David Bridges
At a glance
- GlobalBuildingAtlas: A high-resolution 3D map of all buildings worldwide was created by researchers at TUM.
- Data Volume: The dataset includes 2.75 billion building models, surpassing the previous dataset of 1.7 billion.
- Interactive Features: Users can access an interactive map and search specific addresses for building details.
- Socioeconomic Insights: The project aims to provide insights into urbanization and poverty levels globally.
The world has a lot of buildings. Now you can see them all with a single glance.
A research team at the Technical University of Munich (TUM) in Germany published the GlobalBuildingAtlas, a high-resolution 3D map of all buildings worldwide. The map consists of 2.75 billion building models, which the team gathered from satellite images taken since 2019.
This is a huge leap from the previous global dataset, which contained about 1.7 billion buildings, and offers much better resolution, which is about 30 times finer than comparable databases, according to the researchers in a statement. A detailed account of how the map was created was also published in the journal Earth System Science Data on December 1.
You can access the interactive map here. There’s also a tab for entering a specific address, which also tells you the building placement and elevation. Here’s the result for Gizmodo’s office in New York:
Alternatively, you can also download the data and code for GlobalBuildingAtlas from GitHub.
Why buildings?
From satellite data, the team generated 3D models of metrics like a building’s height, volume, and position compared to other buildings in its vicinity. All the data was also treated with a filtering strategy to account for differences in the quality of satellite data for certain regions.
Other than being cool to look at, the researchers created the project to serve a socioeconomic and environmental purpose. For instance, such a comprehensive, bird’s-eye-view map offers detailed information about the “footprint” of urbanization and poverty around the world, Xiaoxiang Zhu, study lead author and a data scientist at TUM, said in the statement.
For this purpose, the researchers also took special care in finding and incorporating data from often-omitted regions in global maps, such as Africa, South America, and rural areas. The dataset also includes a metric known as building volume per capita, or the total building mass relative to population. Assessing this metric makes it possible to investigate social and economic disparities in certain areas, Zhu explained.
Buildings on Earth
But the team also hopes that mapping urban distribution this way can inform climate studies. For instance, having a good grasp of where and how buildings and populations are located could help improve models on energy demand or carbon emissions, according to the paper.
Already, agencies such as the German Aerospace Center have expressed interest in using the map to evaluate risks from natural and human-made disasters, the researchers added in the statement.
“Buildings anchor human life and define the form and function of urban environments,” the paper stated. “Such 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.”
Now, for the non-scientist and non-politician who just wants to tinker around with the map, please do. The map offers a good range of customizable options for how you want things to look, so definitely go in and explore.
There are certain areas that, zoomed out, look dark and devoid of data, but if you scroll to zoom, you’ll actually see the map has 3D models for maybe three or four buildings, even in areas that look empty and isolated. At least, it worked when I typed in the address of a relative of mine in South Korea who, for better or worse, actually lives in the middle of nowhere with endangered water deer—so the map’s coverage is pretty impressive, I’d say.
Daniel Mercer
