Urban carbon emissions: evolution, analysis, and application based on 53-year data

As one of the core drivers of global climate change, the accumulation and emission of carbon dioxide (CO₂) in urban environments are profoundly affecting urban development from ecological, economic, and social dimensions, while also posing significant hazards 123. From a positive perspective, CO₂ is a key raw material for plant photosynthesis. Rational green space planning in cities can utilize CO₂ to promote vegetation growth, thereby enhancing urban ecological resilience and improving air quality 4.Meanwhile, with the development of low-carbon technologies, the resource-oriented utilization of CO₂ has also provided a new direction for urban industrial upgrading, driving the rise of green industries such as new energy and environmental protection, and emerging as a potential growth point for urban economic transformation 567.However, its negative impacts and hazards are more prominent and far-reaching. Ecologically and environmentally, cities are high-density CO₂ emission areas, and massive emissions have intensified the urban heat island effect-a phenomenon where urban center temperatures are significantly higher than those in surrounding suburbs. This not only increases energy consumption for cooling in summer but also may trigger extreme high-temperature weather, threatening residents' health 89. Additionally, global climate anomalies caused by the greenhouse effect expose cities to higher risks of disasters such as frequent rainstorm-induced waterlogging and sea-level rise, undermining the safety of urban infrastructure 10. Economically, to address environmental issues caused by excessive CO₂ emissions, cities need to invest huge funds in upgrading high-energy-consuming infrastructure, controlling pollution, and restoring ecosystems, which will increase fiscal pressure in the short term. Furthermore, the damage to urban lifeline systems caused by extreme climate disasters directly results in economic losses, affecting the normal operation of cities and their investment environment. Socially and in terms of public health, although high-concentration CO₂ itself is non-toxic, it indirectly exacerbates urban air pollution 111213. This may reduce air visibility and induce health problems such as respiratory diseases and cardiovascular diseases. Moreover, issues like agricultural output reduction and water scarcity caused by climate anomalies may indirectly affect the stability of urban food supply and intensify the pressure of social resource allocation 141516. In summary, CO₂ exerts a "dual-edged sword" effect on urban development. However, the hazards caused by current excessive emissions have far exceeded its limited positive role. Promoting urban low-carbon transformation and controlling CO₂ emissions have thus become core tasks for ensuring the sustainable development of cities.In the process of analyzing the impact of carbon emissions on urban development, the establishment of a carbon emission database holds irreplaceable significance. Firstly, by systematically collecting carbon emission data from various sectors within a city-including industry, construction, transportation, and residential life-the database can accurately identify high-emission areas and key sources of carbon emissions 1718. This provides a data foundation for targeted analysis of the specific harms caused by carbon dioxide emissions from different sectors to urban ecology, economy, and public health. Secondly, the functions of dynamic monitoring and trend analysis in a carbon emission database serve as core tools for assessing the evolutionary patterns of carbon emissions 19. Through long-term accumulation of carbon emission data, the database enables the construction of correlation models between carbon emissions and urban development (e.g., GDP), facilitating quantitative analysis of the corresponding relationship between changes in carbon emission intensity and urban development. It is evident that a carbon emission database is not only a measuring tool for quantifying carbon dioxide emissions but also a microscope for analyzing its urban impacts and a navigator for formulating governance plans 20. The level of its establishment and improvement directly determines the depth of understanding of urban development impacts and the effectiveness of response measures 21. It constitutes the core foundation for promoting the transformation of urban carbon governance from experience-based judgment to data-driven decision-making. Collectively, these studies establish the importance of emission driver analysis but suffer from two gaps: (1) reliance on fragmented or short-term data, and (2) insufficient attention to intra-urban regional differences.To mitigate the hazards of excessive carbon dioxide (CO₂) emissions on urban development, China has implemented a series of targeted technological measures and built a multi-dimensional carbon reduction technology system, focusing on key areas such as energy structure optimization, industrial upgrading, and infrastructure renovation. For instance, on the supply side, China has vigorously promoted the application of renewable energy substitution technologies. Within urban areas, clean energy power generation systems-including photovoltaic (PV), wind, and biomass energy-have been widely deployed 222324. Meanwhile, ultra-high voltage (UHV) power transmission technology is used to deliver cross-regional clean energy to cities, gradually reducing urban reliance on thermal power and cutting CO₂ emissions related to energy consumption at the source. In the construction sector, green building technologies have been fully popularized 25. Through new thermal insulation materials, passive building designs, and renewable energy-based heating and cooling technologies, energy consumption throughout the entire life cycle of buildings has been reduced. At the same time, energy-saving retrofits of existing buildings have been advanced to improve the energy efficiency of established structures. These technological measures work in synergy to drive urban carbon reduction across the entire chain of source control, emission reduction, recycling, and sequestration 2627. They provide core technological support for alleviating the hazards of CO₂ on urban development and realizing low-carbon, sustainable urban development.Meanwhile, China put forward the Dual Carbon goals in 2020, namely achieving carbon peaking by 2030 and carbon neutrality by 2060 28. To this end, a series of policies have been formulated. For example, the 2024-2025 Action Plan for Energy Conservation and Carbon Reduction issued by the State Council emphasizes the following key measures: Strictly controlling coal consumption, advancing the low-carbon transformation of coal-fired power and the triple renovation coordination, and cutting non-electric coal use 29. In key regions for air pollution prevention and control, newly built, renovated, and expanded coal-consuming projects shall implement coal substitution with equivalent or reduced quantities. Vigorously developing large-scale wind and photovoltaic power bases, with a focus on deserts, Gobi, and arid areas; developing offshore wind power in a rational and orderly manner; and promoting the utilization of distributed new energy 3031. Advancing the low-carbon transformation of transportation infrastructure and encouraging the construction of green highways and railways.Promoting the shift of transportation equipment to new energy, advancing the electrification of vehicles in the public sector, and developing zero-emission freight fleets 3233.For newly built buildings, requiring that by the end of 2025, all newly built urban buildings fully comply with green building standards; increasing the coverage rate of photovoltaic systems on the rooftops of newly built public institution buildings and factory buildings; raising the renewable energy substitution rate for urban buildings; and expanding the floor area of newly built ultra-low energy consumption and near-zero energy consumption buildings 3435 The total carbon emissions of Hangzhou from 1970 to 2023 are shown in Figure 1.As can be seen from the figure, the total carbon emissions of Hangzhou exhibited a rapid growth trend. Specifically, the total carbon emissions of Hangzhou stood at 4152.27×10³ As a core district of Hangzhou, Xiaoshan maintains intensive economic and industrial linkages with neighboring districts-for instance, the spillover of manufacturing activities and energy consumption between Xiaoshan and Binjiang, Gongshu, and Yuhang districts directly affects their mutual carbon emission dynamics. Furthermore, this article uses the carbon emissions of eight adjacent regions to predict the emissions of Xiaoshan, with the goal of capturing the spatial spillover effects of interrelated regional development patterns and regional level carbon emissions.Combined with the artificial neural network (ANN) method, the carbon emission data of Binjiang District, Fuyang District, Gongshu District, Shangcheng District, Xihu District, Yuhang District, Qiantang District, and Linping District were used as input, while the carbon emission data of Xiaoshan District were used as output to analyze the impact of the prediction horizon on the model's prediction performance. Specially, MLP was selected because it effectively models the nonlinear relationships between multi-source regional emission data without imposing strict assumptions on data distribution, which is suitable for our cross-district correlation analysis. The hidden layers use Meanwhile, when the number of predictions is the same, the prediction performance fluctuates with the increase of the number of hidden layers. For example, when the predicted number is 1 year and the number of hidden layers is 1, the MAPE is 0.03548; When the number of hidden layers is 2, the MAPE is 0.07236; When the number of hidden layers is 3, the MAPE is 0.16655; When the number of hidden layers is 4, the MAPE is 0.20551; When the number of hidden layers is 5, the MAPE is 0.10015; When the number of hidden layers is 6, the MAPE is 0.08698; When the number of hidden layers is 7, the MAPE is 0.14562; When the number of hidden layers is 8, the MAPE is 0.3485;When the number of hidden layers is 9, the MAPE is 0.20822; When the number of hidden layers is 10, the MAPE is 0.07561; When the number of hidden layers is 11, the MAPE is 0.02716; When the number of hidden layers is 12, the MAPE is 0.10328; When the number of hidden layers is 13, the MAPE is 0.26043; When the number of hidden layers is 14, the MAPE is 0.05878; When the number of hidden layers is 15, the MAPE is 0.10927; When the number of hidden layers is 16, the MAPE is 0.06443; When the number of hidden layers is 17, the MAPE is 0.02863; When the number of hidden layers is 18, the MAPE is 0.00364; When the number of hidden layers is 19, the MAPE is 0.02547;When the number of hidden layers is 20, the MAPE is 0.03163. It can be seen that the predictive performance of BP neural network will fluctuate to some extent under different hidden layers and different prediction quantities, indicating poor robustness. However, for the prediction of carbon emissions in Hangzhou, multiple predictions can be averaged to reduce prediction errors; At the same time, errors can be reduced by debugging the number and layers of hidden layers.Clearly, the MAPE values in Figure 4 As an emerging environmental right, carbon emission rights serve as the cornerstone for building a carbon emission market 36. Amid the global effort to address climate change, China has actively promoted the development of a carbon emission rights trading system. By setting scientific and reasonable total carbon emission targets, emission rights are allocated to key emitting enterprises in the form of quotas. These enterprises can trade their quotas in the market based on their own emission reduction performance. This market-oriented mechanism incentivizes enterprises with low emission reduction costs to exceed their reduction targets and sell surplus quotas for profit. Conversely, enterprises facing high difficulties in emission reduction can purchase quotas to meet compliance requirements. In this way, the mechanism not only stimulates enterprises' initiative to reduce emissions independently but also achieves the overall emission reduction target at a lower social cost. Specifically, the annual total carbon emission quotas are formulated in a scientific and reasonable manner, taking into account factors such as the industrial development stage, historical emission data, future emission reduction potential, and the needs of economic and social development. In the initial stage, quota allocation is dominated by free distribution, with appropriate inclination toward high-energy-consuming, high-emission industries that face significant challenges in emission reduction-this ensures a stable industrial transition. As the market matures, the proportion of paid allocation will be gradually increased to enhance the efficiency and fairness of quota allocation. Meanwhile, a quota reserve mechanism will be established to respond to sudden market fluctuations and stabilize market expectations.Gradient carbon pricing is an innovative price regulation tool that plays a key role in optimizing the allocation of carbon emission resources. The traditional single carbon price struggles to fully reflect differences in the marginal abatement costs and emission reduction potential among different enterprises. In contrast, gradient carbon pricing sets different price ranges based on enterprises' carbon emission levels. When an enterprise's carbon emissions are at a low level, a lower carbon price applies-this encourages the enterprise to maintain and further reduce its emissions. For high-emission enterprises, however, as their emissions exceed the specified threshold, the carbon price increases significantly and progressively. This raises their emission costs, creating a strong constraint. This mechanism compels enterprises to fully consider carbon emission factors when formulating production decisions and technological innovation strategies, guiding resources to flow toward low-carbon and green sectors. For example, in some pilot regions, after the implementation of gradient carbon pricing, high-energy-consuming enterprises have accelerated the technological transformation for energy conservation and emission reduction; some enterprises have even proactively adjusted their industrial structure to transition to low-carbon industries, effectively driving the green development of the regional economy.Overall, Xiaoshan's emissions are dominated by energy-intensive industries, that is, automobile manufacturing, chemical engineering, textile printing and dyeing, and air logistics, that is, Xiaoshan International Airport's passengers over 40 million and 800,000 tons of cargo in 2023. Enterprises exceeding this threshold would face a progressive carbon price increase, e.g., 10% premium for 10-20% overshoot, 30% premium for over 20% overshoot. This directly aligns with Hangzhou's post-2020 emission stabilization by incentivizing Xiaoshan's high-emission sectors to reduce emissions without disrupting industrial transition. For carbon emission rights trading, it suggest that connect the free initial allocation + gradual paid allocation mechanism to Xiaoshan's industrial structure.Initially, 80% of quotas are freely allocated to Xiaoshan's key high-emission enterprises to ensure stable operation during low-carbon transformation. As the market matures, the paid allocation ratio will increase to 30% by 2030, encouraging enterprises to optimize energy efficiency. For Binjiang District, surplus quotas from low-emission IoT/AI enterprises can be traded to Xiaoshan's enterprises, creating a district-level carbon market that supports Hangzhou's overall stabilization goals.The paper's core objective is decomposed into three specific, measurable