Our Digital Earth Insights and Opportunities from the World Settlement Footprint Insights and Opportunities from the World Settlement Footprint Suggested citation: Our Digital Earth—Insights and Opportunities from the World Settlement Footprint This document is the property of the World Bank. It is permissible to copy and use any of the material in this report provided that the source is appropriately acknowledged. Further information is available from: © The World Bank 2023 Adaptations—If you create an adaptation of this work, please add the following disclaimer along with the attribution: This is an adaptation of an original work by The World Bank. Views and opinions expressed in the adaptation are the sole responsibility of the author or authors of the adaptation and are not endorsed by The World Bank. Third-party content—The World Bank does not necessarily own each component of the content contained within the work. The World Bank therefore does not warrant that the use of any thirdparty-owned individual component or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to re-use a component of the work, it is your responsibility to determine whether permission is needed for that re-use and to obtain permission from the copyright owner. Examples of components can include, but are not limited to, tables, figures, or images. This work is a product of the staff of The World Bank and the Global Facility for Disaster Reduction and Recovery (GFDRR) with external contributions. The findings, analysis and conclusions expressed in this document do not necessarily reflect the views of any individual partner organization of The World Bank, its Board of Directors, or the governments they represent. Although the World Bank and GFDRR make reasonable efforts to ensure all the information presented in this document is correct, its accuracy and integrity cannot be guaranteed. Use of any data or information from this document is at the user’s own risk and under no circumstances shall the World Bank, GFDRR or any of its partners be liable for any loss, damage, liability or expense incurred or suffered which is claimed to result from reliance on the data contained in this document. The boundaries, colors, denomination, and other information shown in any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. RIGHTS AND PERMISSIONS: The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Cover: WSF3D. Data: © DLR/EOC | Basemap: © DLR/EOC © OpenMapTiles © OpenStreetMap contributors © KlokanTech.com. OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 3 FOREWORD T he World Development Report 2021: Data for better lives takes a critical look at the evolving data landscape, from data’s unrealized potential to the challenges of ensuring that they can meaningfully contribute to the sustainable development objectives. The report ultimately calls for a new social contract to enable the use and reuse of data to create more economic and social value, while simultaneously striving for better data access and literacy, as well as providing safeguards against misuse, including discrimination and cybercrime. The report outlines the huge potential of both public and private intent data. These data can be used by governments and the public sector to enhance service delivery and prioritize scarce resources, and by the private sector to improve their operations, and by individuals to increase their agency and hold governments accountable. To reach this potential, these data must have adequate coverage and be of high quality, easy to access and combine with other sources of data, and safe to use and reuse. Earth observation data can help drive social, economic, and sustainable development. It is routinely captured for the entire globe, providing both spatial and temporal coverage in standardized manner across countries, with historical catalogs that can inform change detection analysis. High-resolution sensors provide increasingly superior granularity, and developments in machine learning and artificial intelligence are steadily increasing the accuracy of image-derived data which capture patterns in built-up areas, vegetation, land use, and environmental monitoring. This report introduces the World Settlement Footprint (WSF) datasets as a novel and important contribution to development data. It showcases clear examples of how the WSF datasets can be used to standardize many recurring use cases within the World Bank. And perhaps more importantly, it also provides readers with information on where to access these datasets or get support for integrating them into new use cases.By increasing data use and reuse for greater value and making data easily available in a safe-to-use manner, Earth observation data can provide important contributions to development. They will also bring us closer to realizing the World Development Report 2021’s new social contract for data that is guided by the principles of value, equity, and trust. Haishan Fu Chief Statistician of the World Bank, Director of the World Bank’s Development Data Group Co-Chair of the Bank’s Development Data Council Bukavu, in the Democratic Republic of the Congo, has expanded up steep slopes amid rapid urban growth. Photo: Adobe Stock/Katya Tsvetkova. OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 5 EXECUTIVE SUMMARY E arth observation is a crucial source of accurate and up-to-date information of Earth’s natural and manmade environments that are critical when planning for, responding to, and mitigating the effects of natural hazards. Satellites that regularly collect images of the entire globe combined—with machine learning algorithms to process them more efficiently—have the potential to provide timely, standardized, verifiable, and scalable information. This report focuses on the use of Earth observation to identify built-up areas exposed to natural hazards. It describes the World Settlement Footprint (WSF) suite of derived datasets, developed by the German Aerospace Center (DLR) in collaboration with the European Space Agency (ESA), the Google Earth Engine team, and the World Bank. These gridded datasets capture the extent of built-up areas from 1985–2015 and again for 2019, estimated building heights, impervious surfaces, and estimated population. The datasets leverage recent advances in technology to obtain particularly high accuracy and precision. The datasets combine multitemporal imagery from radar and optical satellites, which increase the accuracy, while delivery at 10–30 meters ensures high precision. Earth observation derived information is particularly useful for standardized and recurring World Bank operations. The report looks at several World Bank operations, and the key insights provided through analysis incorporating the various WSF suite products. To date, the WSF datasets are known to have supported more than 67 City Scans, 15 Country Climate and Development Reports, and nine Urbanization Reviews. This report outlines specific examples of how teams have used WSF datasets to better understand urban growth and form, disaster risk management, and public and environmental health. Information from Earth observation has the potential to unequivocally inform and standardize World Bank operations and thereby support informed development strategies and preventive disaster risk management. This report demonstrates how the WSF suite of tools can be used effectively to support analysis for operations, access the tools themselves, and integrate these and other earth observation datasets. OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 6 ACKNOWLEDGEMENTS W orld Bank Digital Earth team members Caroline Gevaert and Nuala Cowan led by Edward Anderson, Pierre Chrzanowski, and Nicholas Jones prepared this report. Additional members from the Digital Earth team included Grace Doherty, Mira Gupta, Mattia Amadio, and Alex Chunet contributed their review. Thomas Esch led the German Aerospace Center (DLR) team, which included Daniela Palacios Lopez, Elisabeth Brzoska, Mattia Marconcini, and Julian Zeidler Integrating the World Settlement Footprint (WSF) datasets into World Bank operations has been an effort supported by many teams. In this work, the Digital Earth Partnership has collaborated closely with DLR, the European Space Agency Global Development Assistance partnership and the Geospatial Operations Support Team (GOST). The team expresses gratitude for the leadership of GFDRR manager Niels Holm-Nielsen and Data Development Group Data Analytics and Tools unit manager Keith Patrick Garret for their support of this initiative. The content of this report showcases several use cases and projects from teams throughout the World Bank, each of whom contributed to the insights of this report and to the development of the WSF as a digital public good. The team would especially like to thank and recognize the efforts of: Aanchal Anand, Antonios Pomonis, Benjamin Stewart, Brian Blankespoor, Giuseppe Molinario, Harris Selod, Hoguen Park, Joaquin Munoz Diaz, Johanna Lee Belanger, Jun Rentchler, Klaus Deiniger, Lande Bosch, Laurent Corroyer, Luc Marius Jacques Bonnafous, Maria Edisa Soppelsa, Mathilde Lebrand, Msilikale Msilanga, Nancy Lozano Gracia, Niina Käyhkö, Olivia D’aoust, Paolo Avner, Peter Kristensen, Rashmin Gunasekera, Ross Marc Eisenberg, Sameh Whaba, Somik Lall, Steve Rubinyi, Thea Hilhorst, Vivien Deparday, Walker Kosmidou-Bradley, Yves Barthelemy, Fabio Cian, Chandan Deuskar, Jun Rentschler, Reda Aboutajdine, and Philippe Ciais. The Global Facility for Disaster Reduction and Recovery (GFDRR) funded this report and ULTRA Designs, Inc. designed it. The report was also supported by a DLR-funded three-month research visit of Thomas Esch (DLR) to the World Bank in Washington D.C., USA. OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 7 TABLE OF CONTENTS Foreword 3 Executive Summary 5 Acknowledgments 6 Introduction 9 1. Datasets 10 Key Features 14 2. Key Insights 17 Where are built-up areas expanding the most rapidly? 17 Where are built-up areas most exposed to natural hazards? 18 Where are people residing in areas exposed to natural hazards? 19 Is the most growth in hazard areas occurring in large or small cities? 20 3. Use Cases 21 Urban growth and form 21 Disaster risk management 30 Public and environmental health 34 Upcoming applications 36 4. The Way Forward 39 Appendix A: How to access the WSF datasets 40 Appendix B: Where to get help with the WSF datasets 41 Pakistan inundated. © Contains modified Copernicus Sentinel data (2019), processed by ESA, CC BY-SA 3.0 IGO. OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 9 INTRODUCTION G lobal area covered by human settlements grew by 85 percent in a mere 30 years from 1985–2015 (Rentschler 2022). During this period, very rapid and unplanned growth characterized urban areas in low- to middle- income countries (LMICs) in particular. Previously, city officials often lacked data regarding where such growth was occurring, and whether this growth was happening in areas exposed to natural hazards. Now, satellites from NASA and ESA’s Copernicus program, among others, regularly capture images of the Earth’s surface. Imagery can be processed by state-of-the-art machine learning algorithms to provide timely and accurate updates on how cities are changing. This greater spatial and temporal resolution of exposure data, as well as the timeliness of these data, facilitate a forward-looking approach to disaster risk management (DRM). The data enable planners to monitor the prevalent situation closely and identify trends allowing them to quickly respond to and mitigate future risks. Such proactive action is also more cost effective than corrective risk management. So why is Earth observation playing such a key role in the switch to the preventive DRM paradigm? First, the data are timely as satellites regularly capture images of the entire Earth’s surface. Second, EO produces standardized information. Data collected at a national level are often subject to different data collection methods, making it difficult to compare datasets across regions. It is possible to compare objectively the estimated population exposed to natural hazards between cities, countries, and continents by using EO data and standardized processing methods. This makes EO data particularly useful for portfolio management. Third, owing to the data’s objectivity, openness, replicability of the data processing workflows, and robust validation campaigns, EO-based data form an independent and verifiable input to support decision making. Fourth, satellite imagery has a historical archive and new imagery is constantly being acquired. It therefore provides continuity over time and enables prediction. Finally, EO is particularly scalable to national, regional, and global coverage and is particularly cost effective at larger scale and over multiple projects. Thus, EO-derived information forms a timely, standardized, scalable, and reliable way to characterize and obtain insights about the evolution of exposure, enabling a shift from corrective to preventive risk management. Recognizing these benefits, satellite-based settlement extent maps have been under development for some years. However, their utility was sometimes limited owing to misclassification errors —from both over and underestimation of settlement areas—derived inherently from the processing framework. The framework included employing only optical or radar imagery to classify settlements, not enough spatial detail, or restricted systematic updatability owing to enlisting commercial, and frequently expensive imagery data. In this context, the German Aerospace Center (DLR) in collaboration with the European Space Agency (ESA), the Google Earth Engine team, and the World Bank worked together to develop and validate the World Settlement Footprint (WSF) suite, a state-of-the-art collection of global settlement datasets based on open satellite imagery which provides unprecedented accuracy and detail. This report introduces the WSF suite and describes how the datasets can be accessed (Appendix A). It then lists features of the data and provides key insights on exposure that can be obtained with these datasets. Datasets demonstrate where built-up areas are expanding most rapidly and where growth is happening in areas exposed to flood, landslide, or earthquake hazards. The paper also describes recurring World Bank use cases that demonstrate how the WSF datasets can be used for global analytics, country-level analyses, and even city-level analytics when incorporating local data—all of which furnish forward-looking DRM and climate adaptation strategies. References Rentschler, J., Avner, P., Marconcini, M., Su, R., Strano, E., and Hallegatte, S. 2022. Rapid Urban Growth in Flood Zones: Global Evidence since 1985. Policy Research Working Paper; 10014. World Bank, Washington, DC. Available at: https://openknowledge. worldbank.org/handle/10986/37348 OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 10 1 DATASETS T he World Settlement Footprint (WSF) suite consists of gridded datasets that describe built-up areas and that estimate building density, building volume, and population (table 1.1). The quality of each WSF layer was verified through robust validation campaigns that compared the qualitative and quantitative accuracy of each dataset against independent reference data and other established datasets when possible.1 Validation campaigns have demonstrated that WSF-Evolution, WSF-2015, and WSF-2019 are consistently and systematically more accurate than other established global built-up layers, whereas WSF-3D and WSF-Imperviousness are new datasets for which no comparable global datasets are available. Built-up areas WSF-2019 (Jakarta) WSF-2015v2 (Jakarta) OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 11 Built-up areas (cont.) WSF-Evolution 1985–2015 (Jakarta) Urban Centers Building heights WSF-3D (Jakarta) OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 12 Impervious surfaces WSF-Evolution imperviousness 1985 (Jakarta) Population WSF-2019-Population (Jakarta) OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 13 Table 1.1. WSF-Suite: Product specifications Spatial Characteristic Dataset name resolution Coverage Competitive advantage Publication WSF-2019 Validation data—collected through a WB Digital Public Marconcini Binary mask outlining the extent of Works program—demonstrated that the average et al., 2021 human settlements globally for the accuracy in Africa of 84.8% outperform other publicly 10mx10m Global year 2019. available datasets: ESRI-2020 (83.0%), WC-2020 (80.0%), GHSL-2018 (79.1%), GAIA-2018 (72.0%), and GISA-2019 (73.2%). WSF-2015v2 Validation data demonstrated that the average accuracy Marconcini Binary mask outlining the extent of in Africa of 83.2% outperform other publicly available et al., 2020 human settlements globally for the 10mx10m Global datasets: GUF (79.5%), GHSL (70.4%), and GLC30 year 2015. (67.4%). Built-up areas WSF-Evolution Validation data collected with support from Google Marconcini Binary mask outlining the extent of demonstrate that WSF-Evolution consistently et al., 2021 30mx30m Global human settlements on a yearly basis outperforms competitors such as GAUD, GAIA, GISA, and from 1985 to 2015. GHSL (only 1990, 2000, 2015). Urban Centers The Urban Centers are based on data from 2019, whereas Esch et al., The detection of urban clusters in other datasets (e.g., Africapolis) represent the state 2021 2019, based on built-up density, in 2015. Although no explicit validation was conducted cluster area, and the number of [vector for the WSF urban centers, they were obtained from Africa inhabitants. The dataset contains the data] WSF-2019, which has proven to be more accurate than following information for each cluster: comparable datasets. population, area, population density, settlement extent 1985-2015. WSF-3D The WSF-3D dataset is the first dataset that provides Esch et al., The datasets quantify the fraction detailed quantification of the fraction, total area, 2022a Building (BF), total area (BA), average height average height, and total volume of buildings at a global 90mx90m Global heights (BH) and total volume (BV) of buildings scale and unprecedented resolution. Other citywide in 2015. datasets of building height must be collected from other sources, such as aerial LiDAR surveys. WSF-Imperviousness Unlike most existing global human settlement datasets, Esch et al., Quantifies the percent of impervious the WSF-Imperviousness provides 2D built-up density 2018 surface within each 10mx10m information at a spatial resolution of 10 m in the form settlement pixel reported by the of percent surface imperviousness. This layer also Impervious WSF2019. formed the basis for estimating the total built-up area 10mx10m Global surfaces as reported in WSF-3D. Methodology for the WSF- Imperviousness processing can also be used to model the vegetation fraction per built-up cell. An extensive global validation campaign is underway with the support of Google crowdsourcing technology. WSF2019-Population This will be the first open-and-free population dataset Palacios- Quantifies the number of people available at a spatial resolution of 10m at a global scale. Lopez et al. per pixel. It has been created using Compared to multilayer approaches of population (2020) population estimates and vector data modeling (e.g. those employed by WorldPop, LandScan), Population collected and harmonized by CIESIN 10mx10m Global the independent weighting framework provided by the for the year 2019. WSF-Impervious layer allows for an easy updatability and replicability of the generated population datasets, while at the same time, delivers qualitative results which are consistent across space. OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 14 KEY FEATURES ● Multi-temporal and multi-sensor analysis Compared to previous iterations of other established built-area layers—GUF, GHSL, GLC30—WSF2015 version 1 was the first dataset produced with a novel method that exploited the synergies between open-and-free multitemporal optical imagery and multitemporal radar satellite imagery. This had two main advantages. First, the use of images of the same area at different dates—besides an evaluation of the spectral signature—considers the temporal behavior of a surface. At the same time, data gaps owing to cloud coverage and shadow areas are avoided. This reduces misclassification errors that are present in single data scene acquisition conditions. Second, using only radar or optical imagery has several limitations. For example, bare soil tends to be misclassified as settlements in optical imagery, while complex topographies are frequently classified as settlements in radar data. The joint analysis of optical and radar imagery, as used by WSF-2015 and WSF-2019, reduces these ambiguities and therefore improves the identification of settlements on the ground. ● High precision WSF layers are produced with an increased spatial resolution of ≈10m and ≈30m at the Equator as opposed to the commonly available 250m and 1km datasets. One of the main advantages of this refinement is that it facilitates more options and higher flexibility to integration of the WSF- layers with other geospatial data that are available at similarly high granularities (e.g. flood data, climate models). The WSF-Evolution data covers a 30-year time span with an unprecedented annual updating frequency. The WSF technology has also successfully been used for quarterly updates of human settlements growth at 10m spatial resolution. Comprehensive validation campaigns demonstrated that WSF-Evolution, WSF-2015, and WSF-2019 are more accurate than other established global built- up layers. OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 15 ● Urban density and 3D settlement characteristics The WSF-Imperviousness defines the 2D built-up density globally at 10m spatial resolution in form of the percent imperviousness surface. The WSF-3D dataset is the first dataset that provides a detailed 3D quantification of the built-up density in form of the fraction, total area, average height, and total volume of the building stock at a global scale and unprecedented resolution of 90m x 90m (aggregated from a measurement at 12m spatial resolution). The availability of this dataset will bridge the data gaps that exist in countries where height information on building heights is not available or expensive to produce, enabling the performance of cross-comparative analyses at local, regional, national, and global scales. ● High-resolution population dataset The WSF-Population dataset (produced on the basis of the WSF-Imperviousness layer), will be the first open- and-free population dataset available at a spatial resolution of 10m at a global scale. Compared to multi-layer approaches of population modeling (e.g. those employed by WorldPop, LandScan), the independent weighting framework provided by the WSF-Impervious layer allows for an easy updatability and replicability of the generated population datasets, while at the same time, delivering qualitative results which are consistent across space. An improved version of the WSF-Population is envisioned, produced on the basis of the WSF-3D dataset. This dataset will be the first global dataset that integrates building height, building volume and building type information as a measure to allocate populations from administrative unit levels to grid cells of 10mx10m spatial resolution. The integration of this additional information has proven to greatly improve the accuracy of population models by up to 30% in highly urbanized areas, delivering population estimates with higher reliability than existing models. All details related to the outcomes of a validation campaign for Europe, including WSF 3D data are provided in Palacios-Lopez et al (2021). Note 1. See Table 1 and Esch et al. (2021) for details. References Esch, T., Bachofer, F., Heldens, W., Hirner, A., Marconcini, M., Palacios-Lopez, D., Roth, A., Üreyen, S., Zeidler, J., Dech, S., and Gorelick, N. 2018). Where We Live - A Summary of the Achievements and Planned Evolution of the Global Urban Footprint. Remote Sensing, 10(6), 895. https://doi.org/10.3390/rs10060895 Esch, T., Marconcini, M., Metz-Marconcini, A., Zeidler, J., Brzoska, E., Palacios Lopez, D., and Leutner, B. 2021. Satellite Monitoring Service of Urbanization in Africa: Final Report. DLR, Munich. Available for World Bank staff at the Development Data Hub: https://datacatalog.worldbank.org/int/data/dataset/0060310/satellite_monitoring_service_of_ urbanization_in_africa_world_settlement_footprint Esch, T., Brzoska, E., Dech, S., Leutner, B., Palacios-Lopez, D., Metz-Marconcini, A., Marconcini, M., Roth, A. andZeidler, J. 2022a. World Settlement Footprint 3D-A first three-dimensional survey of the global building stock. Remote Sensing of Environment, 270, 112877. https://doi.org/10.1016/j.rse.2021.112877 Esch, T., Palacios Lopez, D., and Brzoska, E. 2022b. Satellite Monitoring Service of Urbanization. Marconcini, M., Metz-Marconcini, A., Üreyen, S., Palacios-Lopez, D., Hanke, W., Bachofer, F., Zeidler, J., Esch, T., Gorelick, N., Kakarla, A., Paganini, M. and Strano, E. 2020. Outlining where humans live, the World Settlement Footprint 2015. Scientific Data, 7(1), 1-14. https://doi.org/10.1038/s41597-020-00580-5 Marconcini, M., Metz-Marconcini, A., Esch, T., and Gorelick, N. 2021. Understanding Current Trends in Global Urbanisation - The World Settlement Footprint Suite. GI Forum 2021, 9(1), 33–38. https://doi.org/10.1553/giscience2021_01_s33 Palacios-Lopez, D., Bachofer, F., Esch, T., Marconcini, M., MacManus, K., Sorichetta, A., Zeidler, J., Dech, S., Tatem, A.J. and Reinartz, P. 2021. High-resolution gridded population datasets: Exploring the Capabilities of the world settlement footprint 2019 imperviousness layer for the African continent. Remote Sensing, 13(6), 1142. https://doi.org/10.3390/rs13061142 Johannesburg—South Africa’s largest and most populous city, can be seen at the top of the image. Photo: © ESA, CC BY-SA 3.0 IGO. OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 17 2 KEY INSIGHTS Where are built-up areas Regions across the world are growing at different speeds. The rate of settlement expansion is slowing down in Latin expanding the most rapidly? America and the Caribbean (LAC), but steadily increasing Human settlements are growing rapidly. From 1985– in Sub-Saharan Africa (SSA) (figure 2.1). 2015, in just 30 years, the area of the globe covered by The WSF-suite shows relatively high growth rates for the human settlements grew by 85 percent (Rentschler period of 2010–2015 across countries in the Sahel region 2022). In 2019, the total global settlement footprint and East Africa (figure 2.1). Several provinces in northern was 707,832 square kilometers. Of this, 291,577 square Uganda grew by 40 to 100 percent in this period, more kilometers were covered with buildings that showed than twice the continental average of 13.2 percent (map an average height of 5.55 meters. The total worldwide 2.1) (GFDRR 2023). building volume is estimated at 1,632 cubic kilometers (Esch et al. 2022a). Figure 2.1. Yearly settlement growth rates from 1985-2015 per world region. 3.5% 3.0% 2.5% 2.0% 1.5% 1.0% 0.5% 0.0% 1985 1990 1995 2000 2005 2010 2015 East Asia & Pacific Europe & Central Asia Latin America & Caribbean Middle East & North Africa North America South Asia Sub-Saharan Africa Grand Total Source: Adapted from: Rentschler et al., 2022 OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 18 Map 2.1. Growth of settlement extent across Africa from 2010 to 2015 per ADM1 unit. n Country borders Growth rate (2010–2015) n 0 – 10% n 11% – 25% n 26% – 50% 500 km Source: GFDRR 2023. Note: The settlement extent of an average ADM1 region grew by 13.2% in this period. ADM1 is Administrative Unit 1 - the highest order administrative/enumeration unit for a country. Where are built-up areas most 15 centimeters in a one-in-hundred-year flood event. An estimated 72.4 million people, or 5.9 percent of the exposed to natural hazards? continent’s population, were in these areas exposed to Combining settlement extent and population datasets high flood hazard (GFDRR 2023). from the WSF with global hazard datasets provide In Africa, 373 out of 9,414 settlements had more than insights into the distribution of exposure to natural 70 percent of their settlement extent located in zones hazards such as flooding, earthquakes, and landslides. exposed to high flood hazard. 280 of these were in Egypt Globally, 145,000 square kilometers of settlement extent and another 51 in Nigeria and Somalia. The countries were located in zones with a high or very high flood risk. with the highest proportions of settlement extent in This is equal to 11 percent of the total settlement area in areas exposed to flooding are: Egypt at 32 percent, 2015 (Rentschler et al. 2022). In Africa, almost 10 percent South Sudan at 29 percent, Sudan at 24 percent, Chad of the 106,494 square kilometers of settlement extent in at 23 percent, and Mali at 21 percent (map 2.2) (GFDRR 2019 was located in areas at risk of flooding more than 2023). OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 19 Map 2.2. The proportion of built-up areas located in areas exposed to flood hazard in Africa. n Country borders n SE in high risk zone n SE in low risk zone Settlement Extent (km2) 15,000 5,000 2,000 500 km Source: GFDRR 2023. Note: The relative size of the symbol is proportional to the total settlement extent in that country in 2019. Where are people residing in areas of people in such zones were the Democratic Republic of Congo at 29.1 million or 23 percent, Ethiopia at 12.7 exposed to natural hazards? million or 27 percent, and Burundi at 6.7 million or 55 In Africa, 72.4 million people were exposed to high flood percent. hazard in 2019, which is 5.9% of the total population. Exposure to medium to high landslide hazard affected Most of these flood-exposed residents or 24.8 million 3.3 million people in Africa, equal to 1.1 percent of the lived in Egypt, 10.6 million in Nigeria, and 4.2 million in population of the respective countries (figure 2.2). Sudan (GFDRR 2023). The countries with the highest number of inhabitants In Africa, 80.1 million people were exposed to medium exposed to landslide hazard were Ethiopia with almost or high earthquake hazard, which is 7.3 percent of the 1.1 million, Sierra Leone with nearly 515,000 people, and total population. The countries with the highest number Rwanda with approximately 280,000. OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 20 Figure 2.2. A visualization of exposure to landslide hazard in Africa. Low Medium High Low High Source: Design by DLR. Note: The colors indicate the majority landslide hazard level within the hexagon and the height of the hexagon indicates the population living in that area. Is the most growth in hazard areas hazards has outpaced the growth into low flood hazard areas for medium—100,000 to a million—and large occurring in large or small cities? centers with more than one million people. For small Settlements across Africa numbered 9,414 in 2019. The settlements, growth into low flood risk zones outpaced vast majority of these settlements—91.0%, or 8,564 growth into high risk zones until 2014, after which the unique urban centers—were small, with a population growth into low and high risk zones shows similar speeds of between 10,000 and 100,000. Since 2000, the (figure 2.3) (GFDRR, 2023). settlement expansion in areas exposed to high flood Figure 2.3. Year-over-year growth rates into low flood hazard and high flood hazard zones for urban centers in Africa. 2.5 Year-over-year growth in settlement extent (%) 2.0 1.5 1.0 Small urban centers, risky growth Small urban centers, safe growth 0.5 Medium urban centers, risky growth Medium urban centers, safe growth Large urban centers, risky growth Large urban centers, safe growth 0.0 1990 1995 2000 2005 2010 2015 Source: GFDRR 2023. Note: Urban centers grouped into small (10,000-100,000), medium (100,000-1 million) and large (>1 million) settlements, 1985–2015. References GFDRR. 2023. Understanding the Changes in Africa’s Urban Exposure to Flood Hazards. In press. Rentschler, J., Avner, P., Marconcini, M., Su, R., Strano, E., Vousdoukas, M., and Hallegatte, S. (2023). Global evidence of rapid urban growth in flood zones since 1985. Nature 662, 87-92. doi: https://doi.org/10.1038/s41586-023-06468-9 OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 21 Moutsamudu, Comoros. Photo: urf. 3 USE CASES T he World Bank used the WSF suite of datasets engaged with the DEP team and DLR in the use of the within the first year of its release for a wide datasets. The total number of projects that use WSF range of its projects and operations. Based datasets is likely to be much higher in practice. on direct collaborations between DLR, Digital The following section explores key thematic focus Earth Partnership (DEP), and other World Bank teams, the WSF datasets are known to have supported projects areas where WSF datasets are being used, often in in more than 35 countries; and used in more than eight conjunction with locally available datasets. They provide global studies. Insights from the WSF have supported at insight and analyses at the global, country, and city level. least six World Bank reports. The datasets are also used The use cases are organized according to four themes: to support recurring country- and city-level analyses to (i) urban growth and form; (ii) disaster risk management; plan and support World Bank operations of more than (iii) public and environmental health, and (iv) upcoming 67 City Scans, over 15 Country Climate and Development applications that are technically possible with WSF data Reports, and in excess of nine Urbanization Reviews. but apparently have not yet been exploited for World These indicators represent teams that have directly Bank operations. OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 22 Kampala, Uganda. Photo: © Stelya | Dreamstime.com. Urban growth and form A novel feature of this methodology is its use of global data on building heights, specifically its use of the Pancakes to Pyramids: City Form to Promote WSF-3D layer to understand city height profiles. Height Sustainable GrowthCity growth management requires profiles indicate vertical layering; and layering pivotally that city leaders solve a three-dimensional problem. provides floor space together with urban density—the Preparation for urban growth mandates planners to combination that reshapes pancakes as pyramids, consider not only vertical layering and infill development enabling cities to be not only dense with people, but that will be enabled by future economic productivity, also thriving, livable, and sustainable (Lall et al. 2021). but also horizontal growth that will occur at the city’s Without accurate estimates of height data, the World edge. Using data to understand what makes a city Bank team could not confidently compare floor space grow outward, inward, or upward can help us better across cities. Nor could they track its distribution within understand the interplay between city density and public the geography of a single city (figure 3.2). transport and non-car modes of transportation, that will help cities reduce their climate footprints (figure 3.1). OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 23 Figure 3.1. Growth of a city along three margins—horizontal spread, influx development, and vertical layering. Buil-up area Defined by the presence of buildings Urban area Cities grow along three margins Horizontal spread Infill development Vertical layering Source: Derived from Lall et al., 2021. Figure 3.2. Average building height in Chicago, Illinois. 64m 1m Source: Derived from Lall et al., 2021. Note: The unit of measurement used in the figure is in meters (m). OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 24 Urbanization Reviews to consistently identify built-up areas over a country or region and the WSF Urban Centers enables settlements Today’s cities are growing at unprecedented and to be compared with regard to estimated population, challenging speeds, and local governments require the population density, and other statistics. WSF-Evolution right kind of diagnostic tools to identify investment shows how built-up areas are extending over time, priorities and inform policy. The World Bank’s identifying which cities show the most recent growth Urbanization Reviews offer a critical framework to and also helping to visualize geographical barriers such inform these challenging development decisions. Since as municipal boundaries or land conflicts that limit the 2021, WSF has already been used for the analytics expansion of settlements. The accuracy, granularity, of at least nine Urbanization Reviews, covering: and comparability across regions are clear benefits of Burundi, Cameroon, Comoros, the Dominican Republic, satellite-derived information products such as the Madagascar, Somalia, Sudan, Zambia, and Zimbabwe. WSF-suite. The WSF suite of products can support the Urbanization Review process in multiple ways. WSF-2019 can be used Figure 3.3. Share of total urban expansion per period and city size, Burundi. The share of built-up expansion is growing quickly in favor of cities below 50K as well as cities between 50K to 150K habitants Share of built-up expansion by city size Below 50K: 34% 33% 41% 8% 50K to 150K: 24% 23% 59% Bujumbura: 41% 36% 1985–2000 2000–2015 2015–2019 Source: Mukim 2022. OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 25 Figure 3.4. Expansion of Garowe’s urban footprint (1985–2015). WSF evolution n 2015 n 2014 n 2013 n 2012 n 2011 n 2010 n 2009 n 2008 n 2007 n 2006 n 2005 n 2004 n 2003 n 2002 n 2001 n 2000 n 1999 n 1998 n 1997 n 1996 n 1995 n 1994 n 1993 n 1992 n 1991 n 1990 n 1989 n 1988 n 1987 n 1986 n 1985 Source: Adapted from: World Bank, 2021a. Urbanization Review Example 1: Patterns Mogadishu’s development historically followed the of urban expansion for Somali cities city grid. But since the 1990s, when civil war broke out, Mogadishu more often expanded in a sprawling In Somalia’s fragile context, urban form and density affect manner, with influxes of internally displaced people (IDP) social integration and equity, whereas social and political coinciding with governance vacuums, because the city contestation have shaped cities’ development and lacked further planned areas where they could settle. spatial forms. The WSF-Evolution product has enabled Sprawl continued in the early 2000s. In this period, the World Bank to observe and analyze the patterns of large satellite settlements dislocated from the main city urban expansion for Somali cities over time. Analyzing emerged or expanded, as thousands of IDPs settled to the these patterns from 1985 to 2015 has helped hypothesize North and North-West of Mogadishu, particularly along how drivers of urbanization in Somalia have contributed the Afgooye Road (figures 3.5 and 3.6). to cities’ spatial forms as they exist (figure 3.4). And consequently, analyze how these spatial forms may deepen or dampen urban economic opportunities, quality of life, and integration. OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 26 Figure 3.5. Urban Expansion in and around Mogadishu (1985–2015). WSF evolution n 2015 n 1999 n 2014 n 1998 n 2013 n 1997 n 2012 n 1996 n 2011 n 1995 n 2010 n 1994 n 2009 n 1993 n 2008 n 1992 n 2007 n 1991 n 2006 n 1990 n 2005 n 1989 n 2004 n 1988 n 1987 n 2003 n 1986 n 2002 n 1985 n 2001 n 2000 0 1.25 2.5 5 Kilometers Source: World Bank, 2021a. Figure 3.6. What defines the boundaries of a city? Qarbis Bosaso Mogadishu Mogadishu Legend Legend GHS GHS Africapolis Africapolis WSF WSF Statistics and anlyses are often conducted at city level - yet it is difficult to define the boundaries of these cities in order to calculate such statistics. Especially as cities grow and the spaces between them are built-up. Various global datasets, including the WSF data, Africapolis (https://africapolis.org/en), and the Global Human Settlement layer (GHS - https://ghsl.jrc.ec.europa. eu/) attempt to delineate city boundaries systematically accross countries and regions. Differences remain between these databases regarding methodology, the spatial scale of settlement boundaries (e.g. see how GHS has coarse boundaries), and the timeliness of updates. Source: Adapted from the World Bank, 2021a. OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 27 Urbanization Review Example 2: factors for Moroni’s expansion have been land conflicts Supporting unified planning efforts with surrounding municipalities and the village of in the Union of the Comoros Iconi. While the municipality’s footprint increases and the validated Preliminary Plan for Grand Moroni has As urban population grows in Comoros, pressure on proposed an extended “Commune Urbaine Grand Moroni” scarce land will intensify. Historically, land was not to facilitate coordination, the municipality’s existing commercialized but often managed as a common good. limits remain contested. Parts of the municipality are This characteristic is changing as the urban areas and claimed by the neighboring municipalities of Bangaani their population grow. Land conflicts will continue and Mavingouni, and by the village of Iconi, making unified to constrain growth scenarios of major cities, as is planning efforts more difficult. The information provided exemplified in Moroni, whose geography allowed for the by the WSF-Evolution has supported the World Bank in expansion of the city outside its municipal boundaries, improving unified planning efforts in Comoros. especially to the South (figure 3.7). The constraining Figure 3.7. Urban expansion of Moroni, Comoros. Buil-Up Area 1985–2015 n 1985 n 1999 n 2008 n 1991 n 2000 n 2009 n 1992 n 2001 n 2010 n 1993 n 2002 n 2011 n 1994 n 2003 n 2012 n 1995 n 2004 n 2013 n 1996 n 2005 n 2014 n 1997 n 2006 n 2015 n 1998 n 2007 Municipal boundaries Moroni municipal boundaries Contested areas, claimed by adjacent municipalities Source: World Bank, 2021b. OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 28 Evaluating urban growth patterns urban growth characteristics of similarly sized cities. This can help identify probable recent or upcoming rural hubs Identifying patterns of urban expansion, and how as well as service area gaps within cities. this expansion evolves over time, is critical to our understanding of urbanization drivers. World Bank teams Both teams make use of the WSF-Evolution layer to such as the Geospatial Operations Support Team (GOST) calculate the expansion per year—possible between use the landscape expansion index1 to characterize urban 1985 and 2015—to highlight periods of rapid growth. growth as: traditional expansion, leapfrog, or infill. The These analyses help identify instances of expansion, Central and West Africa Poverty and Equity Geospatial infill, and leapfrog growth patterns. WSF-Evolution team is conducting similar analyses for large and surpasses existing datasets for accuracy and resolution, secondary cities across Sub-Saharan Africa. The team with the added benefit of cost effective and regionally identified periods of high urban expansion and compared standardized analytics (figures 3.8 and 3.9). Figure 3.8. Burundi: Built-up expansion by city size. Extensions represent the vast majority of built-up expansions in Burundi. In second place, Bujumbara has slightly more infills while the rest of the country sees more leapfrog expansions Infill Extension Leapfrog Bujumbara 16% (capital) 10% 74% Cities between 3% 74% 23% 50K to 150K Cities 24% below 50K 5% 71% Source: Mukim 2021. OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 29 Figure 3.9. Built-up area expansion for large and secondary cities in Mali and Niger. NIAMEY MARADI ZINDER High-density metropolis and capital. Agriculture and trade outpost to Agricultural processing hub and Nigeria. refinery. SEGOU SIKASSO BAMAKO Small hub for fishing and small-island Large agricultural hub. High density metropolis and capital. scale farming and pastoralism. Before 1986 2000 2015 First year of build-up 2016 or later (maps only) Source: The World Bank Central and West Africa Poverty and Equity Geospatial Team, 2022. Assessing impacts of urban land policies when—as is increasingly possible with digitization—they —rights registration, planning, zoning can be linked to land records. Many policies and programs supported by the World A recent evaluation of regulatory reform with measures Bank aim to enable landowners to use their land more to register women’s land rights in Maseru, the capital of effectively. Some instances are access credit to invest in Lesotho, linking WSF-Evolution data to land ownership it, transfer it to others via markets or via land use planning records show that reforms improved women’s access to improve coordination and access to services. WSF to credit and resulted in house construction and denser data provide an opportunity assess the extent to which development (Deininger 2022). these policies are effective. This is particularly relevant OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 30 Disaster risk management To date, CCDRs have been produced for more than 15 countries. The project aims to provide detailed risk Country Climate and Development scores for each country around the globe, which should Reports (CCDR) be easily accessible by the end users via application The effects of climate change are becoming more programming interface (API). The WSF19 layer is used prominent across the world and will greatly influence to identify anthropogenic land cover and thus quantify the pathways for sustainable development. In this exposure. WSF appears to be the best product in context, the World Bank is developing Country Climate accuracy of built-up modeling compared to the other and Development Reports (CCDRs) as diagnostics2 that dataset considered. It has overall better resolution, the address these two issues simultaneously. The World best recognition of built-up areas, and as a global layer, it facilitates comparison between countries (figures 3.10 Bank in conjunction with DLR is developing a replicable, and 3.11). consistent risk screening tool3 for hydromet phenomena to support these reports, which draws from hazard, exposure, and climate data to produce risk scores at the administrative level three (ADM3) level. Figure 3.10. Nepal landslide hazard: Built-up area exposed to high landslide hazard. Landslide hazard Rainfall trigger | Built-up exposed to HIGH landslide hazard | Administrative level 3 (Hectars) < 0.1 n 0.1 – 2.5 n 2.5 – 10 n 10 – 20 n 20 – 50 n > 50 Source: World Bank, 2022. OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 31 Figure 3.11. Nepal landslide hazard: Population exposed to high landslide hazard. Landslide hazard | Rainfall trigger | Built-up exposed to HIGH landslide hazard | Administrative level 3 (Population) < 100 n 100 – 2,000 n 2,001 – 10,000 n 10,001 – 25,000 n 25,001 – 50,000 n 50,001 – 100,000 n > 100,000 Source: World Bank, 2022. Nepal. Photo: © Aleksei Gavrikov | Dreamstime.com OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 32 City Scan and Urban Climate Risk Analyses types of hazards under present and future contexts, taking into account climate change impacts. The City Scan is an assessment tool and framework— developed by the City Resilience Program (CRP) team at Recent City Scans and urban climate risk analyses include the World Bank—that provides a rapid assessment of the analytics using WSF datasets, so far representing more critical development challenges that cities face using than 165 cities around the globe. The scans use WSF- publicly available data. The City Scan aims to serve as a Evolution to identify changes in the built-up area and conversation starter between the World Bank task team how much of it is exposed to flood hazards (figure 3.12). and client city to assess cities’ investment priorities WSF-2019 identifies areas that display considerable and financing needs in six broad areas: population, city recent growth. WSF-Imperviousness is utilized to proxy competitiveness, infrastructure and public services, building density and as a conversation starter for long transport, climate, and municipal finance. Thus far, 67 term policy planning and to respond to disease outbreaks. cities conducted City Scans and the tool is a continually Global coverage of WSF datasets is crucial as City Scans evolving product. The team has also produced several are designed to be deployed anywhere in the world multi-city Urban Climate Risk Analyses—China, Egypt, (figure 3.13). All these analyses have been incorporated Jordan, India, Pakistan, South Africa, and West Bank and into CRP’s scripting to produce these products, enabling Gaza—which detail the level and distribution of various significant savings in time and effort. Figure 3.12. Urban built-up area exposed to river and rain flooding in Dushanbe, Tajikistan. Built-up aerea affected in river and rainwater flood risk zone n Flood risk zone Date built n Pre 1985 n 1986 – 1995 n 1996 – 2005 n 2006 – 2015 Source: World Settlement Footprint Landsat 5/7”. Source: City Resilience Program, 2022. OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 33 Figure 3.13. Built-up area affected by river flood risk in Surat, India. Built-up aerea affected in river flood risk zone River Flood probability n < 1% in any given year n 1 – 10% in any given year n > 10% in any given year Date built n Pre 1985 n 1986 – 1995 n 1996 – 2005 n 2006 – 2015 Source: World Settlement Footprint Landsat 5/7”; SSBN arc second (90 m) Global Hazard Data (World Bank License). Source: City Resilience Program (n.d.). Building exposure models has a high benefit for urban seismic risk assessment incorporating building heights when used along with soil information. Finally, the WSF suite captures spatiotemporal expansion, vertical and The Global Program for Disaster Analytics utilizes horizontal, in hazards. building height data from the WSF-3D layer to develop an adaptive global exposure model to inform disaster risk The primary output is an enhanced exposure model of quantification and climate risk assessments. WSF data building heights, an output that directly enhances CCDR capturing the height of buildings have high relevance profiles.4 The data can also be combined with hazards in urban flood risk assessment when used along with data for more detailed building exposure and vulnerability flood depth hazard layers. The height information also analytics. OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 34 Public and environmental health 2023, p. 16). It helped target emergency interventions that avoid the rapid spread of a virus, such as providing Cities, crowding, and the coronavirus: additional infrastructure services —for example additional Predicting contagion risk hotspots water distribution points and portable hand washing sites At the onset of the coronavirus pandemic, one of the in highly affected neighborhoods, just as investing in long primary containment strategies focused on extreme term activities in high risk zones. social distancing. Wealthier people were able to isolate High risk hotspots are identified by finding locations with themselves and still access amenities and employment extremely high density living conditions—a combination whereas residents from poorer communities often of population density and livable floor space that does not have little choice but to congregate in public places allow for physical distancing—and locations where people to facilitate obligatory daily activities. The World Bank have a necessity to cluster such as public toilets, water team developed a tool to identify places with the highest pumps, and markets. WSF products were combined with exposure and contagion risk. The tool was used support information on population distribution and data indicating city leaders with resource prioritization decisions during key service points to identify areas of higher risk, derived pandemic events (for more information see World Bank from OpenStreetMap (figure 3.14). Figure 3.14. Potential hotspots in Kinshasa (72% of the population) showing population density, service location and mobility. Source: This map combines population data from WorldPop and height data from DLR. The population was adjusted to reflect the people living in Kinshasa, based on a recently conducted city survey. The hotspots are in red and purple. OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 35 Coal fired power station on the Vaal River, Free State province, South Africa. Photo: Jay Roode Urban air pollution exposure within the urban environment. The index allows for the comparison of the situation between well-defined areas The correlation between air pollution and the with similar features, like urban areas. The data used for exacerbation of certain health outcomes entails this analysis are CAMS regional–European reanalysis for identifying urban air pollution hot spots at the city PM2.5 concentrations. Demographical data of settlement level. This is done by combining air pollution data with extent and population density were derived from the the number of people exposed to defined pollution World Settlement Footprint suite of products. This first concentrations. Exposure in this context is represented approach has the potential to become a scalable tool to by the health burden index (HBI); a parameter for the support decision-making process for public health in assessment of health risk that considers the total urban environments. number of people impacted by a certain pollution level OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 36 Upcoming applications funded projects, not least those that focus on disaster risk management, public health, and climate change. Several emerging use cases employ the WSF suite, Teams at the World Bank are exploring multiple ways to which exhibits great potential for standardization in the detect and monitor urban vegetation, and the WSF suite near future. These include urban vegetation and energy may provide additional perspective. The development of demand and air quality. the WSF Imperviousness layer relies on a strong inverse relationship between impervious surface and vegetation: Urban heat stress the higher the vegetation index, the lower the percent of Densely built-up areas with little or no vegetation can impervious surface within a given built-up pixel. Hence, lead to urban heat islands that can have significant health the WSF Imperviousness can be seen as a suitable basis effects on the local population such as heat stroke, heat for urban vegetation use cases going forward, such as exhaustion, and related respiratory difficulties. Extreme quantifying green spaces in urban areas or correlating heat particularly affects the health and livelihoods greenness and air pollution or livability. of poorer communities. Modeling predictions for Johannesburg, South Africa, suggest an increase in the Energy demand and air quality heat-related excess mortality rate by 2050, resulting in Existing studies in the fields of energy demand and air several hundred additional deaths per year. The WSF- quality research tend to focus on hyper-local analyses, suite can be used to capture patterns in building density as data are cost intensive and not readily available. and form and the presence of vegetation to help identify Although these studies tend to employ high resolution urban heat islands and to help target interventions. data, methods and ideas can be transferred, and case Outputs include maps indicating present day urban heat studies exist for both fields in which relatively lower island intensity, temperature variations per city, and resolution data on urban density or building height have gridded predictions of the number of hot nights per year been incorporated. The WSF products can serve as a to be expected according to climate change scenarios. valuable basis for approximations of either the calculation of energy demand in urban areas or of air pollution. Global income dataset Understanding the links between income distribution and Near real-time carbon dioxide urban characteristics can help understand what drives emissions from cities the development of poorer settlements. This insight can The Laboratoire des Sciences du Climat et inform planning interventions to avoid repeating patterns l’Environnement (LSCE), Tsinghua University, and Kayrros of poorly planned development. The City Resilience are building an evolution of the Carbon Monitor Cities Program team is working to generate the first global high near real-time fossil carbon dioxide emissions analysis resolution spatial dataset of income. The WSF-Population system for urban areas, covering a selected list of cities and WSF-Imperviousness layers are critical to model the in Egypt, South Africa and Turkey. The data will be made distribution of income levels across built-up areas globally, publicly available. The gridded emission dataset will cover at the highest possible resolution. The analysis outputs six sectors based on near real-time local activity data and also extend to changes in income over time, for example emissions models, with daily emissions maps at a spatial annual updates, and describe the relationship between resolution of at least 500 meters. The WSF building area income distribution to other urban spatial factors—urban and height dataset and population density are critical to growth patterns, vertical and horizontal growth, population downscale residential energy use and allocate carbon density, and accessibility to public transport. dioxide emissions from air conditioning to consumers. Urban vegetation The identification of urban vegetation—from baseline establishment to change over time and the potential for future regreening—is critical to many World Bank OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 37 Notes 3. Code and tutorials for the risk screening tool is available on 1. Code for the Landscape Expansion Index is available on Github: https://github.com/GFDRR/CCDR-tools/tree/main/ Github: https://github.com/worldbank/GOST_Urban/wiki/ Top-down Landscape-Expansion-Index 4. Disaster Risk Analytics website: https://www.gfdrr.org/en/ 2. CCDR website: https://www.worldbank.org/en/publication/ disaster-risk-analytics country-climate-development-reports References Lall, S.V., Lebrand, M.S.M.P., Sturm, H., Venables, D.M., and Bhardwaj, G., Esch,T., Lall, S.V., Marconcini, M., Soppelsa, M.E., Anthony, J. 2021. Pancakes to Pyramids: City Form to Promote and Wahba, S. (2020). Cities, Crowding, and the Coronavirus: Sustainable Growth. World Bank, Washington D.C. Available at: Predicting Contagion Risk Hotspots. World Bank, Washington, https://documents.worldbank.org/en/publication/documents- D.C. Available at: https://openknowledge.worldbank.org/ reports/documentdetail/554671622446381555/cityform-to- handle/10986/33648 promote-sustainable-growth City Resilience Program. 2022. City Scan —Dushanbe, Tajikistan. World Bank. (2021a). Somalia Urbanization Review: Fostering GFDRR / World Bank, Washington, D.C. Cities as Anchors of Development. World Bank, Washington DC. Available at: http://hdl.handle.net/10986/35059 City Resilience Program (n.d.) Urban Climate Risk Analysis - India. GFDRR / World Bank. Washington, D.C. World Bank. (2021b). Comoros Urbanization Review: Reimagining Urbanization in Comoros. World Bank, Washington, D.C. Deininger, K.W., and Ali, D.A. 2022. How Urban Land Titling and Available at: http://documents.worldbank.org/curated/ Registry Reform Affect Land and Credit Markets: Evidence from en/471791612255765765/Comoros-Urbanization-Review- Lesotho. Policy Research Working Paper 10043. The World Bank, Reimagining-Urbanization-in-Comoros Washington D.C. Available at: http://hdl.handle.net/10986/37458 World Bank. (2022). Nepal - Country Climate and Development Gilardi, L., Metz-Marconcini, A., Marconcini, M., and Erbertseder, Report. World Bank. Washington, D.C. Available at: https:// T. 2021. Urban air pollution exposure: an assessment exploiting openknowledge.worldbank.org/bitstream/handle/10986/38012/ world settlement footprint and land use data. Proceedings of FullReport.pdf SPIE, Remote Sensing Technologies and Applications in Urban Environments VI (Vol. 11864, pp. 7-16). Available at: https://doi. World Bank. (2023). Digital-first approach to emergency cash org/10.1117/12.2600414 transfers: STEP-KIN in the Democratic Republic of Congo. World Bank. Washington, D.C. Available at: https://documents1. Mukim, M. (2021). Burundi Urbanization Review: Investing worldbank.org/curated/en/099935104272316767/pdf/ in Resilient and Inclusive Growth. World Bank Group, IDU05debcd9500bf004a580a48b0c6c201068bdc.pdf Washington, D.C. Available at: http://documents. worldbank.org/curated/en/099300012012154930/ P1753940ba0bfc0dd0b3ef08f535a797337 OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 38 A township in the East Rand area of Gauteng, near Johannesburg. Photo: RapidEye. OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 39 4 THE WAY FORWARD T his document provides early examples of The utility and need of global urbanization datasets such development use cases for the WSF suite as WSF are further demonstrated by the willingness of of datasets. These datasets provide timely, World Bank teams to invest in future improvements. standardized, replicable, and up-to-date information Collaborations with the Central and West Africa Poverty regarding global urbanization patterns. The WSF and Equity team are working on harmonizing WSF- also supports key insights at a global level into where Evolution and WSF-2019 to create a seamless time series settlements are expanding more rapidly. Insights that that enables the comparison of urbanization trends can be extracted from these datasets are enhancing the from 1985 through 2019. In addition, numerous teams cost-effectiveness, accuracy, and timeliness of results for have expressed strong interest in exploiting the unique teams and programs across the World Bank. capabilities provided by the novel WSF-3D to describe and analyze built-up density worldwide more comprehensively Close collaboration with World Bank teams identified and precisely than ever before. This potential ranges from several improvements to global urbanization datasets the derivation of innovative indicators on morphologic such as the WSF suite that would unlock even more settlement characteristics over improved modeling of applications for World Bank operations. The foremost population distribution and risk exposure to improved request from teams was the development of an urban climate mitigation and adaptation strategies, or operational service of regularly updated gridded data on enhanced modeling of economic development. settlement extent as with WSF-2019. To conclude, it is clear that global datasets from Earth The second most common demand was related to observation data have great potential to standardize data accessibility. Previously, WSF products were World Bank operations and provide regular updates to downloadable tile-wise—one by one degree latitude and monitor development. This is proven by the integration longitude—or as a global data package via DLR’s EOC of EO datasets such as the WSF-suite into so many Geoservice.1 Third, users requested that WSF products existing operational use cases despite the recency of be made available in different spatial resolutions so that their release. EO products are constantly improving, and preprocessing efforts to harmonize multiple datasets many new datasets and applications are expected in the could be minimized. With the newest developments, coming years. To help navigate these changes, the report most datasets —WSF3D still pending—are now indexed presents an overview of teams that can support the in a catalog Spatio Temporal Asset Catalog (STAC).2 STAC utilization of EO products and services within World Bank provides capabilities to search, download, mosaic, and operations (Appendix B). resample data, using simple Python libraries, allowing users to seamlessly integrate the available datasets Notes 1. https://geoservice.dlr.de/web into their analytical workflows. This eliminates manual 2. STAC - https://geoservice.dlr.de/eoc/ogc/stac/ download of large tiles and allows accessing data on arbitrary areas of interest. Simple examples on how to access the data are expected to be available within the service. OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 40 APPENDIX A: HOW TO ACCESS THE WSF DATASETS The following table provides an overview of how to building (Africa), vector outlines of urban centers with an access the WSF datasets. Note that all World Bank estimated population above 10,000 inhabitants (Africa), staff and consultants can also access WSF datasets and quarterly updates of the WSF-2019 (selected cities). for Africa through the World Bank’s Development Data STAC-based data access is available for all datasets Hub (DDH). These datasets were produced through the that are available via the DLR EOC Geoservice (https:// Monitoring Urbanization from Space project and include geoservice.dlr.de/eoc/ogc/stac/). New datasets will also the WSF datasets described in this paper as well as be available upon release. pilot datasets regarding the estimated population per Spatial Spatial Availability Dataset Resolution Coverage Publication (Public) Link 10mx10m Global Marconcini et Attribution 4.0 DLR EOC Geoservice: al., 2020 International https://geoservice.dlr.de/web/maps/eoc:wsf (CC BY 4.0) Africa region available on the World Bank Data Development Hub WSF2015v2 Google Earth Engine (v1 is publicly available, v2 on request): https://developers.google.com/earth-engine/datasets/catalog/ DLR_WSF_WSF2015_v1 30mx30m Global Marconcini et Attribution 4.0 DLR EOC Geoservice: al., 2021 International https://geoservice.dlr.de/web/maps/eoc:wsfevolution WSF-Evolution (CC BY 4.0) Africa region available on the World Bank Data Development Hub GEE (on request) 10mx10m Global Marconcini et Attribution 4.0 DLR EOC Geoservice: al., 2021 International https://geoservice.dlr.de/web/maps/eoc:wsf2019 WSF2019 (CC BY 4.0) Africa region available on the World Bank Data Development Hub GEE (planned) 90mx90m Global Esch et al., Attribution 4.0 DLR EOC Geoservice: WSF3D 2022. International https://geoservice.dlr.de/web/maps/eoc:wsf3d (CC BY 4.0) Africa region available on the World Bank Data Development Hub 10mx10m Global Pending Planned Global dataset pending release WSF- Publication release in Q3 Africa region available on the World Bank Data Development Hub Imperviousness 2023 10mx10m Global Palacios- Planned Global dataset pending release WSF2019- Lopez et al., release in Q3 Africa region available on the World Bank Data Development Hub Population 2021 2023 References Marconcini, M., Metz-Marconcini, A., Esch, T., and Gorelick, Esch, T., Brzoska, E., Dech, S., Leutner, B., Palacios-Lopez, N. 2021. Understanding Current Trends in Global Urbanisation D., Metz-Marconcini, A., Marconcini, M., Roth, A. andZeidler, - The World Settlement Footprint Suite. GI Forum 2021, 9(1), J. 2022. World Settlement Footprint 3D-A first three- 33–38. https://doi.org/10.1553/giscience2021_01_s33 dimensional survey of the global building stock. Remote Sensing of Environment, 270, 112877. https://doi.org/10.1016/j. Palacios-Lopez, D., Bachofer, F., Esch, T., Marconcini, M., rse.2021.112877 MacManus, K., Sorichetta, A., Zeidler, J., Dech, S., Tatem, A.J. and Reinartz, P. 2021. High-resolution gridded population Marconcini, M., Metz-Marconcini, A., Üreyen, S., Palacios- datasets: Exploring the Capabilities of the world Lopez, D., Hanke, W., Bachofer, F., Zeidler, J., Esch, T., settlement footprint 2019 imperviousness layer for the Gorelick, N., Kakarla, A., Paganini, M. and Strano, E. 2020. African continent. Remote Sensing, 13(6), 1142. https://doi. Outlining where humans live, the World Settlement org/10.3390/rs13061142 Footprint 2015. Scientific Data, 7(1), 1-14. https://doi. org/10.1038/s41597-020-00580-5 OUR DIGITAL EARTH INSIGHTS AND OPPORTUNITIES FROM THE WORLD SETTLEMENT FOOTPRINT 41 APPENDIX B: WHERE TO GET HELP REGARDING WSF DATASETS DEP: Digital Earth Team DLR: German Aerospace Centre The Digital Earth team provides recommendations DLR is the point of contact and support for all matters on how to utilize Earth observation products in World concerning WSF research and production, product Bank operations. It can facilitate the coordination of development and enhancement, quality assurance, WSF-related activities at the World Bank as well as the data access, as well as application development and coordination of improvements. The team can provide downstream services. In this capacity and role, DLR advice for teams willing to access the datasets, perform provides technical assistance and thematic expertise basic analytics, and recommendations on how the data to World Bank staff and clients on how to effectively could be used to support specific World Bank programs analyze and efficiently utilize the WSF data for and activities. The team can also help connect the informed decision making and their dedicated field(s) of interested reader to the use cases described in this application. As a third-party funded research initiative, report. the WSF product suite and analytics functionalities can be further developed, expanded, and adapted on- Contact: DEP_Team@worldbankgroup.org demand to specific user needs and new application scenarios, respectively, by direct funding of the required activities for the WSF team. GOST: Global Operations Support Team The Global Operations Support team (GOST) in the Contact: wsf@dlr.de World Bank Group supports teams and colleagues with the institution on extracting novel insights from data, focused primarily on geospatial, big data, and surveys. GOST support teams with bespoke analysis, TOR design and review, procurement assistance, and any other issue related to extracting insights from novel data. Contact: gost@worldbank.org