What humans have wrought is heavy. In a new study, researchers calculated the weight of more than 600 million buildings in cities across the world and came up with a total of 835 billion metric tons, or gigatonnes (Gt) [1]. That’s almost as much as the weight of all the world’s plants.
All of that weight—in steel, wood, glass, cement, stone, and other materials—means carbon emissions. “If rapidly urbanizing regions follow historical development pathways, this stock could grow by another 419 Gt by 2050,” says study team member Jinchao Song, a research fellow at the University of Michigan’s Center for Sustainable Systems.
Conventional life-cycle analyses and green building certification programs like LEED are built on the principle of material efficiency and less is more. But reducing the environmental impacts of buildings isn’t just about making them lighter, according to a second paper [2].
In that work, Fernanda Cruz Rios, a built environment scientist at Drexel University in Pennsylvania, argues that what matters is how resilient a building is to fires, earthquakes, storms, floods, and so on —especially in the current age of increasing climate disasters.
Consider two otherwise identical buildings: one built directly on the ground and one on a raised concrete platform. The concrete platform increases the carbon emissions involved in construction. But when a major flood comes, the building on the concrete platform stays high and dry, while the one on the ground is inundated requiring costly and time-consuming repairs.
When all the impacts of the flooding are added up—new materials, mold remediation, the need for the building’s occupants to decamp to an alternate location for months—the building on the concrete platform may well come out ahead.
“Once you account for what happens when a building actually gets hit by a disaster, designing it to stay functional can prove to be the sustainable choice, not a trade-off against it,” says Cruz Rios
Existing life-cycle analyses sometimes account for the impact of the materials necessary for repairs. But when Cruz Rios analyzed 40 previously published life-cycle analyses focusing on flood, wind, heat, and seismic hazards she found that none of them captured all dimensions of building resilience.
“Almost no studies look at what happens during a disaster response (the emergency actions like evacuations, temporary fixes, and workarounds a building or community relies on while the event is unfolding),” says Cruz Rios. “Those actions clearly carry environmental costs, but they’re essentially invisible in the data.”
In the paper, Cruz Rios sketches out a new approach to life-cycle analysis that foregrounds resilience, making the benchmark of sustainability a building’s material use in relation to the function it delivers over its lifetime. Cruz Rios reports that her lab is now “working to gather the kind of data that could make this analysis fully usable in practice.”
At the same time, no building is an island. According to the first study, urban form—meaning the height, density, and uniformity of buildings across a city—makes a big difference in material use. Cities with a dense, uniform arrangement of low-rise buildings maximize per capita floor space per pound of building material.
“I was also surprised by how much urban form could influence future material demand,” Song says. Going forward, if rapidly growing cities in the Global South adopt the dense, low-rise arrangement they could require about 30% less material than if they were to adopt the least efficient urban form (a sparse arrangement of uniformly high-rise buildings).
The finding “suggests that early urban planning decisions can substantially shape a city’s long-term material needs,” says Song—underlining the need to consider building sustainability across multiple levels of space and time.
Sources:
[1] Song J. et al. “Weighing 600 million buildings reveals global urban material inequality and efficiency paths.” Nature Cities 2026.
[2] Cruz Rios F. “Resilience in building life cycle assessment: a critical review and framework for time‐integrated functionality.” Journal of Industrial Ecology 2026.
Image: John Fleck/FEMA.
