Researchers at the Technical University of Munich (TUM) in Germany have built a new online atlas that plots 2.75 billion buildings worldwide in 3D-putting familiar navigation tools to one side.
The technical leap rests on an enormous bank of satellite imagery combined with machine-learning methods. In essence, the system uses height data from buildings that have already been measured to estimate the dimensions of buildings that have not.
GlobalBuildingAtlas: what it is and how it compares
The map, called the GlobalBuildingAtlas, is already available online. Against earlier resources of this kind-such as the Microsoft Building Footprint database-it adds information for more than a billion additional buildings, and it also provides relatively high-resolution 3D models for almost all of those structures.
Now that the project is finished and publicly accessible, the researchers behind it hope it will help guide choices connected to climate change, city infrastructure, disaster preparedness, and other domains where urbanization is a major factor.
Why 3D building data matters for urbanization and poverty
"3D building information provides a much more accurate picture of urbanization and poverty than traditional 2D maps," says Xiaoxiang Zhu, data scientist at TUM.
"With 3D models, we see not only the footprint but also the volume of each building, enabling far more precise insights into living conditions."
Resolution, coverage, and what the models can show
The atlas is built on a resolution of 3-by-3 metre blocks. That is not detailed enough to reveal finer architectural features, but it is sufficient to represent each building’s scale and overall form. The team reports that this is 30 times finer than what existing building footprint databases have achieved, which makes the result notable even with the limitations.
Having volumetric information-rather than footprints alone-can materially change how population density is calculated and how the layout of urban centres is understood.
Planning and policy uses: UN goals and new indicators
As a growing share of the global population relocates to cities, the United Nations has made the creation of cities that are "inclusive, safe, resilient and sustainable" one of its central Sustainable Development Goals within its 2030 Agenda. The researchers see the new atlas as one tool that could support progress towards that objective.
They also point to an analytical shift enabled by the dataset: assessing built-up areas via 3D volume instead of 2D coverage. This could provide a more faithful estimate of how many people live in a given place-and, as a consequence, how many essential public services (including hospitals and schools) will be required.
"We introduce a new global indicator: building volume per capita, the total building mass relative to population, a measure of housing and infrastructure that reveals social and economic disparities," says Zhu.
"This indicator supports sustainable urban development and helps cities become more inclusive and resilient."
Limitations and improvements still needed
Because machine-learning AI is involved, the researchers caution that the 3D measurements will not be perfectly accurate everywhere. They note that parts of Africa need more training data and additional validation, and that tall buildings are often underestimated in height overall.
Even so, the team describes this as the most accurate and wide-ranging 3D building map produced to date. They also indicate that the dataset’s quality is intended to improve over time, which would increase the atlas’s usefulness.
"Buildings anchor human life and define the form and function of urban environments," write the researchers in their published paper.
"3D insights are essential for urban planning, infrastructure management and policy-making – especially in resource-limited contexts where the strategic allocation of funding and intervention is critical."
The research has been published in Earth System Science Data.
Comments
No comments yet. Be the first to comment!
Leave a Comment