Decoupling of land-use net carbon flux, economic growth, and population change in China

In the process of China’s modernization, promoting the sustainable development of resource-based cities is a major strategic issue and it has now also become a worldwide issue. This study uses the coupling model to validate the coupling relationship between China’s land-use net carbon flux and economic growth and population change during 2009–2017. The study for the first time draws the conclusion that the coupling degree among the three is getting lower, the correlation is gradually weaker, and the independent relationship is becoming more and more prominent. Utilizing the Tapio decoupling model, we obtained the weak decoupling conclusion that the economic growth rate is higher than the growth rate of the land-use net carbon flux, while negative decoupling of sprawl is where the rate of population growth is less than the rate of net land-use carbon flux growth.


Introduction
Land-use and land-cover change (LULCC) is the central link between human activities and environmental change and the concentrated embodiment of the relationship between man and land (Geng et al. 2000).To date, numerous studies have demonstrated that land-use change and land management can directly or indirectly affect the greenhouse gas exchange and carbon cycle between terrestrial ecosystem and atmosphere (Gurney et al. 2021;Harper et al. 2018;Houghton and Nassikas 2017;Rjoub et al. 2021;Walsh et al. 2017).The carbon emission due to change of land use has become the focus of the world and environmental protection organizations.Based on the United Nations 2020 emission gap report, the fossil carbon dioxide (CO 2 ) emissions (from fossil fuels and carbonates) dominate total GHG emissions including LUC (65 percent) and consequently the growth in GHG emissions.Preliminary data suggest that fossil CO 2 emissions reached a record 38.0 GtCO 2 (range ± 1.9) in 2019.The annual growth rate of global greenhouse gas emissions has been 1.4% since 2010.Owing to the significant increase in vegetation forest fires, the growth rate reached 2.6% in 2019 (Harper et al. 2018;Peylin et al. 2013).The primary reason for the abovementioned issues is that there exist a strong uncertainty and variability of greenhouse gas emissions in the change of land use, and there are more serious consequences of the activities caused by humans.
The early carbon emission literature primarily conducted theoretical analysis and empirical research on the environmental Kuznets curve from the standpoint of economic growth.The predominant view of scholars in this field instead of is that the assumption of a conventional concave environmental Kuznets curve is not robust, and its Responsible Editor: Ilhan Ozturk Xianke Huang, Yujie Huang, and Ruiliang Li contributed equally to this work.
establishment correlates with the selection of measurement methods and research fields (Semieniuk et al. 2021).In addition, scholars in this field extensively researched the correlation between population factors and carbon emissions, primarily from the viewpoint of population size (Miao andChen 2020, Pongratz et al. 2008;Wang et al. 2017).Reportedly, population size significantly affects carbon emissions.To achieve these objectives, the existing environmental IPAT model (Ehrlich and Holdren 1971;Raskin 1995;York et al. 2003;Paramati et al. 2016Paramati et al. , 2017;;Ozcan et al. 2021) has been explored to determine the relationship between CO 2 emissions and 12 variables for the G20 countries.This model discusses the influences of population, GDP, and technology factors on CO 2 emissions, where I is the environmental impact sourced from the total population (P) of the underlying nation, (A) is the economic influence or per capita consumption, and (T) is the level of technology efficiency per capita or per dollar of GDP.This model has been further extended by Paramati et al. (2017) with stochastic impacts by regression on population, affluence, and technology (STIRPAT) model to identify the relationship between renewable energy, stock markets, and CO 2 emissions.With the continuous extension of the research scope of the carbon emission literature, the correlation between economic growth and land-use carbon emission garnered increasing attention (Cheng et al. 2016;Fu et al. 2017;Harper et al. 2018;Houghton et al. 2012;Sun et al. 2006).The research methods primarily followed the research methods of carbon emission and economic growth and carbon emission and population factors.Besides, the research on the correlation between population growth and carbon emission attained fruitful results, which provided a reference for this study (Terrell 2020).Nevertheless, the existing research has the following limitations: (i) it lacks separation of direct carbon emission from land use and examination of its correlation with economic growth and population change; (ii) focus on the impact of economic development and population change on carbon emissions.Indeed, the three are interactive.Coupling can better measure the interaction between economic development, population change, and net carbon flux of land use; (iii) most of these studies are static and lack research on the spatiotemporal coupling dynamic evolution trend of carbon emissions due to economic development and population change.Accordingly, this study explores the temporal and spatial pattern and influencing factors of the coupling coordination degree of population, economy, and carbon emission from the standpoint of coupling.
To explain the correlation between economic growth and resource consumption and the degree and trend of asynchronous change, the Organization for Economic Co-operation and Development (OECD) proposed the concept and basic theory of "decoupling"; that is, different decoupling states are used to explain the different trends of economic development and resource consumption.Generally, if a region decreases resource consumption while attaining economic growth, an ideal decoupling situation exists in this region, which is also the objective of decoupling research.In the 1960s, Carter (Carter 1966) posited the "decoupling" theory to measure the correlation between economic development and resource consumption.Decoupling theory has been extensively used by numerous organizations, including the United Nations Environment Program (Programme 2019).A decoupling model is established to examine the decoupling of environmental pollution, and two dimensions of primary decoupling and secondary decoupling are proposed.Vehmas et al. (Vehmas et al. 2007) and Tapio (Tapio 2005) improved and refined the decoupling model and established a relatively complete decoupling evaluation framework.Per the framework, the decoupling relationship can be divided into decoupling, coupling, and negative decoupling.The authors further calculated the percentage increase or decrease in resources due to economic changes and classified them into eight states (Extended Data Table 1).
Thus, this study is aimed at analyzing the net carbon flux generated by direct and indirect land use, economic development, and population change.Taking the data of China from 2009 to 2017 as an example, through the analysis methods of the coupling model and the Tapio decoupling model, this study discusses the correlation between the net carbon flux of land use and economic growth and population change in various provinces and cities in China and expounds the correlation between the economic and social development, population-scale change, and the net carbon flux of land use in major carbon-emitting countries from the theoretical level.It provides a scientific reference for further carbon emission reduction decision-making and testing.

Economic development
Economic development implies the production and reproduction of the materials of the whole society.Thus, we selected GDP as the reference index of economic development, reflecting the final results of production activities of all permanent residents in a certain period.

Population growth
Population is one of the crucial factors affecting land-use change.
To unify the impact of population size, population mobility, and population size change on land-use types, we selected the number of permanent residents at the end of the year as a reference index, including all permanent residents living in urban and rural areas.With the enhancement of socioeconomic level and the acceleration of population growth, the structure of the types of regional land use will be further changed.

Land-use carbon flux
Direct land-use carbon flux The change in carbon flux due to land-use change includes vegetation carbon cycle and soil carbon cycle (including aboveground and underground).The direct carbon flux of land use includes the change of carbon emission due to land use, carbon storage per hectare of soil, and carbon absorption in the process of regrowth (Houghton and Nassikas 2017, Li et al. 2018, McCalmont et al. 2017, Mendoza-Vega et al. 2021, Montgomery 2007, Pugh et al. 2019, Searchinger et al. 2008, Shi et al. 2012, van Groenigen et al. 2014, Yan et al. 2020, Zhao et al. 2021, Zhu et al. 2019).Extended Data Table 2 shows the carbon emission and carbon absorption coefficient caused by various land-use changes.

Determination of carbon emission coefficient of cultivated land
The carbon emission coefficient of cultivated land should consider not only the carbon emission of cultivated land, that is, the CO 2 greenhouse gas produced by respiration during the growth of crops, but also the carbon absorption of cultivated land, that is, the carbon absorption due to photosynthesis during the production of crops as green plants (Montgomery 2007).Thus, the difference between the two was taken as the carbon emission coefficient of cultivated land.With reference to the previous research results (Carlson et al. 2017;Collins et al. 2019;Pugh et al. 2019;Zhu et al. 2019), the carbon absorption coefficient of cultivated land was 0.0007 kg/(M 2 •a), and the carbon emission coefficient of cultivated land was 0.0504 kg/(M 2 •a).Hence, the difference between the two was taken as the carbon emission coefficient of cultivated land 0.0479 kg/(M 2 •a).

Determination of carbon emission coefficient of forestland
Woodland mostly absorbs CO 2 and other greenhouse gases in the air through photosynthesis and has strong carbon sequestration capacity in the land-use carbon emission.The carbon sequestration capacity of woodland fluctuates with different types.Per related studies (Fang et al. 1996;Hosonuma et al. 2012;Pugh et al. 2019;Zhu et al. 2019), the net carbon fixation of forest land was − 0.0585 kg/(M 2 •a), while that of scrub and young trees was − 0.0577 kg/(M 2 •a).
In this study, the average value of the two was taken as the forest carbon emission coefficient − 0.0581 kg/(M 2 •a).

Determination of carbon emission coefficient of pastureland
Grassland is a critical carbon sink that can attain carbon sequestration mostly through the absorption of CO 2 and other greenhouse gases by different types of grasslands.According to related studies (Cai et al. 2003;Duan et al. 2008;He 2006;Shi et al. 2012), the carbon emission coefficient of grassland was determined as − 0.021 kg/(M 2 •a).

Determination of carbon emission coefficient of water area
Under the action of aquatic organisms, rivers and lakes are usually carbon sinks, but the irrigation and water reservoirs lead to the emission of greenhouse gas CO 2 , which is equivalent to carbon source.According to Lai and Huang (Lai and Huang 2005) and research from other groups on the carbon sink capacity of China's regional waters, the average carbon sink coefficient was 0.0248 kg/(M 2 •a).In this study, we took the lake area into the calculation of carbon emissions; however, before 2011, the lake areas of Sichuan, Gansu, Ningxia, Shanxi, Shaanxi, Guangdong, Guangxi, Henan, and Taiwan were combined in the China Water Conservancy Statistical Yearbook.In this study, we first excluded the data from Taiwan and used the mean interpolation method to supplement the data.

Determination of carbon emission coefficient of unused land
There are multiple types of unused land, mainly including wasteland, saline-alkali land, sandy land, bare land, and rocky land.Such land types also have photosynthesis or respiration of vegetation, which have both carbon sequestration capacity and carbon emission capacity; however, the overall impact on the land-use carbon emission is small, which has been ignored in this study (Gitz and Ciais 2003).
Indirect land-use carbon flux Most energy resources in human life and production are primarily consumed by construction land.Energy consumption is the major source of carbon emission in current land use.Thus, we cannot directly calculate the carbon emission coefficient only by using the data of construction land area (Su and Zhang 2011; Wang et al. 2020;Xu et al. 2006).We can only calculate their consumption in the process of constructing land use through different energy carbon emission coefficients and then indirectly calculate the construction land carbon emission (Hurtt et al. 2020;Kuang et al. 2016;Schwalm et al. 2020;Searchinger et al. 2008;Xu et al. 2006;Zhao et al. 2021).The intensity of carbon emission of other land-use types is significantly lower than that of construction land.

Estimation method of carbon dioxide emission
The carbon emission data are the carbon dioxide emissions from fossil fuel combustion.The carbon emission of different industries was obtained by referring to the abovementioned energy data processing.To ensure the accuracy of carbon emission accounting, we used the data method of different fuel types in the IPCC reference method to calculate the carbon emission of various energy sources (Hurtt et al. 2020;Kuang et al. 2016;Schwalm et al. 2020;Xu et al. 2006).The specific accounting formula is as follows: where E K denotes the terminal consumption of the kth energy; EF k denotes the carbon emission factor of the kth energy; NCV k , A k , and O k denote the average low calorific values, carbon content per unit calorific value, and carbon oxidation rate of the kth energy; 44/12 implies the conversion factor between carbon and CO 2 ; and C j denotes the carbon emission due to energy consumption activities of sector j.The net calorific value and carbon oxidation rate of all types of energy required in the formula are from China's "guidelines for the compilation of provincial greenhouse gas inventories."Extended Data Table 3 provides the CO 2 emissions for each fossil energy source (Xu et al. 2006).

Measurement model of coupling degree
In physics, the concept of coupling is extended to multiple systems, and the coupling degree model is as follows: The model mainly discusses the case of n = 3, where u 1 , u 2 , and u 3 are the comprehensive indexes of the land-use net carbon flux, economic change, and population growth respectively.Therefore, Eq. ( 3) is expanded into Eq.( 4).
Among them, C ∈ the closer the value of C is to 1, the greater the degree of correlation among the three.On the contrary, when C = 0, the smaller the degree of correlation, and the three are in an independent state (Extended Data Table 4).

Decoupling model
The concept of "decoupling" comes from the category of physics, which originally refers to the cessation of mutual relations between two or more physical quantities with response relations.The decoupling development theory plays a significant role in reflecting the uncertain relationship between energy consumption, environmental pollution, and economic development (Peylin et al. 2013;Zhou et al. 2021).This study is aimed at analyzing the relationship between the land-use net carbon flux and economic growth.By constructing a decoupling index system, it can analyze whether economic growth and land-use net carbon flux decouple and the degree of decoupling.This will help in clarifying the mutual relationship between the two.The current general decoupling index construction can be divided into the OECD model and the Tapio model.Compared to the OECD model, the Tapio model has advantages of stability, accuracy, and not being affected by statistical dimensional changes.Therefore, this study adopts Tapio to build the model, as shown below: where DI is the decoupling index; CI is the carbon footprint index of energy consumption, which refers to the land-use net carbon flux in this study; GI refers to the gross domestic product of each region; PG refers to the total population; t 0 and t 1 are the start and end time of the time period.In terms of the time scale of decoupling analysis, due to the lag relationship between economic growth and environmental pollution change, a period of at least 5-10 years should be used.This study employs 2009-2017 as the time interval.
The Tapio decoupling model is essentially an elasticity analysis and is divided into eight decoupling degrees according to the elastic value, as shown in Extended Data (4)

Results
We reexamined the correlation between net carbon flux of land use and economic development and population growth.Unlike the previous conclusion that economic development and population growth are the main driving forces of carbon emission growth (Houghton and Nassikas 2017;Peylin et al. 2013;Zhou et al. 2021), we believe that simple economic development and population growth will no longer be the main factors affecting the net carbon flux of land use and infer that the net carbon flux of land use, economic development, and population growth in most cities in China are at a medium coupling level and gradually decline.The three can play a synergistic effect, but their independence is becoming stronger and stronger.

Analysis results of coupling model
First, this study used the calculation of coupling concept in physics (detailed in the "Methods" section) and extends it to this study.According to the coupling degree index in Table 1, the results of the coupling model between the landuse net carbon flux and economic development and population growth in most cities of China ranged 0.3 < C ≤ 0.5, depicting a medium coupling level, which shows that the three can have a synergistic effect.There were differences in the coupling relationship among different regions from the perspective of spatial distribution, which indicates that the supporting policies and strategic planning of economic and population changes in different regions were different.These can be roughly categorized into eight regions for detailed analysis.The coupling degree of Jiangxi Province, which belongs to the middle reaches of the Yangtze River, was marginally higher than that of other regions; the coupling degree of Southwest China, including Chongqing, Yunnan, Sichuan, Guangxi, and Hunan Province in the middle reaches of the Yangtze River, reached 0.40; the coupling degree of Henan, Hubei, and Anhui provinces in the middle reaches of the Yangtze River, and Fujian, Guangdong, and Hainan provinces in the southern coast, was marginally lower than that of other regions in equal level coupling.In contrast, the coupling model indexes of Ningxia and Xinjiang in Northwest China, Inner Mongolia, and Shanxi in the middle reaches of the Yellow River were 0 < C ≤ 0.3, thereby depicting a low coupling level.Based on the spatial distribution, the economic development level of the middle reaches of Yangtze River, southern coast, and southwest region is significantly higher than that of the northwest region and part of the middle reaches of Yellow River, and the population flow and growth are more, but the land resources and availability are relatively lower.Therefore, it also verifies to some extent that the coupling results are true and reliable.
With reference to time distribution, the coupling degree of China's provinces, cities, and districts during 2009-2017 is in a continuous downward trend.During the observation period, the degree of variation of Xinjiang was the largest, from medium coupling to low coupling, and the coupling degree decreased by 0.06.Jiangsu Province, Fujian Province, Shaanxi Province, Ningxia Hui Autonomous Region, and Jiangxi Province followed closely.Correspondingly, the coupling degree of Yunnan, Guizhou, Hebei, Jilin, and Tianjin changed marginally.Although the coupling degree of most regions showed a downward trend, a medium coupling relationship existed between the land-use net carbon flux and the changes of economy and population.
The concept of "decoupling" stems from the category of physics, which originally denotes the mutual correlation between two or more physical quantities with response relationship.The decoupling development theory plays a significant role in reflecting the uncertain relationship between energy consumption, environmental pollution, and economic development.Thus, "decoupling" in this study implies a departure from the correlation between economic growth and resource consumption or environmental pollution.In addition, we analyzed the correlation between land-use net carbon flux and economic growth.By constructing the decoupling index system, we analyzed whether the economic growth, population growth, and land-use net carbon flux are decoupled and the degree of decoupling to elucidate the correlation between them.The coupling degree in this study implies the change of economic value and population distribution caused by land resource utilization.The use of land resources not only implies the amount of ecological supply, that is, the level of carbon absorption generated by direct land use, but also indicates to undertake the task of greater resource output; that is, cities with intensive industries can use abundant energy to acquire large and rapid economic income, so that it can attract more people to flow in.The high coupling degree in the early stage could be attributed to the fact that China's energy distribution is concentrated, resulting in the concentration of economic resources and population resources.For example, coal resources are concentrated in Beijing, Tianjin, and Hebei provinces in the northern coastal areas, Shanxi, Inner Mongolia, and northwest areas in the middle reaches of the Yellow River; oil resources are primarily concentrated in the northeast, northern coastal, and northwest areas; energy consumption level is relatively low in the southeast coastal areas, but the use of energy is huge in southwestern China, as in Shanghai, Zhejiang, and Guangdong, still relatively low in Guizhou and Guangxi.
The decrease in coupling means that the long-standing correlation between the three gradually shifts to independence.Land resources lead to a yearly increase in carbon emissions from land use, but do not necessarily lead to sustained population growth, such as in Jilin and Heilongjiang Provinces.

Analysis results of decoupling model
Furthermore, we used the Tapio decoupling model to examine the decoupling degree of the land-use net carbon flux from economic growth and population change in China.As shown in Table 2, the decoupling state between the land-use net carbon flux and economic system of China's provinces (cities) during 2009-2017 is primarily weak decoupling, which suggests that the GDP growth rate is higher than that of the land-use net carbon flux.Only Ningxia and Xinjiang have a relatively synchronous state of growth connecting economic growth and net carbon flux of land use.
However, during 2009-2017, the population growth of Heilongjiang Province and Jilin Province was negative, and the decoupling state of the two provinces exhibited strong negative decoupling; that is, the population decreased and the land use increased.In the same period, the population growth was higher than the land-use net carbon flux growth in Beijing.The decoupling index of Tianjin was 1.16, which showed that the population growth and land-use net carbon flux growth were relatively synchronous.However, according to the growth proportion of the population growth and land-use net carbon flux, the later period was very reasonable; it showed negative decoupling of expansion.

Discussion
Although similar research has been carried out using different data and models from different countries, the conclusions tended to imply a long-term positive equilibrium relationship between economic growth and land-use net carbon flux.However, while there may be a positive relationship, it is not a long-term equilibrium (Miao and Chen 2020).
This study offers a new perspective for the study of global carbon dioxide.For the implementation of the Paris Agreement policy and China's speech at the general debate of the 75th UN General Assembly, China announced: "China will enhance its national independent contribution, adopt more powerful policies and measures, strive to reach the peak of carbon dioxide emissions by 2030, and strive to achieve carbon neutrality by 2060."This study reveals that although a coupling relationship exists between economic development and population growth and land-use net carbon flux, this relationship will gradually weaken and become independent.Most previous studies only examined the relationship between economic development and population growth and land-use net carbon flux.Our study is unique from the previously published studies and conducts further comprehensive analysis of the relationship among the three.
We used the coupling model to validate the coupling relationship between China's land-use net carbon flux and economic growth and population change from 2009 to 2017.This study for the first time concludes that the coupling degree among the three is getting lower and lower, the correlation is gradually getting weaker, and the independent relationship is becoming increasingly prominent.Using the Tapio decoupling model, the decoupling degree between the land-use net carbon flux and economic growth and population change of China's provinces, cities, and districts was comprehensively analyzed.We found that the decoupling state between the land-use net carbon flux and economic growth of China's provinces (cities) was primarily weak decoupling, and the GDP growth rate was faster than that of the land-use net carbon flux.However, the decoupling states between the land-use net carbon flux and population change system in China are mostly expansion negative decoupling, which suggests that the growth rate of population is less than that of the landuse net carbon flux.We believe that the reason for this is that China's economic development no longer completely depends on the development and utilization of land resources, and more economic development is owing to the tertiary industry.With the advancement of science and technology, the economic development of most provinces in China is separated from the industry that completely depends on land.Likewise, China has fully implemented the family planning policy since the 1970s, and the population growth rate is relatively slow.Thus, the correlation between population change and land-use net carbon flux is relatively independent in the research period selected in this study.With the increasing aging of the population and the full liberalization of the two-child policy in China, the correlation between population change and land-use carbon emissions merits further research, which is consistent with Shi et al. (Shi et al. 2012), who believed that adjusting industrial structure, technological progress, and adjusting population structure can help alleviate the pressure of carbon emission.
China is a big country with large regional disparities.Thus, regional policies are crucial to China.The majority of the existing regional action mitigation measures on the land-use net carbon flux follow the national plan and do not fully consider the regional characteristics of different regional industrial structures.China's central government should not take unified actions, but should motivate local governments to adopt specific regional policies that match the unique low-carbon consumption potential of the region and carry them out stepwise.
Although China's energy demand will continue to grow, the Chinese government, by incorporating the task of decreasing greenhouse gas emissions into the "14th five year plan" and the long-term goal of 2035, strives to attain such policies as "carbon peak by 2030" and "carbon neutral by 2060."By optimizing and upgrading industries that have high energy consumption and high carbon emissions, we can enhance energy efficiency and realize a low-carbon economy, which can promote the comprehensive and sustainable development of society and promote the green transformation of the world's comprehensive production and lifestyle.
This study summarizes three policy suggestions for China and other developing countries.(1) The spatial difference of coupling coordination degree is apparent, and the spatial pattern is relatively stable.Appropriate measures should be formulated per the actual development of each region.For regions with high dependence on energy and lack of financial, technical, and skilled labor support, resulting in relatively weak decoupling between population and carbon emission, these regions should adjust the mode of economic development; optimize the industrial structure; decrease their dependence on energy; actively introduce funds, technology, and skilled labor; and develop green economy and low-carbon economy.(2) The coupling coordination degree has a significant spatial spillover effect.When formulating policies, local governments should consider neighboring cities, augment regional cooperation and exchanges, and realize complementary advantages.Meanwhile, it can lead the development of other cities by setting up pilot cities for the coupling and coordinated development of population, economy, and carbon emission, as well as provide experience for other cities. (3) Regulating the total population is the key way to change the degree of coupling and coordination.China's dependence on land resources to drive rapid economic growth is mainly a short-term growth brought by the use of fossil energy.Along with optimizing the population structure, improving the quality and management of the population, reducing regional differences, and accelerating the implementation of regional balance strategies, it is also necessary to optimize the industrial structure, reduce the proportion of secondary industries, improve low-carbon technologies, and develop a green and low-carbon economy.China's use of land resources should also be further transformed to seize the opportunity of renewable energy development.

Table 1 .
(Peylin et al. 2013;Zhou et al. 2021ied to resource and environmental economics, strong decoupling indicates the best state, with economic growth occurring as environmental pollution decreases; strong negative decoupling is the worst decoupling state, which indicates that economic growth is negatively correlated with environmental pollution, and environmental pollution intensifies at the same time of economic recession; other states are between the two(Peylin et al. 2013;Zhou et al. 2021).