From Urban Growth to Material Accumulation: Reconstructing Building Stock Dynamics in Gothenburg, from 1960 to 2015
This study reconstructs the long-term dynamics of Gothenburg's residential building stock and material flows from 1960 to 2015 by integrating historical GIS, street-network characteristics, and machine learning to quantify a net accumulation of approximately 9.75 Mt of materials, demonstrating how such a framework can analyze the spatial formation and turnover of urban material stocks.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Cities are not just collections of buildings; they are vast, living reservoirs of physical matter. Every wall, floor, and beam represents a quantity of resources that has been extracted from the earth, processed, and assembled into a structure that can last for decades or even centuries. This concept, often called urban metabolism, views a city as an organism that constantly takes in resources and releases waste, but with a crucial difference: unlike food that is eaten and digested quickly, the materials in our buildings accumulate over time. They sit there, locked in place, forming a massive stock of concrete, steel, and wood that defines the city's physical footprint. Understanding how this stock grows, changes, and eventually breaks down is essential for planning a sustainable future, yet it has been incredibly difficult to trace. We know how many people live in a city, but we rarely know exactly how many tons of material were added to the city's buildings in a specific year, or where those materials ended up.
To solve this puzzle, researchers Jonathan Cohen at Chalmers University of Technology undertook a massive reconstruction of Gothenburg, Sweden, looking at every single building in the city over a fifty-five-year period, from 1960 to 2015. The challenge was that historical records are often incomplete; old maps might show a building existed, but they rarely tell us what it was made of, how many floors it had, or whether it was a home or a factory. To fill these gaps, the team combined detailed historical maps with modern machine learning. They taught computer models to recognize patterns in the shape and location of buildings, allowing them to predict the missing details of old structures with high accuracy. By layering these reconstructed facts over time, they created a dynamic, building-by-building history of the city's physical growth, turning a static map into a moving story of material accumulation.
The results reveal a city that grew much faster in terms of physical space than in terms of people. Between 1960 and 2015, Gothenburg's population increased by 29 percent, but the total floor area of its residential buildings more than doubled, rising by 102 percent. This means that while the city added more people, it added even more space for them to live in. Over this period, the city's residential buildings accumulated a net total of approximately 9.75 million tons of material. This figure represents the difference between the massive amount of new material brought in through construction and the much smaller amount removed through demolition. In total, about 10.9 million tons of residential materials entered the city's stock, while only 1.14 million tons were released when buildings were torn down. The city is essentially a one-way street for materials, with a vast majority of what was built remaining in place.
The timing of this growth tells a specific story about the city's development. The most intense period of building occurred in the late 1960s and early 1970s. During these years, the city added the largest volume of material in its history, with a peak of over 3 million tons entering the stock in just five years. This era of rapid expansion was followed by a significant slowdown in the 1980s and 1990s, before a modest increase in the 2000s. Interestingly, the number of new buildings added in the early 1970s was actually higher than in the late 1960s, but the total amount of material added was lower. This suggests that the buildings constructed in the earlier period were larger or used heavier materials, while the later surge consisted of more numerous but smaller structures. Demolition, by contrast, was a minor event in the grand timeline. Most tearing down of buildings happened in the earlier decades, and very little residential demolition has occurred since 1990, meaning the buildings from the mid-20th century boom are still standing today, holding the bulk of the city's material wealth.
Where this growth happened is just as important as when it happened. The new materials did not spread evenly across the city. Instead, they concentrated in a ring roughly 5 to 10 kilometers from the city center, with a particularly dense band around 8 kilometers out. This area saw the heaviest influx of concrete, steel, and aggregates, forming the backbone of the modern residential stock. While the city center remained a hub for commercial and industrial construction, the residential expansion pushed outward, creating a distinct spatial pattern of material accumulation. The study found that areas with higher population density today generally correspond to areas where more material was added in the past, but this relationship is not perfect. Some neighborhoods have a high density of people but a lower history of material addition, while others, particularly in the northern and western parts of the city, show massive material accumulation relative to their current population.
The composition of these materials is dominated by the heavyweights of the construction world. Concrete accounts for the largest share of both the new materials added and the old materials removed, followed by sand, aggregates, and steel. Because so little demolition has occurred in recent decades, the city's material stock is heavily weighted toward the concrete and steel poured during the mid-century boom. This creates a long-term legacy: the buildings standing today are not just shelters for people but also a massive, static archive of the city's industrial past. The study demonstrates that by combining historical data with modern computational tools, it is possible to see the city not just as a collection of addresses, but as a physical entity that has grown, shifted, and solidified over time. This detailed view provides a necessary foundation for understanding how cities store resources and how they might one day recover them, turning the built environment from a permanent sink of materials into a potential source for the future.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.