How satellites see our cities – and can help them become more sustainable

Cities around the world are undergoing unprecedented transformations. Over the last four decades, the global urban population has more than doubled, reaching 3.5 billion in 2015. It is projected to reach 5 billion by 2050. At the same time, cities need to manage their land take and impact on the environment. How are they rising to the challenge?

Urban landscapes are being reshaped

New high-resolution satellite data shows how the urban landscape is shifting in real-time. The recent built-up estimates, released by European Commission’s Joint Research Centre on a global scale, show that in the last two decades, some cities have experienced a rapid surge in residential construction, in many cases outpacing population growth. This can not only impact energy demand and housing markets, but also contribute to increased emissions by depleting carbon sinks and undermining biodiversity. In Gwangsan (Korea), residential building space increased by more than 40% over the past 20 years, whereas population increased by only 9%, the largest differential observed across OECD large metropolitan areas.

Where can you live large?

New OECD land-use estimates also reveal a huge difference in the residential land (both buildings and private gardens) available per inhabitant. While in Korea, Colombia and Türkiye residential land per inhabitant is only 40-60 m2, in the United States, Finland and New Zealand, available residential land is 450 m2. Bogotá (Colombia) is the most compact OECD large metropolitan area (i.e. with more than 1.5 million inhabitants) with only 25 m2 per capita, while residents of while New Haven (United States) have a whopping 561 m2/capita.

Land-use estimates in the metropolitan area of Vienna (Austria), 2018

Source: OECD estimates based on AI land use model

Of course, the land take is only part of the story. We are also interested in the total dwelling space available, which is built upwards as well as outwards. Thanks to recent estimates of building heights released by the EC-JRC on a global scale, we can now explore disparities in dwelling space across OECD metropolitan areas. This shows that residential building volume ranges from 220 m3 per inhabitant in Finland to almost 1400 m3 in New Zealand. The metropolitan area of Auckland (New Zealand) records the highest residential space per capita across OECD large metropolitan areas, 10 times higher than for the metropolitan area of Bogota (Colombia), which has the least residential space per person. Interestingly, Finland shows both a relatively high level of residential land per capita but a low level of dwelling space per capita, indicating that buildings in the country are more spread out, characterised by a lot of open space such as gardens, coupled with a trend of lower building height.

How green is your city?

We can also see which cities are greenest. Green spaces are increasingly valuable not only to the well-being of urban inhabitants, but also in reducing the temperature spikes that have left many sweltering in recent summers. We find cities in the United States record on average 380 m2 of green space per capita, almost 7 times more green space per capita than urban centres located in Mexico (55 m2 per capita).

Accounting for policy makers

This combination of AI and satellite imagery used for this analysis is revolutionising our ability to plan and track urban development – and its impact on people and the environment. It will also provide a helpful tool for citizens to plan their next move, and hold policy makers to account in the context of new initiatives, such as the EU’s drive to achieve ‘no net land take by 2050.’

Read more about our studies: Monitoring land use in cities using satellite imagery and deep learning, OECD Regions and Cities at a Glance 2022. All the indicators presented in this article are now available on the OECD metropolitan database.

Data Scientist at OECD | + posts

Alexandre Banquet works as a Geospatial Data Scientist for the OECD Centre for entrepreneurship, SMEs, regions and cities. He actively contributes to the development of subnational indicators in all 38 OECD member countries on various topics, including climate change, land use, environment, energy, demographic change, and housing, by leveraging both conventional and alternative sources of data, such as Earth Observation. Alexandre also contributes to the activities of the OECD Laboratory for Geospatial Analysis.

Before joining the OECD, Alexandre worked as an Energy Data Scientist for the US R&D centre of Electricité de France (EDF) in the Silicon Valley. Alexandre holds a Master's degree in Science and Executive Engineering from Mines Paris - PSL, as well as a Master of Science in Mathematics from ENS Paris-Saclay.