Unveiling patterns in human dominated landscapes through mapping the mass of US built structures

The study addresses how human-made structures have transformed landscapes to the point where their mass rivals and surpasses global biomass, a hallmark of the proposed Anthropocene. Socioeconomic material stocks—also termed manufactured capital, technomass, human-made mass, or in-use stocks—have grown to exceed 1000 Gt globally and have doubled roughly every two decades, closely tracking inflation-adjusted GDP. Buildings and mobility infrastructures constitute the majority of these stocks. The paper’s purpose is to quantify and map, at high resolution, the mass of built structures across the conterminous United States, to reveal patterns of material dominance, assess per-capita material intensity, and explore determinants and implications for resource use, environmental impacts, and planning for a circular economy.

The authors situate their work within debates on humanity’s dominance of ecosystems and the Anthropocene. Prior studies estimate global socioeconomic material stocks now exceed 1000 Gt (dry matter), approximately equal to the total biomass, with rapid doubling aligned with economic growth. Several conceptual frameworks exist for these stocks (manufactured capital, technomass, human-made mass, in-use stocks). Buildings and mobility infrastructure are recognized as the principal components by mass. The study builds on and extends this literature by providing spatially explicit, high-resolution mapping and per-capita intensity analyses for the US.

• Study scope: Conterminous United States; assessment year ca. 2018.
• Stock categories: Eight building types and nine mobility infrastructure types.
• Materials: Quantified the mass of 14 stock-building materials, assigning material composition factors to each structural category.
• Mapping resolution and products: High-resolution mapping at 10 × 10 m grid cells (with visualization also at 10 × 10 km and 100 × 100 m for figures), including a pseudo-3D map of total material stock. An interactive web viewer presents the full 2D map in t per 10 × 10 m grid cells.
• Data sources and geodata streams (examples cited): Microsoft building footprints; OpenStreetMap (OSM) for infrastructure; US Geological Survey National Land Cover Database; Council on Tall Buildings and Urban Habitat for skyscraper heights; various building height reference datasets; US Dept. of Energy building climate zones; US Census Bureau and US Bureau of Economic Analysis for socio-economic variables; state and county boundaries from the US Census Bureau.
• Typology harmonization: Structural types identifiable in geodata were aligned with material typologies. Material factors (composition and intensities) were derived by averaging multiple published sources per category to generalize and harmonize across the nation.
• Material factors and assumptions: Factors are conservative, based on minimum-standard specifications per construction codes. Variations due to local ground conditions, historical construction standards, and compliance were acknowledged.
• Roads and mobility infrastructure: For local and rural roads, assumptions on ratios of paved vs. unpaved and gravel vs. dirt surfaces were stratified by climate zones to estimate material stocks where direct differentiation was not available.
• Parking surfaces: All impervious areas not attributable to other above-ground built structures were classified as parking-related infrastructure, with validation against local case studies (e.g., Los Angeles County and Phoenix metro area) indicating good agreement.
• Uncertainty analysis: Variability in published material composition factors across categories was used to estimate uncertainty. Highest uncertainty occurs in climates requiring reinforced construction (wet and cold) and in abundant categories (low-rise residential buildings, local roads). Additional uncertainties arise from geospatial product errors and interactions between datasets; some uncertainty propagation is quantifiable (e.g., material factor variability), while others are more complex.
• Population and intensity metrics: Per-capita material intensity (t per capita) computed at county level to analyze spatial patterns and urban–rural differences.
• Visualization: Maps and histograms at county level; regional zooms (e.g., New York City; Lubbock, TX).

• Built structures in the conterminous US are 2.6 times heavier than all plant biomass across the country; most inhabited areas are mass-dominated by buildings or infrastructure.
• Population distribution by dominance: 26% of the US population lives in counties where mobility infrastructure prevails; this share is 18% for urban populations and 62% for rural populations. Urban populations predominantly live where building stocks dominate.
• Average per-capita material intensity of built structures: 391 t per capita.
• Extremes in per-capita intensity: Bronx, NY has the lowest at 90 t per capita; Loving County, TX has the highest at 42,691 t per capita (potentially 33,445 t per capita under lower-bound road surface assumptions).
• Spatial patterns: Highest building-related intensities in the northern Great Plains; large regions with lower intensities in the West, Southwest, and Southeast; urban centers (e.g., Boston–Washington corridor) exhibit particularly low intensities.
• Mobility vs building disparities: Mobility infrastructure shows much higher spatial disparity in per-capita intensity (interquartile range ≈ 595 t per capita at county level) compared to buildings (≈ 110 t per capita), reflecting that roads serve broader functions beyond local shelter and are more spatially extensive to enable access for agriculture, forestry, resource extraction, processing, and trade.
• Parking surfaces account for approximately 11% of the total stock, or about 43 t per capita, with estimated areal shares aligning with local studies (e.g., ~13.53% in LA County; ~10.33% in Phoenix metro).
• Densely built settlements have substantially lower per-capita material stocks; sparsely populated regions exhibit the highest intensities due to ubiquitous infrastructure.
• Out-migration intensifies per-capita stocks in rural areas because built structures persist while population declines.

The findings demonstrate that the mass of built structures now surpasses plant biomass nationwide and that this mass is distributed such that most inhabited areas are dominated by buildings or infrastructure. By quantifying stocks at high spatial resolution and linking them to population, the study reveals that dense urban settlements achieve lower per-capita material stocks, whereas sparsely populated regions face high per-capita intensities driven largely by extensive infrastructure networks. This suggests that compact urban form can reduce material requirements per person, with implications for land take, resource demand, and lifecycle impacts (including GHG emissions) associated with construction, maintenance, and operation. The pronounced disparity in mobility infrastructure intensity underscores that many infrastructures serve functions disconnected from local population counts, underpinning economic activities and interregional connectivity. The study highlights that per-capita metrics localized to place may not capture responsibilities and drivers where stocks serve distant populations, indicating the value of a spatially explicit, consumption-based perspective to complement production- or location-based assessments. Overall, the results provide a biophysical basis for policies targeting resource-efficient settlement design and strategies for a sustainable circular economy, emphasizing the need to manage and retrofit existing stocks and to optimize infrastructure provisioning relative to service demand.

This work delivers the first high-resolution, nationwide mapping of the mass of US built structures, disaggregated by building and mobility infrastructure types and material composition. It shows that built structures outweigh plant biomass by a factor of 2.6 and that per-capita material intensity varies markedly across the urban–rural gradient, with dense urban areas achieving lower intensities and sparsely populated areas exhibiting very high intensities due to extensive infrastructure. The study demonstrates the value of integrating geospatial datasets with harmonized material factors to quantify and visualize the biophysical foundations of society. Future research directions include: improving parameterization of material composition factors with more localized and time-resolved data; refining assumptions for road surface types using direct observations; advancing uncertainty propagation across geospatial layers; linking local stocks to non-local demand via consumption-based accounting; and enhancing validation through detailed local audits and longitudinal analyses of stock dynamics.

• Material composition uncertainties: High variability across published material factors, especially in climates requiring reinforced construction (wet and cold), and in abundant categories (low-rise residential buildings, local roads).
• Conservative assumptions: Factors based on minimum-standard specifications may understate actual material quantities where local ground conditions, historical standards, or compliance issues increase material use.
• Road surface assumptions: County-level shares of paved vs unpaved and gravel vs dirt roads are assumed using climate-zone stratification; deviations can affect local estimates (e.g., Loving County’s peak intensity could drop from 42,691 to 33,445 t per capita under lower-bound assumptions).
• Parking classification: All impervious areas not otherwise classified are treated as parking infrastructure; while comparisons suggest reasonable accuracy, misclassification remains possible.
• Geospatial data errors and interactions: Uncertainties in input datasets (building footprints, heights, land cover, infrastructure mapping) and their interactions may compound or offset; not all such effects are fully quantifiable.
• Service allocation: Localized per-capita intensities do not account for stocks serving non-local populations (e.g., intercity infrastructure, government, industrial/commercial facilities), potentially biasing interpretations of local responsibility.