China’s CO2 peak before 2030 implied from characteristics and growth of cities


China pledges to peak CO2 emissions by 2030 or sooner under the Paris Agreement to limit global warming to 2 °C or less by the end of the century. By examining CO2 emissions from 50 Chinese cities over the period 2000–2016, we found a close relationship between per capita emissions and per capita gross domestic product (GDP) for individual cities, following the environmental Kuznets curve, despite diverse trajectories for CO2 emissions across the cities. Results show that carbon emissions peak for most cities at a per capita GDP (in 2011 purchasing power parity) of around US$21,000 (80% confidence interval: US$19,000 to 22,000). Applying a Monte Carlo approach to simulate the peak of per capita emissions using a Kuznets function based on China’s historical emissions, we project that emissions for China should peak at 13–16 GtCO2 yr−1 between 2021 and 2025, approximately 5–10 yr ahead of the current Paris target of 2030. We show that the challenges faced by individual types of Chinese cities in realizing low-carbon development differ significantly depending on economic structure, urban form and geographical location.


The Paris Agreement stresses the urgency of building a new climate-safe future by constraining the increase in global average surface temperature to less than 2 °C (or even 1.5 °C) above pre-industrial levels1. Responsible for 71–76% of global carbon dioxide (CO2) emissions derived from energy use, cities are recognized as a key arena for efforts to mitigate future climate change2. The urban shares of global CO2 emissions may continuously grow, reflecting the projection that the global urban population will increase from 3.4 billion today to 6.3 billion in 20503. A number of cities around the world have taken independent action to mitigate the potential impacts of climate change4,5,6. The C40 cities, a network of 91 megacities dedicated to tackling climate change, committed to invest US$375 billion in low-carbon infrastructure over three years to ensure that their emissions would peak by 2020 and be nearly halved by 20301. Recently, more than 300 city mayors in the United States have committed to “adopt, honour and uphold the commitments to goals enshrined in the Paris Agreement” in response to President Trump’s withdrawal from the agreement.

In China, urbanization and industrialization are two pillars of economic development7. They are associated with intensive fossil-fuel consumption, especially of coal, bringing serious climate and air pollution challenges to cities and their surrounding regions. In 2010, China’s urban population reached 670 million, nearly 50% of the total population8. Another 280 million formerly rural residents are expected to migrate into cities by 20253, equivalent to the creation of a megacity like Beijing or the New York metropolitan region every year from 2010 to 2025. As urban residents on average consume nearly three times as much energy as their rural counterparts9,10, this high speed of urbanization will exert huge pressures on China’s future energy systems and thus its ability to meet CO2 mitigation targets. Simultaneous challenges also arise for China to address its severe urban air pollution, which mainly originates from combusting the same fossil fuels that produce CO2.

To combat the intertwined challenges of global climate change and domestic air pollution, the Chinese government has set targets to mitigate CO2 emissions and more than 100 Chinese cities since 2009 have declared their own ‘eco-city’ plans11. This trend received further impetus from a National Development and Reform Commission programme that selected 8, 26 and 45 cities, respectively, as low-carbon pilots in 2010, 2012 and 2017. Judging China’s success at realizing low-carbon urbanization, however, is impeded by weak quantification of CO2 emissions from its cities due to the scarcity and low quality of basic data (for example, for energy consumption) at the city level.

Several studies have reported emissions of greenhouse gases from Chinese cities, especially megacities such as Shanghai, Beijing and Tianjin12,13,14,15 and provincial capitals16. We have developed urban-scale CO2 emission inventories in China17,18, including those for 12 cities during 2000–200819. Recently, researchers20,21,22 have tried to classify Chinese cities into groups and to investigate carbon mitigation strategies for each group based on emissions of prefecture-level cities. However, the types of cities defined in those studies differed in various respects (for example, economic output, industrial structures, urbanization level and geographical characteristics), reflecting diverse research objectives, and their inconsistency with China’s official classification weakens their value as direct references for cities designing the low-carbon strategies. Also, the methods, scopes and primary data applied in these earlier studies are different, making their results difficult to compare. Furthermore, these previous studies usually report emissions of cities for a single year, for example, 2012 in Cai et al.21 and 2010 in Shan et al.22, or time-series emissions for a single city17,18, making it challenging to draw conclusions about dynamic changes of emissions across cities. This is particularly important for developing countries such as China, which are experiencing swift rates of urbanization across cities23.

Here, we explore carbon emissions for 50 Chinese cities over a 17 yr period from 2000 to 2016, using an evaluation framework suggested by the Intergovernmental Panel on Climate Change. Our analysis also considers prominent transboundary emissions relating to electricity and heat consumption in cities as evaluated by Kennedy et al.12,14 and in other studies24,25,26. Here we focus on CO2 emissions by sector from energy consumption and industrial processes, grouping sources into transportation, buildings and industry categories (the last including both industrial energy consumption and industrial processes, for example, calcination in cement making). We emphasize the implications of the 50 cities’ emission trajectories for China’s aggregate national emissions. The research framework (Supplementary Fig. 1) and details of the analytical approach are presented in the Methods. We find significant diversity in the levels and trends of CO2 emissions for individual Chinese cities, highlighting their disparities in terms of wealth, economic structure, urban form and geographical location. However, emissions of most cities reach their peaks and begin to decline when their levels of per capita gross domestic product (GDP) (2011 purchasing power parity (PPP)) reach around US$21,000 (80% confidence interval (CI): US$19,000 to 22,000), suggesting a peak in China’s national emissions of 13–16 GtCO2 yr−1 at some point between 2021 and 2025.

Emission characteristics of 50 Chinese cities

The 50 cities analysed here represent substantial differences in terms of physical conditions (for example, geographical location) and socioeconomic status (for example, economic development and structure) (Fig. 1a and Supplementary Table 1). We selected at least one city in each of the 31 provinces and provincial-level municipalities across China except for Tibet, each chosen to be representative of the most densely populated areas of these jurisdictions. The socioeconomic characteristics of the 50 cities exemplify differences in economic wealth, from less-developed (for example, per capita GDP of US$4,480 (2011 PPP) in Zunyi in 2010) to more-developed (for example, Shenzhen at US$27,900) cities, and differences in economic structures, from industrial (for example, Linfen and Tangshan) to service-oriented (for example, Beijing and Shanghai) cities. Cumulatively, the 50 cities were responsible for 3.50 billion tons of CO2 emissions in 2015, nearly 35% of China’s total emissions estimated from fossil-fuel combustion and cement production27. The combined share of the population in these cities is 30%, while their contribution to the GDP is 51% (Supplementary Fig. 2). As China’s national carbon emission intensity (that is, CO2 emissions per unit of GDP) has declined continuously over time28, these results suggest greater progress in mitigating carbon emission intensity in these cities, reflecting an inverse relation between trends in shares of emissions and GDP over the period 2000–2016.

Fig. 1: CO2 emissions and trends of 50 Chinese cities from 2000 to 2016.

a, Population density and geographical locations and boundaries of the 50 cities. b, Per capita CO2 emissions of each city. The cities are coded by numbers as follows: 1, Shenzhen; 2, Guangzhou; 3, Beijing; 4, Shanghai; 5, Chengdu; 6, Xi’an; 7, Chongqing; 8, Shenyang; 9, Hangzhou; 10, Wuhan; 11, Tianjin; 12, Nanjing; 13, Haikou; 14, Nanchang; 15, Zhengzhou; 16, Nanning; 17, Fuzhou; 18, Changsha; 19, Baoding; 20, Nantong; 21, Hefei; 22, Xiamen; 23, Yangzhou; 24, Linfen; 25, Changchun; 26, Qingdao; 27, Harbin; 28, Kunming; 29, Guiyang; 30, Xuzhou; 31, Jinan; 32, Dalian; 33, Hohehot; 34, Wuxi; 35, Ningbo; 36, Handan; 37, Shijiazhuang; 38, Changzhou; 39, Taiyuan; 40, Lanzhou; 41, Suzhou; 42, Urumqi; 43, Xining; 44, Yinchuan; 45, Tangshan; 46, Yancheng; 47, Zunyi; 48, Lianyungang; 49, Zhenjiang; 50, Yichang. Consistent with the new standard of city classification published by China State Council in 2014, megalopolis, metropolis, large cities, and middle and small cities represent the cities with a residential population living in the urban area of more than 10 million, 5–10 million, 1–5 million, 0.5–1 million and less than 0.5 million in 2010, respectively.

In 2016, the average per capita emissions for Chinese megalopolises, metropolises, large cities and small cities (defined in Fig. 1a and Supplementary Table 1) were 7.3 ± 1.6 (mean ± s.d.), 8.5 ± 4.3, 9.6 ± 6.3 and 6.5 ± 1.8 tCO2, respectively. Most of the 50 cities selected here have seen a significant increase in per capita emissions during 2000–2016, with an average growth rate of 192% (Beijing is one of the exceptions, having declined 7.4%). The average emission growth rate of the four megalopolises—Shenzhen, Guangzhou, Beijing and Shanghai—was 49.3% (range −7.4–152.5%), lower than that for the metropolises at 110.3% (range 67.7–124.1%), large cities at 190.0% (range 12.4–664.0%) and middle/small cities at 336.4% (range 127.6–643.0%). The megalopolises are located in China’s eastern coastal region, identified with advanced economies, large populations and high urbanization rates, and with strong policy incentives to address environmental pollution. Industries, especially heavily polluting ones, account for very small shares of the economic production in these cities, and their industrial carbon emissions have already peaked and have begun to decline in recent years (Supplementary Table 1 and Supplementary Fig. 3). Their future emission trends depend more on sources associated with transportation and buildings, relating to the improvement of living standards. The megalopolises seem to have reached the turning point of their overall emissions, peaking around 10 tCO2 per capita (except Shenzhen at ~6.0 tCO2 per capita) between 2000 and 2016 (Fig. 1b and Supplementary Table 2).

The metropolises (Supplementary Table 1) are usually the provincial capital cities, with large populations and advanced socioeconomic statuses like the megalopolises. They follow similar paths for emissions as megalopolises, and it appears that related industrial emissions either have peaked or are approaching a peak (for example, in Hangzhou and Xi’an). The large cities have the highest average per capita emissions among the four types of cities, reflecting their rapid recent industrialization and urbanization. They are mostly the fastest-growing cities in China, with large proportions of heavy industries or resource-based economies (for example, Tangshan and Yinchuan), and their emission trends will have an important influence on China’s future emission trajectory. Medium and small cities are usually the less-developed regions that are at early stages of industrialization and urbanization, which likely explains why they have the lowest values but highest growth rates in average per capita emissions. As their urbanization rates and industrial structures are far from mature, they have also more opportunities to leapfrog the traditional development mode characterized by a mindset of ‘pollute now and clean up later’.

Factors explaining the disparities in emissions

The great diversity in CO2 emissions (for example, per capita emissions of Tangshan are six times those of Shenzhen) and trends across individual cities, even those in the same size category (Fig. 1b), highlights disparities for these cities in terms of wealth, economic structure, urban form and geographical location.

Per capita GDP appears to be the primary driver for CO2 emissions from Chinese cities (Supplementary Fig. 4), indicating that the economic wealth of cities may be a key factor in explaining their emissions14,19. However, the significant variability in per capita emissions with the same affluence across the 50 cities (Supplementary Fig. 5a) implies that factors other than affluence, especially economic structure19, might greatly influence emissions. Chinese cities usually represent hubs driving regional economies and serving as the bases of many emission-intensive industries that produce goods for domestic and global export markets29. It is thus not surprising that the industrial shares of total emissions average nearly 70% (with a range of 35–90%) in the 50 cities (Fig. 2), which is very different from cities of developed countries. It means that the high per capita emissions in most Chinese cities are driven more by production than consumption. The major contribution from industrial emissions, reflecting very different industrial or subindustrial distributions, thus accounts for most of the great variability of emissions, as well as the weak correlation between total/industrial emissions and the total/industrial outputs across Chinese cities (Supplementary Fig. 5a,b). By contrast, the buildings and transportation sectors, closely related to consumption, contributed little to the total emissions from Chinese cities, averaging 14% for buildings and 12% for transportation. Their emissions are found to have grown greatly, however, and are significantly correlated with urbanization levels and tertiary industrial outputs across the 50 cities (Supplementary Fig. 5c,d).

Fig. 2: Per capita CO2 emissions of the 50 Chinese cities in 2010.

a, Northern Chinese cities. b, Southern Chinese cities.

For an individual city, its consumption of energy and related emissions may be influenced by the geographical location. For example, the central government established winter heating practices for homes and offices through the provision of coal for heating boilers in North China, specifically restricted to north of the Huai River and the Qinling Mountain range30. These long-lived heating systems continue to make indoor heating much more common in North China today. The average per capita CO2 emissions of cities in North China reached 10.2 t in 2010, 54% higher than for southern cities (Fig. 2), partly as a result of the consumption of fuels, especially coal31, deployed to heat buildings during the cold season.

Geographical location is also a key factor determining the local energy endowment and thus influencing emissions per unit of energy use in the city. The emissions from electricity use32 in some northern cities, such as Tianjin, Jinan, Tangshan and Taiyuan, is nearly 1,100 tCO2 GWh−1, attributable to the extensive coal production in regions surrounding those cities and the consequent high reliance on coal-fuelled power (for example, over 95% electricity consumption was generated by coal in 2010). But the aggregate emission factor for electricity generation was as low as 860 tCO2 GWh−1 for some coastal cities such as Shanghai, Hangzhou and Fuzhou in southeast China, benefiting from proximity to hydro and nuclear power plants in these areas (for example, 12–45% of electricity used was generated by zero- and low-carbon energy sources, including hydro, nuclear and natural gas). Therefore, switching from coal-fired to cleaner electricity generation, especially in northern Chinese cities where wind and solar power potentials are abundant but have not been fully exploited, could be critical for mitigating their emissions.

Urban form, including land use and the existence of public transit systems, influence the population distribution and density, another factor accounting for energy consumption and related emissions from Chinese cities, especially those from traffic (Supplementary Fig. 5e). Per capita traffic emissions amounted to 0.8–1.0 tCO2 for most Chinese cities, with population densities defined by the number of people per square kilometre for constructed areas ranging from 15,000 to 30,000 in 2010. The lower-density cities, except Chengdu and Kunming, have per capita traffic emissions greater than 1.0 tCO2, while emissions for higher-density cities (>30,000) are usually under 0.7 tCO2. As cities with higher population densities are usually more affluent, as defined by per capita GDP, these results reflect the fact that people prefer to live in more developed cities, usually where they have access to better public transportation systems. This increases the use of public transport (for example, metro and bus) and reduces per capita traffic emissions in China’s developed cities. A recent study33 supports the conclusion that urban population density tends to be more critical than energy-efficient technologies in mitigating energy use. Urban form also has lock-in effects that influence the baseline and trajectories of CO2 emissions in cities34. For example, Shenzhen, a newly developed city with high-tech industries that has grown at an accelerated rate since 1979, had a very low emission intensity of 0.2 tCO2 per US$1,000 of GDP in 2000, 30% of that for Shanghai, a traditional manufacturing-based city with similar per capita GDP. During 2000–2016, Shanghai greatly decreased its emission intensity, by 44%, but this is still more than two times that of Shenzhen, the emission intensity of which has been relatively stable.

Turning point of China’s future emissions

Although local emissions and GDP are weakly correlated across the 50 Chinese cities (Supplementary Fig. 5a), per capita emissions for individual cities (that is, controlling for geographical location, urban form and economic structure) correlate significantly with economic wealth over time. In addition, the relationship between per capita emissions and per capita GDP (in 2011 PPP, US$) for each city is well described by the traditional shape of the environmental Kuznets curve (Supplementary Table 2 and Supplementary Fig. 6). It suggests that a bell-shaped curve with a foreseeable peak in per capita CO2 emissions may describe well Chinese cities as they become richer, similar to what has been observed already in the megalopolises and some metropolises. As shown in Supplementary Table 2, peaks of per capita emissions and the years that these peaks occurred differ significantly across the 50 Chinese cities (for example, Beijing peaked at 9 tCO2 around 2007 while Yinchuan seems far from its peak emissions). However, most cities tend to reach peaks in emissions when per capita GDP level reach around US$21,000 (80% CI: US$19,000 to 22,000).

Chinese average per capita GDP was US$13,500 (2011 PPP) in 2015 with per capita emissions of 7.5 t. Although many researchers believe that the environmental Kuznets curve is reflected in the trends of China’s CO2 emissions, there is no consensus as to the time and peak value of CO2 emissions when China reaches the turning point nationally. As discussed above and in the Methods, Chinese cities are defined by administrative boundaries that extend beyond their urban cores, and the 50 cities display very different characteristics and growth rates, offering opportunities for a reasonable sampling of China. We therefore assume China’s national turning point for per capita emissions will be consistent with the interval of per capita GDP when the 50 Chinese cities reach their turning points, that is, US$19,000–22,000 (80% CI). The national per capita emissions would peak in this case at 10.2 tCO2 yr−1 (80% CI: 9.6–10.8 tCO2 yr−1) (Fig. 3), equivalent to the average peak emissions of the socioeconomically developed megalopolises in China. We therefore optimistically project that China’s total emissions should peak at 13–16 GtCO2, that is, 30–60% higher than 2015 values, and should do so around 2021–2025 (that is, China’s 14th Five-Year Plan), based on China’s future population and level of economic development as estimated by the World Bank ( This result reflects that China has begun to implement stringent strategies to improve air quality in cities since 2010, for example, curbing the use of coal and shifting the industrial structure away from high-emission sectors, with an important co-benefit of accelerating CO2 mitigation at the national level.

Fig. 3: The relationship between annual per capita GDP and CO2 emissions for China.

The CIs represent the intervals of per capita GDP corresponding to the per capita emissions of all Chinese cities reaching a peak at 66%, 80% and 90% confidence levels, respectively. The range of China’s per capita GDP corresponding to the emission peak widens from US$19,500–21,500 to US$18,500–22,500 when the CI changes from 66% to 90%. Values for China overall are based on data from the Carbon Dioxide Information Analysis Center (CDIAC)27 counting emissions of fossil-fuel use and cement production process.

Policy implications

We project that China’s total emissions from fossil fuel and industrial processes will peak at 13–16 GtCO2 at some point 5–10 yr ahead of 2030 on the basis of data from the 50 Chinese cities studied here over the period 2000–2016. However, the great diversity in CO2 emissions and trends among individual cities highlight the different challenges they are facing, concerning economic wealth and structure, urban form and geographical location.

As industries are the primary drivers of CO2 emissions in most Chinese cities, sustaining technological improvements of some key industrial sectors22 may be effective in reducing emissions while maintaining economic growth under the current industrial structure and energy systems. At the same time, in contrast to the recent slowdown or even decline in China’s industrial emissions35,36, emissions of transportation and building sectors, which are more closely related to living standards10,19, have grown rapidly and become key factors in determining the emission trends in many Chinese cities (Supplementary Fig. 3). Megalopolises and metropolises are usually sited in the most populated and economically advanced regions of China and their industrial emissions appear to have reached or at least approached peaks, implying the cities must make additional effort to incentivize low-carbon lifestyles and consumption patterns, to improve energy efficiency in buildings and transportation, and to accelerate deployment of clean and renewable energy. For the large and small Chinese cities, which are undergoing rapid industrialization and urbanization and therefore facing direct conflict between economic development and CO2 emissions, efforts to constrain emissions should focus more on industrial sectors.

Policymakers should keep the following points in mind when implementing low-carbon strategies for Chinese cities. First, urban forms, difficult and costly to retrofit once they have been already established (the ‘lock-in’ effect), are among the most critical factors determining the long-term trends of emissions from Chinese cities. Therefore, the emphases of low-carbon strategies for existing cities and newly emerging cities should be different. Many Chinese cities (for example, Beijing, Xi’an, Hangzhou) are relatively mature and policies of such cities should focus more on how to improve energy efficiency (for example, in existing buildings) and how to change the emission trajectories rather than their initial carbon-intensive infrastructure endowments. Emerging cities (for example, Xiong’an) and the new urban areas around old cities, which are currently expanding their infrastructures, however, may have opportunities to leapfrog and bypass carbon-intensive growth. For example, emerging cities should plan and build their infrastructures according to sustainable growth principles37,38, which can help greatly in mitigating emissions over the long run, saving materials and demand for energy by planning for more compact layouts and green buildings, and by reducing demand for traffic fuel consumption through purposeful accessibility planning.

Second, developed cities, for example, Beijing and Shanghai, might be tempted to relocate emissions from intensive industries to ease the transition to a lower-carbon future and to mitigate air pollution. They might directly move out carbon-intensive industries, or indirectly relocate emissions by importing large amounts of carbon-intensive materials and goods from less-developed cities. While this could speed the emission peaks of coastal developed cities, it may delay those of less-developed cities. In other words, such relocation of higher-emitting industries outside of city boundaries might offset the effect of CO2 mitigation on the larger spatial scale29. Policies should discourage such simple carbon leakage, and encourage cities to view their carbon mitigation efforts on broader scales of national and even global levels, to accelerate advanced technology transfer among cities within China and internationally. This could not only promote technology and economic development in western cities of China but also take advantage of their abundant renewable energy (for example, wind power generation, including the reduction of ongoing wind ‘curtailment’)39.

Finally, we emphasize that policies should differentiate cities according to their specific characteristics (for example, urban form, economic development and industrial structure) to realize the most cost-effective low-carbon development for each individual city.


Definition of city and data collection

The definition of a city is broader in China than in the United States and European countries. It identifies a prefectural-level administrative unit encompassing urban, town and rural populations. Correspondingly, China’s statistics, especially energy consumption data, are generally available only for the whole administrative unit (not distinguishing urban and rural areas). We collected basic data such as sectoral energy consumption, industrial production, population and GDP from published statistical yearbooks for each of 50 Chinese cities from 2000 to 2016. The GDP values are adjusted by a PPP conversion factor, defined as the number of local currency units required to buy the same amounts of goods and services in the local market that a US dollar would buy in the United States in 2011. These 50 cities are located in different regions of China with different socioeconomic characteristics (Supplementary Table 1). The substantial differences in terms of affluence (that is, per capita GDP) among Chinese cities also reflect their differing abilities to afford carbon-reduction strategies. We chose 2000–2016 as our study period, since urbanization began to accelerate following China accession to the World Trade Organization in 2001 and as China progressively became the world’s factory (Supplementary Fig. 8).

As cities are defined by their administrative boundaries, and per capita emissions of rural residents are lower than for their counterparts in urban areas of China9,10, our calculation of per capita emissions for Chinese cities in this study should be lower than for their urban residents alone. However, this offers an opportunity for an effective sampling of China in analysing the future emissions by incorporating 50 cities with very different characteristics and growth status.

CO2 emissions inventory of Chinese city

We focused on CO2 emissions from energy consumption and industrial processes (for example, production of cement, lime, steel and glass) in Chinese cities. Emissions from wastes and agriculture were not included in this analysis due to the large associated uncertainties and their negligible contribution to overall emissions of CO2 (ref. 19). Emissions from consumption of various types of energy (for example, coal, coke, natural gas and oil) were estimated for socioeconomic sectors similar to those selected by Shan et al.22, and were further grouped into emissions from industry, transportation and buildings. We calculated emissions using China’s city statistics through International Council of Local Environmental Initiatives40 metrics, similar to those used by nations under the United Nations Framework Convention on Climate Change, that is, the Intergovernmental Panel on Climate Change guidelines for national greenhouse gas inventories. Emissions were calculated by multiplying activity levels (for example, fuel consumption) by corresponding emission factors (for example, emissions per unit of fuel consumption). Besides the emissions occurring within the city’s administrative boundary, that is, scope 1 emissions, the emissions of cities include also contributions relating to their consumption of electricity and heat produced outside of the city boundary (scope 2 emissions). If a city exports electricity/heat to other regions outside its boundary, emissions from generating the exported electricity/heat should be subtracted from the scope 1 emissions to avoid double counting. Unless stated otherwise, emissions of Chinese cities refer to the sum of scope 1 and scope 2 emissions. More details of calculating carbon emissions for Chinese cities are summarized in the Supplementary Methods and in our previous papers17,19.

Short-term pathways of CO2 emissions in China

Previous studies have demonstrated that the relationship between economic development and environmental pollution in general followed the theory of Kuznets curves for developed countries41,42 and China43. Based on this theory, the short-term trends in emissions were estimated by fitting a Gaussian Kuznets curve (bell-shape function) (equation (1)) to the observed emissions and GDP trajectories for each Chinese city through 2000–2016, which assumes that their per capita emissions should increase with per capita GDP to reach the peaks and then decline, as generally observed for developed countries (Supplementary Table 3 and Supplementary Fig. 7).

$$E_{\mathrm{p}} = a \times {\mathrm{e}}^{ - \left( {\frac{{G_{\mathrm{p}} - b}}{c}} \right)^2}$$

where Ep and Gp represent the observed annual per capita CO2 emissions and per capita GDP, respectively; the scale parameter a relates to the height of the function, which corresponds to the measure of peak per capita emissions; b controls the position of the function along the horizontal axis, which is the per capita GDP corresponding to the peak per capita emissions; c governs the shape of the function.

We fit the input data (that is, Ep and Gp) to the abovementioned Gaussian Kuznets curve (equation (1)) for each Chinese city using the function of ‘nlsLM()’ from the package ‘minpack.lm’ in R (version 3.5.3). The ‘nlsLM()’ is developed to modify the original nonlinear least-squares regression with the Levenberg–Marquardt algorithm. As shown in Supplementary Table 2 and Supplementary Fig. 6, per capita emissions are significantly correlated with and follow a bell-shaped relationship with per capita GDP for each Chinese city.

To project China’s national per capita CO2 emissions (see details in the Supplementary Methods), we ran a 1,000-trial Monte Carlo analysis, randomly assigning the turning point within the 66%, 80% and 90% CIs of per capita GDP (that is, US$19,500–21,500, US$19,000–22,000 and US$18,500–22,500) for peak emissions of the 50 Chinese cities.


There is a pressing need for more literature investigating detailed sectoral and geographical breakdowns of energy use and greenhouse gas emissions to decompose the great complexity of how the emissions of Chinese cities are currently evolving. We present a step in that direction, but acknowledge also the following limitations should be addressed in future research.

On the one hand, a more representative sample of Chinese cities, especially the small ones, over a longer time frame should be included in future analyses. Although the 50 cities analysed here can offer a reasonable distribution of Chinese cities in terms of geographical location, urban form, economic development and structure, and were responsible for nearly 35% of China’s total emissions, medium and small cities are underrepresented. This might underestimate the growth of emissions per capita for the country as a whole, while overestimating the decline in emissions per GDP (Supplementary Fig. 2). Future extension of the database in both spatial (that is, including more medium and small Chinese cities) and temporal coverages would facilitate more comprehensive understanding of China’s emissions trends, and would also help to identify the distinct development features and the fundamental mechanisms affecting the peak carbon emissions for individual cities.

In addition, the integrity and accuracy of emissions from Chinese cities need further improvement. Restricted by available emission data for industrial processes, we only include the process emissions of cement, lime, glass and steel production in this study, which might slightly underestimate (less than 0.5%, Supplementary Methods) the total CO2 emissions of Chinese cities on average. Differentiation of energy and economic data by urban and rural districts would support more robust differentiation of conditions within the cities. Quantitative analysis of the impact of direct relocation or indirect relocation (that is, via domestic trade) of carbon-intensive industries on the emission peaks for China and for specific Chinese cities also deserve further study in the future.

Data availability

Details on the methodology and data for estimating CO2 emissions of 50 Chinese cities are summarized in the Supplementary Information, and any other datasets generated during this study are available upon request from the corresponding authors.


  1. 1.

    Watts, M. Cities spearhead climate action. Nat. Clim. Change 7, 537–538 (2017).

    Article  Google Scholar 

  2. 2.

    IPCC Climate Change 2014: Mitigation of Climate Change (eds Edenhofer, O. et al.) (Cambridge Univ. Press, 2014).

  3. 3.

    United Nations Department of Economic and Social Affairs World Urbanization Prospects: The 2014 Revision: Highlights (United Nations, 2014).

  4. 4.

    Rosenzweig, C., Solecki, W., Hammer, S. A. & Mehrotra, S. Cities lead the way in climate-change action. Nature 467, 909–911 (2010).

    CAS  Article  Google Scholar 

  5. 5.

    Duren, R. M. & Miller, C. E. Measuring the carbon emissions of megacities. Nat. Clim. Change 2, 560–562 (2012).

    CAS  Article  Google Scholar 

  6. 6.

    Weiss, K. Cities bask in spotlight at Paris climate talks. Nature (2015).

  7. 7.

    Wang, H., Zhang, Y., Lu, X., Nielsen, C. P. & Bi, J. Understanding China’s carbon dioxide emissions from both production and consumption perspectives. Renew. Sustain. Energy Rev. 52, 189–200 (2015).

    CAS  Article  Google Scholar 

  8. 8.

    National Bureau of Statistics of the People’s Republic of China China Statistical Yearbook 2012 (China Statistics Press, 2013).

  9. 9.

    Qi, Y., Wu, T., He, J. & King, D. A. China’s carbon conundrum. Nat. Geosci. 6, 507–509 (2013).

    CAS  Article  Google Scholar 

  10. 10.

    Wiedenhofer, D. et al. Unequal household carbon footprints in China. Nat. Clim. Change 7, 75–80 (2017).

    CAS  Article  Google Scholar 

  11. 11.

    Baeumler, A. et al. Sustainable Low-carbon City Development in China (World Bank, 2012).

  12. 12.

    Kennedy, C. et al. Greenhouse gas emissions from global cities. Environ. Sci. Technol. 43, 7297–7302 (2009).

    CAS  Article  Google Scholar 

  13. 13.

    Liu, Z. et al. Features, trajectories and driving forces for energy-related GHG emissions from Chinese mega cites: the case of Beijing, Tianjin, Shanghai and Chongqing. Energy 37, 245–254 (2012).

    CAS  Article  Google Scholar 

  14. 14.

    Kennedy, C. A., Ibrahim, N. & Hoornweg, D. Low-carbon infrastructure strategies for cities. Nat. Clim. Change 4, 343–346 (2014).

    CAS  Article  Google Scholar 

  15. 15.

    Zhang, Y. et al. A dual strategy for controlling energy consumption and air pollution in China’s metropolis of Beijing. Energy 81, 294–303 (2015).

    CAS  Article  Google Scholar 

  16. 16.

    Dhakal, S. Urban energy use and carbon emissions from cities in China and policy implications. Energy Policy 37, 4208–4219 (2009).

    Article  Google Scholar 

  17. 17.

    Bi, J. et al. The benchmarks of carbon emissions and policy implications for China’s cities: case of Nanjing. Energy Policy 39, 4785–4794 (2011).

    Article  Google Scholar 

  18. 18.

    Wang, H. et al. Mitigating greenhouse gas emissions from China’s cities: case study of Suzhou. Energy Policy 68, 482–489 (2014).

    CAS  Article  Google Scholar 

  19. 19.

    Wang, H., Zhang, R., Liu, M. & Bi, J. The carbon emissions of Chinese cities. Atmos. Chem. Phys. 12, 7985–8007 (2012).

    Article  Google Scholar 

  20. 20.

    Ramaswami, A. et al. Urban cross-sector actions for carbon mitigation with local health co-benefits in china. Nat. Clim. Change 7, 736–742 (2017).

    Article  Google Scholar 

  21. 21.

    Cai, B., Guo, H., Cao, L., Guan, D. & Bai, H. Local strategies for China’s carbon mitigation: an investigation of Chinese city-level CO2 emissions. J. Clean. Prod. 178, 890–902 (2018).

    Article  Google Scholar 

  22. 22.

    Shan, Y. et al. City-level climate change mitigation in China. Sci. Adv. 4, eaaq0390 (2018).

    Article  Google Scholar 

  23. 23.

    Gurney, K. R. et al. Track urban emissions on a human scale. Nature 525, 179–181 (2015).

    CAS  Article  Google Scholar 

  24. 24.

    Dodman, D. Blaming cities for climate change? An analysis of urban greenhouse gas emissions inventories. Environ. Urban 21, 185–201 (2009).

    Article  Google Scholar 

  25. 25.

    Hillman, T. & Ramaswami, A. Greenhouse gas emission footprints and energy use benchmarks for eight U.S. cities. Environ. Sci. Technol. 44, 1902–1910 (2010).

    CAS  Article  Google Scholar 

  26. 26.

    Sovacool, B. K. & Brown, M. A. Twelve metropolitan carbon footprints: a preliminary comparative global assessment. Energy Policy 38, 4856–4869 (2010).

    Article  Google Scholar 

  27. 27.

    Boden, T. A., Marland, G. & Andres, R. J. Global, Regional, and National Fossil-Fuel CO 2 Emissions (Oak Ridge National Laboratory, US Department of Energy, 2017);

  28. 28.

    Liu, Z. et al. A low-carbon road map for china. Nature 500, 143–145 (2013).

    CAS  Article  Google Scholar 

  29. 29.

    Feng, K. et al. Outsourcing CO2 within china. Proc. Natl Acad. Sci. USA 110, 11654–11659 (2013).

    CAS  Article  Google Scholar 

  30. 30.

    Chen, Y., Ebenstein, A., Greenstone, M. & Li, H. Evidence on the impact of sustained exposure to air pollution on life expectancy from China’s Huai River policy. Proc. Natl Acad. Sci. USA 110, 12936–12941 (2013).

    CAS  Article  Google Scholar 

  31. 31.

    Shen, H. et al. Urbanization-induced population migration has reduced ambient PM2.5 concentrations in China. Sci. Adv. 3, e1700300 (2017).

    Article  Google Scholar 

  32. 32.

    Huo, H., Zhang, Q., Liu, F. & He, K. Climate and environmental effects of electric vehicles versus compressed natural gas vehicles in China: a life-cycle analysis at provincial level. Environ. Sci. Technol. 47, 1711–1718 (2013).

    CAS  Article  Google Scholar 

  33. 33.

    Güneralp, B. et al. Global scenarios of urban density and its impacts on building energy use through 2050. Proc. Natl Acad. Sci. USA 114, 8945–8950 (2017).

    Article  Google Scholar 

  34. 34.

    Ürge-Vorsatz, D. et al. Locking in positive climate responses in cities. Nat. Clim. Change 8, 174–177 (2018).

    Article  Google Scholar 

  35. 35.

    Mi, Z. et al. Chinese CO2 emission flows have reversed since the global financial crisis. Nat. Commun. 8, 1712 (2017).

    Article  Google Scholar 

  36. 36.

    Guan, D. et al. Structural decline in China’s CO2 emissions through transitions in industry and energy systems. Nat. Geosci. 11, 551–555 (2018).

    CAS  Article  Google Scholar 

  37. 37.

    McGranahan, G. & Satterthwaite, D. Urban centers: an assessment of sustainability. Annu. Rev. Environ. Resour. 28, 243–274 (2003).

    Article  Google Scholar 

  38. 38.

    Creutzig, F., Baiocchi, G., Bierkandt, R., Pichler, P. P. & Seto, K. C. Global typology of urban energy use and potentials for an urbanization mitigation wedge. Proc. Natl Acad. Sci. USA 112, 6283–6288 (2015).

    CAS  Article  Google Scholar 

  39. 39.

    Lu, X. et al. Challenges faced by China compared with the US in developing wind power. Nat. Energy 1, 16061 (2016).

    Article  Google Scholar 

  40. 40.

    International Council of Local Environmental Initiatives Local Government Operations Protocol for the Quantification and Reporting of Greenhouse Gas Emissions Inventories (ICLEI, 2010);

  41. 41.

    Grossman, G. M. & Krueger, A. B. Economic growth and the environment. Q. J. Econ. 110, 353–377 (1995).

    Article  Google Scholar 

  42. 42.

    Shahbaz, M. & Sinha, A. Environmental Kuznets curve for CO2 emissions: a literature survey. J. Econ. Stud. 46, 106–168 (2019).

    Article  Google Scholar 

  43. 43.

    Li, T., Yong, W. & Zhao, D. Environmental Kuznets curve in China: new evidence from dynamic panel analysis. Energy Policy 91, 138–147 (2016).

    Article  Google Scholar 

Download references


This study was supported by the National Key R&D Program of China (2016YFA0600204), National Natural Science Foundation of China (NNSFC) (41371528, 71433007, 71690244), IGSNRR and Youth Innovation Promotion Association CAS (2019055) and the Harvard Global Institute of Harvard University.

Author information




H.W. conceived and led the research. H.W., X.L. and J.B. designed the paper. Y.S., Y.D. and H.W. calculated emissions. Y.L., G.Z. and M.B. performed emission trends analysis. H.W., X.L., Y.D. and C.P.N. interpreted the data. H.W., X.L., C.P.N. and M.B.M. drew conclusions and wrote the paper with input from all co-authors.

Corresponding authors

Correspondence to Haikun Wang or Jun Bi.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Supplementary Tables 1–8, Figs. 1–9, methods and references.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Wang, H., Lu, X., Deng, Y. et al. China’s CO2 peak before 2030 implied from characteristics and growth of cities. Nat Sustain 2, 748–754 (2019).

Download citation

Further reading


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing