Global Carbon Metabolism in 16 Cities: An Integrated Analysis of Embodied and Virtual Carbon Opens New Perspectives on Emission Reduction

      Cities, as the core carriers of human production and life, contribute more than 70% of global carbon emissions. Their carbon metabolism processes—including the input, transformation, storage, and output of carbon—represent a key entry point for global climate governance. However, traditional urban carbon accounting has long focused on gaseous emissions within the "territorial scope" (such as direct emissions from fossil fuel combustion), while neglecting two core components: first, the cross-border flow and long-term storage of physical carbon (the actual carbon contained in materials and products); and second, the transfer effects of virtual carbon (emissions embedded in upstream supply chains, such as electricity generation in other regions and raw material extraction). This omission has resulted in significant "blind spots" in understanding urban carbon metabolism, making it difficult to support the formulation of precise emission reduction policies.

        To address this research gap, a team led by Professor Chen Shaoqing from Sun Yat-sen University, in collaboration with researchers from Beijing Normal University, the University of Maryland in the United States, and other institutions, has developed an integrated metabolic analysis framework that combines "physical carbon balance" and "fossil fuel-derived virtual carbon footprint." Using 16 representative cities worldwide—including Beijing, Hong Kong, New York, London, Tokyo, and São Paulo, covering high-, middle-, and low-income levels and different geographical regions—the team completed a full-chain analysis of urban carbon metabolism. The findings, published in Nature Communications under the title "Physical and virtual carbon metabolism of global cities," provide a new academic paradigm and practical basis for urban carbon management.
 

        The core innovation of this study lies in breaking the limitations of traditional carbon accounting—its "single-dimensional, local-scope" approach—by integrating multiple methods to achieve a systematic characterization of carbon metabolism. On one hand, material flow analysis (MFA) is used to track the complete path of physical carbon—from "input" (imports from outside the city, local ecological extraction, and material recycling) to "allocation" (consumption by sectors such as agriculture, manufacturing, and construction) and finally to "destination" (gaseous emissions, solid waste, household storage, urban stock, and exports), ensuring the material balance of physical carbon flows. On the other hand, life cycle assessment (LCA) is combined with input-output analysis (IOA) to quantify fossil fuel-derived virtual carbon—the CO₂ emissions generated in the upstream supply chains of imported goods and services—and accurately allocate them to final demand categories such as household consumption, capital formation, and exports.On this basis, the study defines "total carbon inflow (TCI)" as a comprehensive indicator (physical carbon input + virtual carbon input), achieving for the first time a full quantification of the scale and structure of urban carbon metabolism.Using this framework, the research reveals three key patterns of global urban carbon metabolism, directly addressing the blind spots of traditional accounting:

       Physical carbon is highly dependent on external "carbon outsourcing." Among the 16 cities, 88%-92% of physical carbon inputs come from imports outside the city boundaries, only 2%-6% originate from local ecosystem extraction (such as urban forest biomass), and 3%-8% come from material recycling. Even in cities with relatively well-developed recycling systems like Stockholm and Vienna, recycled carbon accounts for only about 8% of physical carbon input and less than 5% of total carbon inflow. In cities such as Moscow and Bangkok, imported physical carbon accounts for more than 50% of the total carbon balance, highlighting the deep dependence of cities on global supply chains.

Virtual carbon constitutes an important component of total carbon inflow, accounting for 33%-68% (highest in Hong Kong, lowest in São Paulo) and contributing nearly half of total carbon inflow on average. From the demand perspective, 30%-53% of virtual carbon is driven by local consumption and investment. In high-income cities like Tokyo and New York, household consumption accounts for more than 30% of virtual carbon, while in developing cities such as Beijing and Delhi, about 45% of virtual carbon originates from capital formation due to infrastructure construction needs. This means that ignoring upstream emissions would seriously underestimate the actual impact of urban carbon metabolism.

Urban stock carbon forms a "hidden emission source." In the 16 cities, 8%-24% of carbon is stored in durable goods (furniture, textiles) or urban stock (buildings, infrastructure). Per capita stock carbon reaches 0.6-1.5 t C, more than twice the global average, and its emissions exhibit significant time lag—this carbon will be slowly released in the coming years to decades through waste incineration, landfilling, and other pathways, making it an unavoidable target for future emission reduction.

        The academic and practical value of this study is particularly prominent. At the theoretical level, its methodology integrating MFA, LCA, and IOA fills the research gap in "physical-virtual carbon" coupling studies involving large-scale urban samples, deepening the understanding of the complexity of urban carbon metabolism systems. At the practical level, the research provides precise directions for urban emission reduction—shifting from "only controlling local emissions" to a dual-track approach of "managing supply-chain virtual carbon + optimizing urban stock carbon"—and emphasizes the need to develop differentiated strategies for different types of cities (e.g., high-income cities should focus on consumption-side virtual carbon, while developing cities should prioritize emissions related to capital formation and stock carbon management). For global climate governance, this study not only reveals the "hidden ledger" of urban carbon metabolism but also offers a reference pathway for collaborative emission reduction among cities at different development levels, contributing to the achievement of global carbon neutrality goals.

      Link：Physical and virtual carbon metabolism of global cities | Nature Communications