Models of all the world’s buildings now available in an interactive 3D map

The open-data GlobalBuildingAtlas is the first high-resolution 3D map of 2.75 billion buildings worldwide. Derived from satellite imagery by researchers at the Technical University of Munich (TUM), it opens up new possibilities for understanding how cities grow, how infrastructure is shaped, and how risks unfold. The dataset provides a powerful new basis for climate research and for tracking progress towards the UN Sustainable Development Goals (SDGs), enabling more precise models for urban development, infrastructure planning and disaster management. Ultimately, GlobalBuildingAtlas represents a tangible step towards cities that are more resilient, more inclusive and better prepared for the challenges ahead.

A research team led by Prof Xiaoxiang Zhu, director of the Munich Data Science Institute at the Technical University of Munich (TUM), has not only counted the world’s buildings, but also developed a dataset visualizing them all in three dimensions. Supported by an ERC Starting Grant, the project resulted in GlobalBuildingAtlas, a global 3D representation of 2.75 billion buildings derived from satellite imagery from 2019. This significantly exceeds the previous global dataset, which contained around 1.7 billion buildings.

Better resolution, scalability and insight

At 3×3m-resolution, the models in GlobalBuildingAtlas are around 30 times more detailed than comparable datasets. Approximately 97% (2.68 billion) of the buildings are provided as Level of Detail 1 (LoD1) models. While less detailed than higher LoD levels, these simplified 3D representations capturing the basic shape and height of each building can be integrated at scale into computational models as the basis for urban infrastructure analysis and planning purposes. Moreover, GlobalBuildingAtlas includes buildings in regions such as Africa, South America and rural areas, which are often missing in other global datasets.

New perspectives for inclusion, sustainability and climate research

Three-dimensional models visualize not only the footprint but also the volume of each building, enabling far more precise insights into urbanization, living conditions and poverty than traditional 2D maps, according to Prof Zhu. The TUM team has introduced a new global indicator: building volume per capita, i.e. the total building mass relative to population. “As a measure of housing and infrastructure that reveals social and economic disparities, this indicator supports sustainable urban development and helps cities become more inclusive and resilient,” he says.

For example, the 3D building data from GlobalBuildingAtlas enables cities to plan additional housing or public facilities such as schools and health centres in densely populated, disadvantaged neighbourhoods.

At the same time, the data plays a crucial role in climate adaptation, improving models on topics such as energy demand and carbon emissions, and supporting the planning of green infrastructure. Meanwhile, by allowing risks from natural events such as floods or earthquakes to be assessed more quickly, GlobalBuildingAtlas enhances disaster prevention.

Real-world application

The German Aerospace Center (DLR) is just one of the organizations to have already expressed an interest in GlobalBuildingAtlas. DLR is examining its potential use as part of the ‘International Charter: Space and Major Disasters’.

To access the interactive map, go to GlobalBuildingAtlas LoD1. All data and code are freely available via GitHub and mediaTUM, TUM’s media and publication server.

For more information, read the full paper: GlobalBuildingAtlas: an open global and complete dataset of building polygons, heights and LoD1 3D models