Published on Sustainable Cities

Insights from Space: Monitoring City Expansion with AI-Powered Satellite Technology

This page in:
WSF Evolution showing urban expansion 1985–2015 (Source: GFDRR Partnership Days 2022: DEP & ESA presentation) WSF Evolution showing urban expansion 1985–2015 (Source: GFDRR Partnership Days 2022: DEP & ESA presentation)

Satellite imagery has been an important basis for development operations for some time. Recent advances in big data computing and machine learning are allowing us to take advantage of growing catalogs of open-source imagery and to produce maps of the ever-changing urban environment at unprecedented scale, timeliness, and accuracy.

 

One example is the World Settlement Footprint (WSF), a suite of map layers developed by the German Aerospace Center (DLR) in collaboration with the European Space Agency (ESA), the Google Earth Engine team, and the World Bank (available on the World Bank Development Data Hub, the DLR EOC Geoservice, and Google Earth Engine). By combining multispectral and radar-based satellite imagery, these grid-based data layers help World Bank teams inform decisions for a myriad of operations, from understanding urbanization patterns over time to estimating how settlements are expanding into hazard-prone areas. These issues have significant consequences for urban planning, public health, and environmental management.

 

Advantages of satellites-derived products

 

Data derived from Earth Observation (EO) satellite images provides information that is timely, standardized, consistent, verifiable, and scalable. Timely, because new imagery is being continuously collected and processed. Standardized, because image acquisition is based on calibrated measurements. Consistent, because the data allow us to conduct comparable and repeatable analyses across countries. Verifiable, because, unlike most field-based surveys, an independent entity can access the original satellite data to cross-examine and reverify the information extracted. Finally, these data are scalable, as the methods used to create these insights can be extended across regions and examined over time.

 

Answering key development questions

 

The WSF datasets make abundant use of satellite imagery, and the report Our Digital Earth - Insights and Opportunities from the World Settlement Footprint summarizes several ways these satellite-derived data can be used for World Bank projects. By overlaying settlement extent with global hazard datasets, we can capture how much urban growth is happening in areas with increased exposure to natural hazards such as floods and landslides. For example, in Africa, almost 10% of the settlement extent is located in areas at risk of flooding, where an estimated 72.4 million people, or 5.9% of the continent’s population live (GFDRR 2023). This knowledge can help prioritize development needs across countries.

Graph showing settlements are slowing down in Latin America and Caribbean and increasing in Sub-Saharan Africa.

Figure 1: Global datasets like the World Settlement Footprint can show us how quickly settlements are growing in different parts of the world. These yearly growth rates show us that settlement expansion in Latin America & Caribbean is slowing down, whereas it’s increasing in Sub-Saharan Africa. (adapted from Rentschler et al 2022).

 

Products such as the WSF also augments World Bank’s analytical and diagnostic products such as Climate Change Development Reports (CCDRs) and City Scans, by helping us visualize settlement expansion over time. It shows us where the most recent growth is happening, allowing us to identify and track informal versus formally approved expansion, and identify geographical barriers that limit the smart expansion of settlements.

 

Looking beyond human expansion

 

Due to the spatial nature of these datasets, it’s easy to combine them with other georeferenced information to get a more detailed snapshot of settlement characteristics. By capturing patterns in building density and the presence of vegetation, the WSF datasets can assess the intensity of urban heat islands. Similarly, urban density and building height can be used to approximate energy demand, air quality, and many more aspects of cities.

 

The integration of EO datasets such as the WSF into so many existing operational use cases is a strong endorsement of this technology, and its potential to support decisions. EO products are constantly improving, with many new datasets and applications expected in the coming years. Watch this space.

 

Excited? Reach out to the Global Facility for Disaster Reduction and Recovery’s (GFDRR) Digital Earth Partnership team (DEP_team@worldbankgroup.org) for more information on how to utilize Earth Observation products in World Bank operations. The suite of WSF geoinformation products is developed by DLR’s team Smart Cities and Spatial Development and all datasets can be discovered, visualized and downloaded at the Geoservice of DLR’s Earth Observation Center (EOC).


Edward Anderson

Senior Technology and Resilience Specialist, Global Facility for Disaster Reduction and Recovery (GFDRR)

Pierre Chrzanowski

Disaster Risk Management Specialist, GFDRR, World Bank

Mattia Marconcini

Development Project Manager and Research Data Scientist

Thomas Esch

Leader of the research team “Urban Areas and Land Management” and project manager at the German Remote Sensing Data Center

Nuala Margaret Cowan

Disaster Risk Management Consultant

Join the Conversation

The content of this field is kept private and will not be shown publicly
Remaining characters: 1000