the cumulative number of patented strains collected from 2001 to 2019 at 31,386, accounting for 39.86% of the global number of patented microorganism strains collected. From 2008 to 2019, China’s annual increase in the number of patented strains ranked first in the world for 12 consecutive years.

Figure 2.21  Status of deposits of patented microorganisms in China from 2001 to 2019

According to the statistical data of the Analyzer of Bio-Resources Citations of WDCM, from 2001 to 2020, 55 countries and regions around the world published a total of 3,223 scientific research papers referring to microbial resources of China’s three major collection institutions (Figure 2.22). Overall, Asian, North American and European countries utilized Chinese microbial resources to a higher degree (Figure 2.23). Among the top 10 countries in publishing the papers, the United States, Republic of Korea, Germany, Britain, Japan, India, Spain, Saudi Arabia and the Netherlands accounted for 24.75% of the total.

Chapter 2  SDG 2 Zero Hunger

Figure 2.22  Status of global annual publication of papers referring to Chinese microbial resources from 2001 to 2020

Figure 2.23  top10 countries of global publications referring to Chinese microbial resources from 2001 to 2020

49

50

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

Highlights  China is at the forefront of microbial resource collections. In terms of the collection of patented bacteria, from 2008 to 2019, the annual newly added quantity of collections of patented bacteria in China ranked first in the world for 12 consecutive years.  The benefit-sharing data of China’s microbial resources in the past 20 years show that Asian, North American and European countries have utilized China’s microbial resources to a higher degree. The cases of benefit-sharing of microbial resources in China provide technical support and information support for the adherence to China’s Convention on Biological Diversity and experience and reference for the transnational transfer and benefit-sharing of global biological and genetic resources.

Outlook By using the big data platform of biological and genetic resources, this case study realized the tracking and monitoring of all gathering, collection, transnational transfer, academic and commercial application and benefit-sharing to assess the global utilization of Chinese microbial resources and provide information support for China’s adherence to the Nagoya Protocol. In this case study, three authoritative Chinese international collection institutions of microbial resources were selected as the main objects of analysis, though they do not represent all collection institutions of microbial resources in China. However, the results of this case study are an important reference for the benefit-sharing of microbial resources at the national level. The data show that China’s collection of microbial resources is abundant, which reflects the high degree of gathering and collection in the country; but in the global sharing and utilization, it is reflected that China is the main country, followed by Asian, North American and European countries, which proves that the operation mode of microbial resources in China needs to be ameliorated, and the level of service in resource acquisition, identification, research and development, and sharing needs to be improved. In addition, the fair and equitable sharing and utilization of microbial resources were analyzed in this case study from the perspective of academic papers that published reference microbial resources. As the microbial industry has been positioned as a strategic emerging industry by major economies, and its position in the social economy is increasingly prominent, in the future, we can consider exploring the transnational transfer and global use of microbial resources based on multisource data of patents and digital sequence information, which will provide a reference for the benefit-sharing of global microbial resources.

Chapter 2  SDG 2 Zero Hunger

2.7  Food security and agricultural sustainability in major regions of Asia, Europe and Africa Targets SDG 2.1: By 2030, end hunger and ensure access to all people, in particular the poor and people in vulnerable situations, including infants, to safe, nutritious and sufficient food all year round. SDG 2.2: By 2030, end all forms of malnutrition, including achieving, by 2025, the internationally agreed targets on stunting and wasting in children under 5 years of age, and address the nutritional needs of adolescent girls, pregnant and lactating women, and older persons. SDG 2.4: By 2030, ensure sustainable food production systems and implement resilient agricultural practices that increase productivity and production, that help maintain ecosystems, that strengthen capacity for adaptation to climate change, extreme weather, drought, flooding and other disasters, and that progressively improve land and soil quality.

Background The selected regions in Asia, Europe and Africa examined in this case study refer to East Asia, Southeast Asia, South Asia, Central Asia, Central and Eastern Europe, West Asia, and North Africa. The study area is composed of 65 countries (see Attached Table at the end of the section), most of which are developing countries or emerging economies. These regions are also characterized by large populations, weak social and economic performance lagging behind other regions in the world, fragile ecological environments, widespread hunger and malnutrition, and grim prospects of food security. Against this backdrop, a common task has emerged for all countries in these regions: find the means to ensure food security and agricultural sustainability, end hunger and reduce malnutrition. It has also been considered an indispensable part of building the Belt and Road Initiative. Every year, many organizations, including the FAO, monitor, evaluate and issue early warnings about the world’s food security situations and crises, including hunger and malnutrition. However, there is still a lack of a holistic and comprehensive study regarding the problems in the study area. For this reason, it is important to review the current status of hunger and malnutrition, gain a full understanding of the food security status, predict future food demand, assess the potential to increase cereal production, identify the constraints on agricultural sustainability and seek effective

51

52

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

responses in these regions. These will play a significant role in enhancing agricultural exchange and cooperation between China and other countries in these regions, helping achieve the SDG of Zero Hunger, tackling malnutrition, and promoting sustainable and high-quality agricultural development in these regions. In this case study, we used land cover data collected through remote sensing technology and statistics from multiple sources to comprehensively examine and assess the food security situation, including hunger and malnutrition, future food demand trends, potential to increase cereal production, constraints on agricultural sustainability and corresponding policy options.

Data ●●

European Space Agency (ESA) CCI-LC land cover data (1992-2018).

●●

ESA MODIS land cover data (2001-2018).

●●

Data on World Food Security and Nutrition (FAO et al., 2020) (2019).

●●

FAOSTAT data on agricultural output (1961-2019).

●●

World Bank Open Data on population, GDP, and other factors (1960-2019).

Method Two basic indicators, namely the number of undernourished people and the prevalence of undernourishment, were used in our analysis of the food security in the study areas in Asia, Europe and Africa to measure the level of hunger or malnutrition in these regions. The prevalence of undernourishment (POU) was calculated through the density function: POU = ∫

x 25 km2) was estimated in 2010, 2015 and 2020 to show the clearness of water bodies in the past 10 years. This case study provides a new remote sensing monitoring dataset of surface water clarity for SDG indicator 6.3.2.

81

82

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

Data ●●

Remote sensingdata: MODIS Terra data, 500 m spatial resolution (2010, 2015, 2020).

●●

In-situ data: field data of water clarity of surface water bodies in China; shared measured datasets of SDD acquired from the European Multi Lake Survey dataset and United States AquaSat dataset.

●●

Preliminary geographic information: global coastal zone vector data.

Method This case study used MODIS surface reflectance products as the main data source. The second atmospheric correction method was used to extract the water-leaving reflectance data from the surface reflectance data, and the automatic bimodal method was used to identify the water area. Then a calculation model was constructed for the Forel-Ule index and hue angle α, which was normalized based on the spectral response function of MODIS blue, green, and red bands. A lake water clarity retrieval model (Wang et al., 2020a) was developed based on the above information and validated by the acquired global datasets of representative surface water bodies. It was evaluated that the accuracy of the lake water clarity retrieval model was 70%. Subsequently, water clarity datasets were produced for lakes in the world larger than 25 km2 for 2010, 2015 and 2020 based on MODIS data. Finally, the spatiotemporal changes in lake clarity were analyzed based on the time-series datasets (Stephens et al., 2015). With the thresholding method, the surface water clarity was divided into six levels (Class Ⅰ: ZSD>4 m; Class Ⅱ: 2 m70). Only 27.83% of lakes in Africa were in an oligotrophic or mesotrophic state, showing that the problem of lake eutrophication is still severe. Eutrophicated lakes often have a frequent occurrence of cyanobacteria blooms. When cyanobacteria blooms occur, a large number of algal toxins will be released into the water body. The high concentrations of algal toxins will hurt residents or wild animals that come in contact with or drink the water. The AFAI index (Fang et al., 2018) was used to determine whether a lake had cyanobacteria blooms. In AFAI, the Rrc(Red), Rrc(NIR), and Rrc(SWIR) are the Rayleigh corrected reflectances of the red, near-infrared, and shortwave infrared bands. When AFAI>0, it was judged that blooms had occurred in the water pixel. AFAI = Rrc (NIR) − Rrc (Red) + ( Rrc (SWIR) − Rrc (Red) ) × 0.5 If mean TSISDD is greater than 70 or cyanobacteria blooms occurred, the lake was defined as a “severe eutrophication lake”, and a list of such lakes in Africa was generated (Table 3.5). The list contains information such as the names, latitude and longitude coordinates, which can provide a basis and reference for cooperation between China and the environmental protection departments of various African countries in the protection or governance of these lakes.

93

94

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

Table 3.5  List of “seriously eutrophicated lakes” in Africa Lake name

N

E

TSI

Bloom

-21.228300°

24.859300°

Botswana

206.73

Yes

Gnagna

12.617600°

-0.126634°

Burkina Faso

198.80

Yes

Lac de Sian

13.101300°

-1.208420°

Burkina Faso

198.36

Yes

Kadiogo

12.129200°

-1.323610°

Burkina Faso

180.29

Yes

Sanmatenga

13.067800°

-0.984088°

Burkina Faso

178.90

Yes

Seno

13.950800°

0.292395°

Burkina Faso

170.12

Yes

Gourma

12.186700°

0.323407°

Burkina Faso

168.55

Yes

Lake Dadin Kowa

10.537700°

11.500200°

Nigeria

167.16

Yes

-32.546700°

24.712100°

South Africa

166.36

Yes

8.461530°

4.669670°

Nigeria

163.55

Yes

Kaya

13.095100°

-1.070480°

Burkina Faso

162.62

Yes

Sanmatenga

12.836600°

-1.027080°

Burkina Faso

161.33

Yes

Victoria West

-31.406200°

23.097000°

South Africa

159.13

Yes

Lake Dem

13.199900°

-1.143950°

Burkina Faso

157.90

Yes

Sokoto

13.534300°

5.929790°

Nigeria

153.04

Yes

Kadiogo

12.303500°

-1.331260°

Burkina Faso

152.32

Yes

8.365930°

-13.203800°

Sierra Leone

151.29

Yes

Tera

14.022900°

0.729371°

Niger

147.43

Yes

Namentenga

12.625800°

-0.275947°

Burkina Faso

143.99

Yes

Chad

13.077083°

14.527083°

Chad

＜ 70

No

Tanganyika

-5.91118°

29.185417°

The Democratic Republic of Congo

＜ 70

No

-14.417702°

35.236458°

Malawi

＜ 70

No

Kivu

-2.48862°

28.892849°

Rwanda

＜ 70

No

Mai-Ndombe

-2.715432°

18.177779°

The Democratic Republic of Congo

＜ 70

No

Bangweulu

-11.430955°

29.8145°

Zambia

＜ 70

No

Chilwa

-15.297917°

35.714583°

Malawi

＜ 70

No

Ntwetwe Pan

-20.938467°

25.633917°

Botswana

＜ 70

No

Sua Pan

-20.939583°

25.952083°

Botswana

＜ 70

No

30.413404°

32.363542°

Egypt

＜ 70

No

Boteti

Aberdeen Plain Unilorin lake

Guma Lake

Malawi

Great Bitter

Countries

Chapter 3  SDG 6 Clean Water and Sanitation

Highlights  This case study produced the first 30 m remote sensing-retrieved lake SDD product of Africa.  A trend in the annual average change in lake transparency was identified in Africa.  The water quality of African lakes was evaluated based on remote sensing technology, and a list of “seriously eutrophicated lakes” was generated.

Outlook Water transparency is one of the basic parameters to describe the optical properties of water, which can directly reflect the clarity and turbidity of water. This study applied the RBRG model to Africa’s long-time series Landsat imagery. The remote sensing inversion mapping of African lake transparency at a resolution of 30 m was realized, producing the first lake transparency product in Africa for the base year from 1985 to 2020. The overall transparency status of water bodies in African lakes shows a significant trend of increasing, which means that the water quality of lakes tends to improve. However, 72.17% of the lakes in Africa are still eutrophic, and the water quality of African lakes still faces challenges. Africa’s economic and social development is relatively backward compared to several other continents, and lake water quality assessment is limited by its limited field monitoring capabilities. Monitoring water quality in African lakes based on remote sensing technology has essential research significance and application value. However, due to the low temporal resolution of Landsat, only interannual lake water quality assessment can be achieved based on Landsat. In future research, it is necessary to use remote sensing data with a higher temporal resolution to gain quarterly, monthly, and ten-day-scale monitoring of lake water quality to better serve the water quality assessment and early warning of African lakes.

95

96

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

3.4  Assessment of changes in global crop water-use efficiency Target SDG 6.4: By 2030, substantially increase water-use efficiency across all sectors, ensure sustainable withdrawals and supply of freshwater to address water scarcity and substantially reduce the number of people suffering from water scarcity.

Background SDG indicator 6.4.1, change in water-use efficiency over time, measures national water-use efficiency in order to help address issues that must be overcome before SDG target 6.4 can be met. It involves water-use efficiency in sectors ranging from agriculture to industry to services. Agriculture has high water use and consumption (consumed through evapotranspiration). Improving agricultural water-use efficiency is important for promoting the sustainable use and development of water resources. A commonly used evaluation indicator of agricultural water-use efficiency is crop water-use efficiency (WUE), which refers to the biomass yield per unit of water and can reflect the wateruse efficiency from the output. Time series of crop water use efficiency (WUE) estimated from Big Earth Data and combined with models can provide scientific support to the assessment of agricultural water-use efficiency and its temporal changes. The spatiotemporal coverage, timeliness, and update frequency of such methods and datasets are significantly better than evaluation methods based on statistical data. In order to target the inadequacy of data for evaluating agricultural water-use efficiency, this case study developed alternative indicators to evaluate the interannual change in crop water-use efficiency in different regions of the world.

Data ●●

Remote sensing data: global datasets derived from different satellite observations with various spatiotemporal scales in 2001-2019, including albedo, normalized difference vegetation index (NDVI), leaf area index (LAI), fractional vegetation cover, snow cover, and land use/ land cover (LULC) from MODIS and Global Land Surface Satellite (GLASS); dynamic water surface area from Aerospace Information Research Institute (AIR) under CAS; global precipitation measurement (GPM) precipitation; European Space Agency-Climate Change Initiative (ESA-CCI) soil moisture and LULC; fraction of absorbed photosynthetically active

Chapter 3  SDG 6 Clean Water and Sanitation

radiation (FAPAR) from Copernicus Global Land Service (CGLS); Shuttle Radar Topography Mission (SRTM) DEM data (2000). ●●

Meteorological forcing and other spatial data: ECMWF ERA5 forcing data in 2001-2019; soil texture data.

●●

Ground measurements of latent heat flux and CO2 flux from the global flux tower network, used for calibration and validation.

Method Crop WUE is calculated as the ratio between crop net primary productivity (NPP) and water consumption (i.e., evapotranspiration, ET). The method to estimate crop WUE using the aforementioned data is summarized as follows. First, the ET was calculated by applying the ETMonitor model (Hu and Jia, 2015; Zheng et al., 2019) to the corresponding multi-source remote sensing data and atmospheric reanalysis data ERA5. The ETMonitor model distinguishes the energy partitioning and water fluxes between soil evaporation and vegetation transpiration upon theories of energy balance, water balance and plant physiology of soil-vegetation canopy systems. Second, NPP was estimated by the difference between GPP and respiration. A model for the crop GPP based on light-use efficiency (Du et al., 2022; Field et al., 1995; Zwart et al., 2010) was modified in two aspects: (1) soil water stress factors were introduced to improve the estimation accuracy of GPP under drought conditions; and (2) model parameters were calibrated and optimized by using the GPP obtained from the carbon flux data observed from the global eddy-covariance flux tower stations. By comparing the GPP observed by flux tower stations, it was found that the improved light-use efficiency model significantly improved the estimation accuracy of GPP. Third, the estimated daily ET and annual GPP of global crops were validated by comparing them with ground measurements. Finally, the time series of global NPP and ET from 2001 to 2019 with 1 km resolution were generated for the assessment of long-term changes in crop water-use efficiency.

Results and analysis There were some spatial variations in the global trend of crop water-use efficiency from 2001 to 2019. In Asia, America, and Oceania, crop water-use efficiency showed a consistent increase. In Europe and Africa, the trend generally was upward, except in a few countries and regions. China and Canada had the most significant increases in crop water-use efficiency (Figure 3.17). Crop water-use efficiency has increased by 16.4% on average in the past 20 years [Figure 3.18(a)]. The water-use efficiency of rainfed and irrigated crops increased by about 16.0% and 20.2%

97

Figure 3.17  Spatial distribution of interannual trend of global crop water-use efficiency from 2001 to 2019

98 Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

Chapter 3  SDG 6 Clean Water and Sanitation

respectively, with the latter remarkably higher than the former [Figure 3.18(b)]. The recent rate of improvement in crop water-use efficiency was higher than the earlier period of 2001-2019, and the turning point came earlier for irrigated crops than for rainfed crops [as shown in Figure 3.18(b) where a positive anomaly for the former comes before that of the latter]. This is due to the joint effect of the increase in crop biomass and the decrease in water consumption caused by technological progress, economic and social development, and a certain degree of climate change. During the overall increase in crop water-use efficiency, there was a clear decline in 2015, probably caused by the global super El Niño event (occurring between October 2014 and April 2016). The event, the strongest since the 20th century, had three important attributes: long life cycle, high cumulative magnitude and high peak intensity. The regional droughts caused by it led to a significant decrease in crop biomass, which then pushed down crop water-use efficiency through its strong linkage with the interannual dynamic changes in crop water-use efficiency.

Figure 3.18  Statistical analysis of interannual variations of global crop WUE from 2001 to 2019

99

100

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

Highlights  A method was developed for the evaluation of global crop water-use efficiency based on multi-source remote sensing data and crop growth processes, and a global dataset from 2001 to 2019 was generated, providing an innovative method and data for monitoring and assessing SDG indicator 6.4.1 in spatially comparable agricultural areas around the world.  In the past 20 years, crop water-use efficiency in agricultural areas around the world has increased (16.4% on average), mainly attributable to the increase in crop biomass thanks to technological progress, economic and social development, and a certain degree of climate change.

Outlook In this case study, an evaluation method of crop water-use efficiency was developed based on multi-source remote sensing data combined with the crop growth process. Based on globally consistent and spatially comparable remotely sensed ET and NPP data, the global crop wateruse efficiency was estimated, and its interannual changes from 2001 to 2019 were extracted and analyzed. The product is instrumental in the accurate understanding of the historical evolution and current state of global crop water-use efficiency. It should be noted that high NPP of crop does not mean high grain yield. Agrometeorological disasters can result in inconsistencies between the spatiotemporal dynamic changes in crop water-use efficiency calculated on crop NPP and ET and the crop water productivity calculated on grain yield and ET. The factors affecting the transformation from crop NPP to grain yield are complex, especially because the sensitivity to agrometeorological disasters is different in various crop growth stages. Given the present severe impact of agrometeorological disasters, farmland management needs to be further improved so as to raise the conversion rate from crop NPP or biomass to grain yield (harvest index), which is necessary for the achievement of a higher level of food security and water resource security in 2030.

Chapter 3  SDG 6 Clean Water and Sanitation

3.5  Comparative analysis of the implementation of IWRM and water stress in Lancang-Mekong countries Target SDG 6.5: By 2030, implement integrated water resources management at all levels, including through transboundary cooperation as appropriate.

Indicator SDG 6.5.1: Degree of integrated water resources management implementation (0100).

Background The Lancang-Mekong River basin (LMRB), involving six countries: China, Myanmar, Laos, Thailand, Cambodia, and Vietnam, is the first region where China has made breakthroughs in promoting the Belt and Road Initiative and the concept of “a community with a shared future for humankind”. In March 2016, The first Lancang-Mekong Cooperation Leaders’ Meeting officially launched the process of Lancang-Mekong Cooperation (LMC). Water resource cooperation is one of the five priority areas. LMRB is at an important stage of economic development and poverty reduction. Rapid economic development has increased the supply of food and energy, putting pressure on regional water resources, environments and ecosystems. At the same time, the impact of climate change has also been amplified in the water resource system, increasing water variability and leading to more frequent and extreme floods and droughts in LMRB. In order to maintain sustainable economic and social development and alleviate the increasingly severe water stress, it is urgent to improve the way water is used and managed. It is necessary to implement and improve IWRM in the Lancang-Mekong countries to balance competing social, economic, and environmental needs and impacts on water resources in an effort to achieve the broader sustainable development goals and improve resilience to climate change. Lancang-Mekong countries need a clear understanding of the current level of water resources management and water stress. Sharing their experience, finding deficiencies, and designing future development directions are important contents of Lancang-Mekong water resource cooperation. Meeting these requirements will help improve the water governance capacity of Lancang-Mekong countries, promote relationships among basin countries and strengthen China-ASEAN relations. It also meets the development requirements of the Belt and Road Initiative. Clean Water and Sanitation (SDG 6) is one of the 17 SDGs, and one metric of progress is SDG

101

102

Big Earth Data in Support of the Sustainable Development Goals (2021): The Belt and Road

indicator 6.5.1, which tracks the degree of the implementation of IWRM by assessing the four key components of IWRM: enabling environment, institutions and participation, management instruments and financing. The World Resources Institute defines water stress as the ratio between total water withdrawals and available renewable surface water (Gasser et al., 2013). It is an index that can comprehensively reflect water risk and evaluate the degree of water competition among water users. This case study focused on six countries involved in LMRB. Through the analysis and comparison of water stress and the degree of implementation of IWRM in Lancang-Mekong countries, this case study examined the experience of IWRM, explored the correlation between IWRM and water stress, and put forward suggestions for IWRM in Lancang-Mekong countries based on the forecast of future water stress, so as to promote the sustainable utilization of water resources in the basin.

Data ●●

Evaluation results of SDG indicator 6.5.1 in Lancang-Mekong countries in 2017 and 2020 (UN-Water SDG 6 Data Portal: https://www.sdg6data.org/).

●●

Baseline water stress assessment results and water stress predicted in Lancang-Mekong countries from the World Resources Institute (www.wri.org/data/water-stress-country) (2020, 2030 and 2040).

Method SDG indicator 6.5.1 is the degree of integrated water resources management (IWRM), which is an average score of four key components, including enabling environment (policies, programs and laws underpinning IWRM), institutions and participation (status of institutions implementing IWRM), management instruments (management tools to support IWRM implementation), and financing (budget and investment for water resources development and management). Countries adopt the questionnaire method recommended in the “Integrated Monitoring Guide for SDG 6: Step-by-step methodology for monitoring integrated water resources management (6.5.1)” from the United Nations Environment Program. Countries employing the guide reference the evaluation index system of IWRM, formulate a questionnaire (including 33 items in four categories and eight subcategories), distribute it to stakeholders for statistical scoring, and obtain the final score after consultation and analysis. The score ranges from 0 to 100 and is divided into six grades, with 0-10, 11-30, 31-50, 51-70, 71-90 and 91-100 being very low, low, medium-low, medium-high, high and very high, respectively. This case study uses the IWRM implementation evaluation results of the 2017 and 2020 SDG 6.5.1 measurements released by the United Nations.

Chapter 3  SDG 6 Clean Water and Sanitation

Water stress is defined by the World Resources Institute as the ratio between total water withdrawals and available renewable surface water, which is an indicator that can comprehensively reflect water risk. Higher scores correspond to greater competition among water users relative to available surface water resources. The score of water stress is on a scale from 0 to 5, divided into five categories, including low (0-1, the ratio of withdrawals to available water, referred to as Ratio