Accounting greenhouse gas emissions of food consumption between urban and rural residents in China: a whole production perspective

Abstract

Food consumption is necessary for human survival. On a global scale, the greenhouse gas (GHG) emission related to food consumption accounts for 19%–29% of the total GHG emission. China has the largest population in the world, which is experiencing a rapid development. Under the background of urbanization and the adjustment of the diet structure of Chinese residents, it is critical to mitigate the overall GHG emission caused by food consumption. This study aims to employ a single-region input-output (SRIO) model and a multi-regional input-output (MRIO) model to measure GHG emission generated from food consumption in China and compare the contributions of different industrial sectors, uncovering the differences between urban and rural residents and among different provinces (autonomous regions/municipalities), as well as identifying the driving forces of GHG emission from food consumption at a national level. The results indicate that the total GHG emission generated from food consumption in China tripled from 157 Mt CO₂e in 2002 to 452 Mt CO₂e in 2017. The fastest growing GHG emission is from the consumption of other processed food and meat products. Although GHG emissions from both urban and rural residents increased, the gap between them is increasing. Agriculture, processing and manufacture of food, manufacture of chemical and transportation, storage and post services sectors are key sectors inducing food consumption related GHG emissions. From a regional perspective, the top five emission provinces (autonomous regions/municipalities) include Shandong, Hubei, Guangdong, Zhejiang, and Jiangsu. Based on such results, policy recommendations are proposed to mitigate the overall GHG emission from food consumption.

Graphical abstract

Keywords

greenhouse gas (GHG) emission / food consumption / industry sectors / mitigation measures / urban governance

1 Introduction

The rapid urbanization and industrial development have posed many environmental challenges, such as air and water pollution, soil contamination, and more importantly, climate change [1]. To respond to the climate change, many countries have initiated their mitigation efforts by developing low carbon economy, optimizing energy structure, and improving energy efficiency. As the largest GHG emission country in the world, China has prepared a grand plan to reduce its GHG emission. For instance, the central government has proposed to decrease its GHG emissions per unit of GDP by 60%–65% (based on the 2005 level) and peak its GHG emissions in 2030 [2], and aimed to achieve carbon neutrality in 2060 [3,4]. In addition, the central government is actively promoting various mitigation measures as well [5], such as the action plan for peaking carbon emissions by 2030, national carbon emission trading market, and renewable energy programs.

Consumption is the ultimate goal of all production activities and the main driving force for GHG emissions from the production sectors [6]. Many studies have been conducted to estimate the total GHG emission from consumption, especially focusing on household consumption. For instance, Zhang et al. [7] used the IO method to examine the GHG emission and energy use of household consumption in 2000–2010. They found that the indirect GHG emission was the main part, accounting for 69%–77% of the total GHG emission. In these studies, both direct and indirect emissions were accounted to identify the driving forces and predict the future GHG emission changes caused by household consumption. Direct GHG emission from household consumption refers to the direct energy consumption of residents in their daily life, such as the consumption of electricity and natural gas, while the indirect GHG emission refers to the GHG emission embodied in the daily consumption of goods of residents. Wu et al. [8] used an environmentally extended multi-regional input-output (MRIO) model to measure the direct and indirect GHG emissions from household consumption, splitting indirect GHG emission into domestic and foreign ones. They found that the per capita GHG emission from household consumption in China doubled in 2002–2012 and the large-scale interprovincial transfer of indirect GHG emissions occurred either within the Yangtze River Delta, the Pearl River Delta, and the Jing-Jin-Ji Region or among these three regions. Generally, the final household consumption can be classified into eight different consumption categories, i.e., food, clothing, housing, facilities, transportation and communication, education, health, and others, and all the related industrial sectors can be matched into these eight categories [9]. Tian et al. [10] matched 45 national economy industries with 8 categories of household consumption, and employed an input-output (IO) model and a computable general equilibrium (CGE) model to investigate the effects of household consumption pattern on its energy consumption and carbon emissions in Shanghai. They found that the household consumption, energy consumption, carbon emissions, output of the industry, and foreign trade in Shanghai would not stop increasing until 2030. In addition, some scholars paid attention to the differences between urban and rural residents. Wang and Chen [11] combined urban and rural household-level expenditure surveys of residents in one MRIO model to assess the roles that different household consumption categories played in contributing to carbon footprint disparity across different provinces (autonomous regions/municipalities) in China in 2010–2014. They found that the urban-rural disparity expanded by 155% in 2010–2014, while the growth rate decreased from 27.5% to 11.4%. They also found that although the per capita carbon footprints are insignificant in some less-developed regions, large differences exist between urban and rural households in such regions than in other more-developed provinces (autonomous regions/municipalities). However, in the above studies, food consumption is just one category of household consumption. The more specific study only focusing on food consumption is still lacking. Thus, it is critical to initiate such a study so that more specific results can be obtained for preparing appropriate mitigation policies.

2 Literature review

Food consumption is the most basic human demand and an important component of household consumption. It is, therefore, necessary to pay attention on the GHG emission from food consumption since the GHG emission from food consumption contributed to 19%–29% of global anthropogenic GHG [12]. This problem is more prominent in China because China has to meet with the basic food needs of its 1.4 billion people. In particular, rapid economic development and urbanization are leading to significant changes of the food consumption pattern and threaten the stability of the food supply chain. Several studies have been conducted to investigate the impact of food consumption on the environment. Major accounting methods include life cycle assessment (LCA) and input-output analysis (IO). While LCA focuses on the consumer side, IO focuses on the production side. The life cycle of a food system consists of the agriculture stage, the processing stage, the packaging stage, the transportation and preservation stage, the food consumption stage, and the final disposal stage [13]. Wu et al. [14] found, by applying the LCA method, that the total carbon footprint of food consumption is 4.77 Mt in Beijing which accounts for 23.3% of the total emission of household consumption. The most crucial part of GHG emission generated from food consumption comes from grain, fruits, and vegetables, accounting for more than 65% of GHG emission generated from food consumption. Cao et al. [15] divided food into plant-based food and animal-based one and examined the food consumption changes of urban and rural residents in 1997–2016. By using the Lorentz curve and Gini coefficient, they uncovered the spatial-temporal agglomeration characteristics and driving factors. Zhi and Gao [16] analyzed food consumption related GHG emission differences between urban and rural inhabitants in 1978–2006 and found that the GHG emission induced by food consumption has increased due to rapid economic development and changes in food consumption structure. They further proposed mitigation measures, such as improving energy efficiency of relevant sectors and promoting low carbon food consumption. Yue et al. [17] examined the carbon footprints of 26 crops and 6 livestock products to assess the climatic impacts of agriculture from farm production on food consumption in China and found significant differences between carbon footprint across different management patterns and farm scales, which helped them propose appropriate mitigation policies, such as improvement in agricultural management and changes in dietary consumption.

In reality, food industry is highly fragmented and complex. When applying LCA to analyze the GHG emission from food consumption, several challenges exist, such as the low availability of measured data and the incomplete assessment in the complicated food system [13]. Therefore, more scholars began to apply the IO method to analyze the GHG emission caused by food consumption. The IO analysis can illustrate the relationship between the food industry and other relevant industries and provide a more comprehensive understanding [18]. For instance, Kucukvar and Samadi [19] employed a time series multi-regional input-output method to analyze the carbon and energy footprints of food manufacturing industries in the 27 member states of the European Union and Turkey. They found that Germany, France, Spain had the largest food production-related energy footprints. They also found that in all the European countries, except Romania, upstream suppliers are the dominant contributors to their total energy consumption, responsible for over 90% of the total GHG emission caused by food consumption. Feng et al. [20] employed the IO model to explore the food-related GHG emission of urban households in China in 1992–2007 and found that the physical volume of food consumption and the related economic expenditures increased by 20.7% and 35.9% respectively. But due to technological innovation and the change of household demographics, the total GHG emission from food consumption experienced a decreasing trend over the time. Song et al. [21] investigated the energy implications of the food system transformation in China by using a time-series multi-sectoral input-output model and found that food-related energy consumption increased in 2002–2012.

In addition to these findings, investigating the driving forces of GHG emission from food consumption is another important research field. Only by clarifying the causes of the changes can be corresponding measures be taken to reduce the corresponding GHG emission. At present, the commonly used methods are structural decomposition analysis (SDA) and index decomposition analysis (IDA). Su et al. reviewed the relevant studies before 2010 and made a detailed analysis of SDA and IDA [22]. They found that the SDA method is widely used to study such embodied resources flows, involving household consumption, government consumption, fixed asset formation, international trade, and other final demand sectors. The advantage of the SDA method is that it can be used to analyze the internal relationship among final demand, industrial structure, and input factors.

In the field of energy consumption, Wang et al. [23] employed a region-based SDA method to capture the spatial heterogeneity of driving factors between eastern, central, and western China from a temporal perspective. Yu et al. [24] decomposed industrial electricity consumption changes in China in 2007–2012 into four factors: electricity intensity, technology-input structural, final demand structure, and total final demand by using the SDA method. Dietzenbacher et al. [25] applied the SDA method to identify the main contributors and geographical distribution of the growth in global renewable energy use in 2000–2014. Similarly, in the field of GHG emission, Zhang [26] applied the SDA method to examine the direct and indirect urban and rural per-capita carbon emissions generated by household consumption in China from 1987 to 2007. Su and Ang [27] used the SDA method to analyze the relationships between emissions and value added from a different viewpoint. Based on the accounting of indirect carbon emission from household consumption, Zhang et al. [28] analyzed the influencing factors of indirect GHG emission by using the SDA method. In addition, Lin and Xie [18] decomposed the GHG emissions of the food industry in China in 1991–2012 and found that GHG emission changes in the food industry mainly depended on the total output effect and the energy intensity effect.

From a geographical perspective, several relevant studies mainly focus on GHG emission from food consumption in one specific province (autonomous region/municipality). For instance, Yu et al. [29] explored the impact of food consumption on energy consumption and carbon dioxide emission in Anhui Province. Chen [30] calculated the carbon footprint of food consumption of residents in Lanzhou, Gansu Province so that the impact of diet on the environment can be uncovered. Ding [31] investigated carbon emission from food consumption in Guangdong Province. However, a holistic study on the GHG emission caused by food consumption in China is yet to be completed, especially from a regional disparity perspective.

Under such circumstances, this study aims to fill these research gaps by accounting the total GHG emission and sectoral emissions generated from food consumption and uncover such differences between urban and rural residents in 2002–2017. Simultaneously, regional disparities of the total and the per capita GHG emissions from food consumption in different provinces (autonomous regions/municipalities) in China will be analyzed based on the MRIO data for years 2010 and 2012. The academic contributions of this study are multiple. This study investigates the GHG emission from food consumption in China from a holistic perspective so that both national and regional perspectives can be presented, and urban and rural emission differences can be uncovered to present the special features of rapid development in China. Moreover, the key emission sectors will be investigated from sectoral perspectives and the SDA method will be applied to decompose the GHG emission from food consumption at a national level. It is expected that these contributions can help prepare more feasible mitigation policies so that the overall GHG emission from food consumption in China can be mitigated. It is also expected that such findings can provide valuable insights for other countries facing similar challenges.

3 Methods and data

3.1 GHG emission coefficients of different sectors

The GHG coefficient refers to the amount of GHG emitted per unit of economic output. To explore the relationship between industrial sectors and food consumption, the first step is to calculate the GHG coefficients of various industrial sectors in China. The basic equation of the GHG emission coefficient of primary energy is

(1)C_i={\mathrm{NCV}}_i{\mathrm{*CC}}_i{\mathrm{*OF}}_i\mathrm{*}{\displaystyle \frac{\mathrm{44}}{\mathrm{12}}},

where NCV means the average low calorific value, CC means the carbon content per unit calorific value, OF means the carbon oxidation rate, and i means the type of primary energy. The calculation formula of GHG emission coefficient of electric and thermal energy are

(2)C_{\mathrm{e}}={\displaystyle \frac{{{\displaystyle \sum }}_{i=\mathrm{1}}^{\mathrm{3}}{{\displaystyle \sum }}_{j=\mathrm{1}}^{\mathrm{27}}{q}_j{C}_j}{Q}},

(3)C_{\mathrm{h}}={\displaystyle \frac{{{\displaystyle \sum }}_{i=\mathrm{1}}^{\mathrm{27}}h{C}_i}{H}},

where q is the quantity of the primary energy consumed by different types of power generation, Q is the total gross generation, h is the quantity of the primary energy consumed by the thermal sector, and H is the total heat production. Based on the above data, the GHG emission coefficients of different sectors can be calculated as

(4)f_j={\sum }^{\mathrm{}}{\displaystyle \frac{E{C}_j}{P_j}},

(5)E{C}_j=\left({{\displaystyle \sum }}_{i=\mathrm{1}}^{\mathrm{27}}{C}_i{E}_j\right)+C_{\mathrm{e}}{E}_{\mathrm{e}j}+C_{\mathrm{h}}{E}_{\mathrm{h}j},

where E_j means the quantity of the energy consumed by the j sector, E_ej and E_hj are the amount of the electricity and heat consumed by the j sector, P_j means the total output of the j sector, and f_j means the GHG emission coefficient of the j sector.

To conduct a more comprehensive analysis, 26 types of energy fuels are considered in this study, as summarized in Table 1, including row coal, cleaned coal, other washed coal, coke, coke oven gas, blast furnace gas, converter gas, other gas, other coking products, crude oil, gasoline, kerosene, diesel oil, fuel oil, naphtha, lubricants, paraffin waxes, white spirit, bitumen asphalt, petroleum, LPG, refinery gas, other petroleum products, natural gas, LNG, and other energy sources reported in the statistical yearbooks. The lower calorific values, carbon content, and oxidation rates for all these types of energy sources were derived from China Energy Statistical Yearbooks, the General Principle of Comprehensive Energy Consumption Calculation (GB/T2589-2008) [32], the Guideline for Provincial Greenhouse Gas Inventory [33] and the Intergovernmental Panel on Climate Change (IPCC) Guidelines [34].

3.2 GHG emission accounting

The IO method can be classified into the single-region input-output (SRIO) method and the multi-region input-output (MRIO) method [35]. SRIO is usually applied to examine the GHG emission of a single region by aggregating the rest of the world as one region [36]. Thus, for the national level, the SRIO method was used to estimate the GHG emission caused by the food consumption. The basic IO equation is [37]

(6)\mathit{X}={\left(\mathit{I}-\mathit{A}\right)}^{-\mathrm{1}}\mathit{Y},

where I is the identity matrix, and A represents the direct consumption coefficient matrix. (I–A)^–1 stands for the Leontief inverse matrix, reflecting the relationship between the final demand and the total output and uncovering the complicated economic relations among various sectors of the national economy. Y refers to the final consumption matrix and X is the economic output matrix. Coupled with the GHG emission coefficient matrix F, the total GHG emission can be calculated as

(7)\mathit{C}\mathit{E}=\mathit{F}\mathit{X}=\mathit{F}{\left(\mathit{I}-\mathit{A}\right)}^{-\mathrm{1}}\mathit{Y},

where CE is the total GHG emission matrix and F represents the GHG emission coefficient matrix. It is worth noting that the SRIO model employed in this study is competitive, which means in the basic equation, the matrix A does not distinguish between domestically produced products and imported products [38]. Generally, there are two ways to solve this problem. One is to treat the imported products as different from the domestic ones and to use the origin of the imports to calculate the corresponding emissions. The other is to treat the imported products to be the same as those produced domestically [39]. Since it is difficult to obtain the raw data of imported products, the competitive import hypothesis is chosen. Although there is a gap between the calculation results of the two different assumptions [39], it is believed that it does not affect the overall trend. Thus, based on the assumption that the intermediate products and the final products use imported goods in the same proportion, Eq. (8) is modified. The new direct consumption coefficient matrix and final demand vectors were derived in which only domestic goods, A_d and Y_d, are included [40], expressed as

(8)S_i={\displaystyle \frac{m_i}{x_i+m_i-e_i}},\mathrm{\hspace{0.17em}}\mathrm{for}\mathrm{\hspace{0.17em}}\mathrm{all}\mathrm{\hspace{0.17em}}i,

(9){\mathit{A}}_{\mathrm{d}}=\mathrm{d}\mathrm{i}\mathrm{a}\mathrm{g}\left(\mathrm{1}-S\right)\mathit{A},

(10){\mathit{Y}}_{\mathrm{d}}=\mathrm{d}\mathrm{i}\mathrm{a}\mathrm{g}\left(\mathrm{1}-S\right)\mathit{Y},

where S_i is the share of imports in the supply of products and services to each sector i, x represents the total output, m is the imported goods, and e refers to the exports. Then, for each year, the GHG emission embodied in import is calculated by Eq. (9), assuming that the imported goods used in the food supply chain were produced in China.

(11)\mathit{C}{\mathit{E}}_i=\mathit{F}{\left(\mathit{I}-\mathit{A}\right)}^{-\mathrm{1}}{\mathit{Y}}_i,

where CE_i refers to the total domestic GHG emission embodied in the total import Y_i.

China is a large country with imbalanced regional development. Each province (autonomous region/municipality) has its own resource endowment, industrial structure, and consumption culture. MRIO can trace spatial transfer of resources use, energy consumption, related environmental emissions and help distinguish the imported goods and services with domestic ones, which is usually used to present the interprovincial emission flows [8,35,41–43]. To uncover regional disparity, the MRIO method is employed to investigate the GHG emission from food consumption. The basic equation of MRIO is the same as that of SRIO and the direct consumption coefficient matrix A of MRIO is

(12){\mathit{A}}_{\mathit{r},\mathit{s}}^{\mathit{i},\mathit{j}}=\left(\begin{array}{lll}{a}_{\mathrm{1},\mathrm{1}}^{i,j}\hfill & \cdots \hfill & a_{\mathrm{1},n}^{i,j}\hfill \\ ⋮\hfill & \ddots \hfill & ⋮\hfill \\ a_{\mathrm{1},n}^{i,j}\hfill & \cdots \hfill & a_{n,n}^{i,j}\hfill \end{array}\right),

where the element a_{r,s}^{i,j}={\displaystyle \frac{z_{r,s}^{i,j}}{x_s^j}} means the direct consumption coefficient of sector s in region j from sector r in region i. In terms of the final consumption Y_{\mathrm{F}}=\left(\begin{array}{lll}{Y}_{\mathrm{F}}^{\mathrm{1},\mathrm{1}}\hfill & \cdots \hfill & Y_{\mathrm{F}}^{\mathrm{1},m}\hfill \\ ⋮\hfill & \ddots \hfill & ⋮\hfill \\ Y_{\mathrm{F}}^{n,\mathrm{1}}\hfill & \cdots \hfill & Y_{\mathrm{F}}^{n,n}\hfill \end{array}\right), each element Y^i,j refers to the final products from province (autonomous region/municipality) i to province (autonomous region/municipality) j for meeting the consumption demand. Then, the inter-provincial GHG emission transfer matrix CE_p can be calculated by replacing the GHG emission coefficient matrix F in Eq. (2) with F_p which is equal to \left(\begin{array}{lll}{F}^{\mathrm{1}}\hfill & \cdots \hfill & \mathrm{0}\hfill \\ ⋮\hfill & \ddots \hfill & ⋮\hfill \\ \mathrm{0}\hfill & \cdots \hfill & F^n\hfill \end{array}\right)and each element Fⁱ is the row vector of the sectoral direct emission coefficient in province (autonomous region/municipality) i. The calculation details can be referred to in Ref. [35].

3.3 SDA method

The core idea of the SDA method is to decompose the change of a dependent variable in the economic system into the sum of the changes of related independent variables, so as to measure the contribution of each variable to the change of dependent variable. Since Leontief proposed the IO model, structural decomposition has been applied and its theory has been gradually improved. In the SDA method, the treatment of the cross phase of each variable after decomposition directly affects the analysis effect, the current popular method is to merge the cross term into an independent variable [44]. In this study, the mainstream two polar decomposition method is used. Taking the case of five factors as an example, if the dependent variable is x and the decomposition factors are a, b, c, d, and e, then

(13)x=a\cdot b\cdot c\cdot d\cdot e,

the decomposition equation based on the base year is

(14)\mathrm{\Delta }x=\mathrm{\Delta }a\cdot b_t\cdot c_t\cdot d_t\cdot e_t+a_{\mathrm{0}}\cdot \mathrm{\Delta }b\cdot c_t\cdot d_t\cdot e_t+a_{\mathrm{0}}\cdot b_{\mathrm{0}}\cdot \mathrm{\Delta }c\cdot d_t\cdot e_t+a_{\mathrm{0}}\cdot b_{\mathrm{0}}\cdot c_{\mathrm{0}}\cdot \mathrm{\Delta }d\cdot e_t+a_{\mathrm{0}}\cdot b_{\mathrm{0}}\cdot c_{\mathrm{0}}\cdot d_{\mathrm{0}}\cdot \mathrm{\Delta }e,

and the decomposition equation based on the current year is

(15)\mathrm{\Delta }x=\mathrm{\Delta }a\cdot b_{\mathrm{0}}\cdot c_{\mathrm{0}}\cdot d_{\mathrm{0}}\cdot e_{\mathrm{0}}+a_t\cdot \mathrm{\Delta }b\cdot c_{\mathrm{0}}\cdot d_{\mathrm{0}}\cdot e_{\mathrm{0}}+a_t\cdot b_t\cdot \mathrm{\Delta }c\cdot d_{\mathrm{0}}\cdot e_{\mathrm{0}}+a_t\cdot b_t\cdot c_t\cdot \mathrm{\Delta }d\cdot e_{\mathrm{0}}+a_t\cdot b_t\cdot c_t\cdot d_t\cdot \mathrm{\Delta }e,

then take the arithmetic mean of the two decomposition equations to get the result

\mathrm{\Delta }x=A+B+C+D+\mathrm{E}

={\displaystyle \frac{\mathrm{1}}{\mathrm{2}}}\left(\mathrm{\Delta }a\cdot b_t\cdot c_t\cdot d_t\cdot e_t+\mathrm{\Delta }a\cdot b_{\mathrm{0}}\cdot c_{\mathrm{0}}\cdot d_{\mathrm{0}}\cdot e_{\mathrm{0}}\right)

+{\displaystyle \frac{\mathrm{1}}{\mathrm{2}}}\left(a_{\mathrm{0}}\cdot \mathrm{\Delta }b\cdot c_t\cdot d_t\cdot e_t+a_t\cdot \mathrm{\Delta }b\cdot c_{\mathrm{0}}\cdot d_{\mathrm{0}}\cdot e_{\mathrm{0}}\right)

+{\displaystyle \frac{\mathrm{1}}{\mathrm{2}}}\left(a_{\mathrm{0}}\cdot b_{\mathrm{0}}\cdot \mathrm{\Delta }c\cdot d_t\cdot e_t+a_t\cdot b_t\cdot \mathrm{\Delta }c\cdot d_{\mathrm{0}}\cdot e_{\mathrm{0}}\right)

+{\displaystyle \frac{\mathrm{1}}{\mathrm{2}}}\left(a_{\mathrm{0}}\cdot b_{\mathrm{0}}\cdot c_{\mathrm{0}}\cdot \mathrm{\Delta }d\cdot e_t+a_t\cdot b_t\cdot c_t\cdot \mathrm{\Delta }d\cdot e_{\mathrm{0}}\right)

(16)+{\displaystyle \frac{\mathrm{1}}{\mathrm{2}}}\left(a_{\mathrm{0}}\cdot b_{\mathrm{0}}\cdot c_{\mathrm{0}}\cdot d_{\mathrm{0}}\cdot \mathrm{\Delta }e+a_t\cdot b_t\cdot c_t\cdot d_t\cdot \mathrm{\Delta }e\right),

where A, B, C, D, and E respectively represent the influence of five factors on the dependent variable x.

In this study, the influencing factors of GHG emission from food consumption at a national level are divided into technological progress t, sector linkages, consumption structure c, consumption level e, and population scale p. Therefore, the equation of GHG emission from food consumption is

(17)\mathit{C}\mathit{F}=\mathit{F}\times {\left(\mathit{I}-{\mathit{A}}_d\right)}^{-\mathrm{1}}\times Y_{\mathrm{d}}=\mathit{t}\times s\times c\times e\times p,

(18)\mathrm{\Delta }\mathit{C}\mathit{E}=\mathrm{\Delta }\mathit{t}\times \mathrm{\Delta }s\times \mathrm{\Delta }c\times \mathrm{\Delta }e\times \mathrm{\Delta }p,

where the technological progress factor represents the impact of research and development and application of low-carbon and energy saving technologies, which is reflected by GHG emission coefficient \mathit{t}=\mathit{F}; sector linkages represent the influence of the degree of connection among economic sectors, which is reflected by Leontief inverse matrix \mathit{s}={\left(\mathit{I}-{\mathit{A}}_{\mathrm{d}}\right)}^{-\mathrm{1}}; the effect of consumption structure represents the influence of the distribution of the consumption of residents, which is reflected by the ratio between the residential consumption in each sector and their total consumption \mathit{c}={\displaystyle \frac{Y_{\mathrm{d}}}{Y_{\mathrm{total}}}}; the consumption level represents the impact of final consumption growth, which is reflected by per capita consumption expenditure e={\displaystyle \frac{Y_{\mathrm{total}}}{P}}; the population scale represents the impact of population growth, which is reflected by the total population at the end of the year p = P.

3.4 Data sources

The China Input-Output (SRIO) tables for the years 2002, 2007, 2012, and 2017 were obtained from the National Bureau of Statistics [45]. The data of the final energy consumption by the industry sector and the energy balance sheet at a national level were obtained from China Energy Statistical Yearbooks [46]. The sectoral data of GHG emission in each province (autonomous region/municipality) in China for the year 2012 were collected from China’s Emission Accounts and Data sets and the rest of the data are referred to in Ref. [47]. The provincial MRIO table for the year 2012 was obtained from the Institute of Geographic Science and Natural Resources Research [48]. In total, 30 provinces (autonomous regions/municipalities) in China are considered in this study, except Hong Kong SAR, Macao SAR, Taiwan Province, and Tibet Autonomous Region due to the lack of relevant data in these regions.

Furthermore, due to the different sectoral accounting items in the IO table and energy statistical yearbooks, the different databases were aggregated into 43 sectors based on the Industrial Classification for National Economic Activities (GB/T 4754-2011) [49]. The merged industrial sectors and the abbreviated names are presented in Table 2.

4 Results

4.1 Results at a national level

GHG emissions from food consumption in China for years 2002, 2007, 2012, and 2017 are summarized in Fig. 1, while such emission differences between urban and rural residents are listed in Fig. 2. In general, GHG emissions from food consumption in China experienced a growth, tripling from 157 Mt CO₂e (carbon dioxide equivalent) in 2002 to 452 Mt CO₂e in 2017. Of the GHG emissions caused from the consumption of different food types, the fastest growing one is from the consumption of other processed food and processed meat, which increase from 65.6 Mt CO₂e and 14.7 Mt CO₂e in 2002 to 199.8 Mt CO₂e and 72.1 Mt CO₂e in 2017 respectively. The GHG emission from the consumption of grinded cereal has also been on the rise, being 16.6 Mt CO₂e in 2002, which was greater than that of processed meat. However, such a figure was only about half of that of processed meat in 2017. Figure 3 shows the GHG emission differences from imported food consumption between urban and rural residents. It is clearly observed that the GHG emission caused by urban residents increases faster than that by rural residents. In 2002, the GHG emission from rural residents was 52% of that from urban residents, but this figure was reduced to 36% of that from urban residents in 2017. The main reason for this is the rapid urbanization. Compared with 2002, urban population in China increased by 62% and reached 813.47 million in 2017, while the rural population decreased from 782.41 million in 2002 to 576.61 million in 2017. Although GHG emission from food consumption of rural residents increased in the study period, such an emission decreased by 4% in 2007 than that in 2002 due to the sharp decline of carbon emission intensity in 2007.

4.2 Results at a sectoral level

Figure 4 illustrates sectoral GHG emissions caused by food consumption in 2017. The food system is composed by agricultural production, food processing and manufacturing, packaging, transportation, sales, and final consumption. Table 2 lists all the relevant food consumption sectors. It is clearly observed that sectoral GHG emissions caused by food consumption increases stably. The top four sectors with the highest GHG emissions include agriculture (AGR), processing of food from agriculture and products and manufacture of food (FAMF), manufacture of chemical raw materials and chemical products (MCC), and transport, storage, and postal services (TSPS), with an amount of 107.5 Mt CO₂e, 80.5 Mt CO₂e, 57.4 Mt CO₂e, and 41.7 Mt CO₂e, respectively.

4.3 Results at a provincial level

Figure 5 exhibits the GHG emissions from food consumption at a provincial level in 2012. The top five provinces (autonomous regions/municipalities) with the largest GHG emissions from food consumption include Shandong Province (34.8 Mt CO₂e), Hubei Province (21.4 Mt CO₂e), Guangdong Province (21.3 Mt CO₂e), Zhejiang Province (17.4 Mt CO₂e), and Jiangsu Province (15.1 Mt CO₂e), while the top five provinces (autonomous regions/municipalities) with the lowest GHG emissions from food consumption include Qinghai Province (0.8 Mt CO₂e) and Hainan Province (1.5 Mt CO₂e), Ningxia Hui Autonomous Region (1.7 Mt CO₂e), and Guizhou Province (2.9 Mt CO₂e) and Gansu Province (3.7 Mt CO₂e). Figure 6 depicts per capita GHG emissions at a provincial level. From 2010 to 2012, per capita GHG emissions from food consumption increased in almost all the provinces (autonomous regions/municipalities), except Shandong Province and Jiangxi Province. Shandong Province, Tianjin and Shanghai Municipalities, Liaoning Province, and Xinjiang Uyghur Autonomous Region had the largest per capita GHG emissions from food consumption in 2010, with an amount of 0.523 t CO₂e, 0.294 t CO₂e, 0.263 t CO₂e, 0.222 t CO₂e, and 0.209 t CO₂e, respectively. Inner Mongolia Autonomous Region had the largest per capita GHG emission from food consumption in 2012, with an amount of 0.372 t CO₂e, followed by Hubei Province (0.369 t CO₂e) and Shandong Province (0.359 t CO₂e), Shanghai (0.355 t CO₂e) and Beijing Municipalities (0.331 t CO₂e). Figure 7 illustrates the comparisons of GHG emissions from food production and food consumption of different provinces (autonomous regions/municipalities) in China. It is clearly observed that Shandong Province had the highest GHG emission from food production, with an amount of 35.76 Mt CO₂e, followed by Henan Province (19.26 Mt CO₂e), Hubei Province (18.83 Mt CO₂e), and Jilin Province (17.61 Mt CO₂e), and Inner Mongolia Autonomous Region (16.91 Mt CO₂e). In particular, several provinces (autonomous regions/municipalities), such as Inner Mongolia Autonomous Region and Jilin Province, had much higher GHG emissions from food production than those from food consumption. However, Beijing and Shanghai Municipalities had higher GHG emissions from food consumption that those from food production. In general, those eastern coastal provinces (autonomous regions/municipalities) were the main food consumption regions, while those north-east and north-west provinces (autonomous regions/municipalities) were major food production regions.

4.4 Decomposition of driving forces

The decomposition analysis of five factors indicates that only technological progress has a negative effect on the GHG emission from food consumption in 2002–2007, leading to a 98.6% of GHG emission reduction. Other factors all contributes to the GHG emission increase. The consumption level has the highest impact, leading to a 69.5% of GHG emission increase. With the overall contribution of these five factors, the GHG emission caused by food consumption increased by 36.1% in 2002–2007. Sector linkages became one negative factor and led to a 6.5% of GHG emission reduction from food consumption in 2007–2012, while technological progress led to a 56.6% of GHG emission reduction in the same period. The consumption level was the largest positive factor and led to 87.5% of GHG emission increase in 2007–2012. Similarly, both technological progress and sector linkages were negative factors, while the other three factors were positive and led to GHG emission increases. Table 3 lists all the related data.

4.5 Uncertainty analysis

Based upon the IO analysis, the GHG emission intensities of various sectors were accounted by using the data from energy statistical yearbooks and those collected by CEADs. The accuracy and reliability of these data may have different impacts on the research results. Therefore, a sensitivity analysis was conducted which indicates that the agricultural sector, the manufacture of food from agricultural products and food manufacture sector, the manufacture of chemical raw materials and chemical products sector, and the transport, storage, and postal services sector were the four dominant sectors for the total GHG emission from food consumption. Taking the GHG emission intensities of these four sectors as the variables, each individual parameter was designed to change by –10%, –5%, 5%, and 10%, with the rest being constant [50]. The changes of the total GHG emission from 2002 to 2017 are listed in Table 4. It is clearly observed that the change of GHG emission intensity does not lead to the significant change of the total GHG emission caused by food consumption. Consequently, this uncertainty analysis can guarantee the accuracy of the research results.

5 Discussion

5.1 Trends of the GHG emission from food consumption at a national level

By using the data of food consumption and GHG emission intensities of various kinds of food, Cao and his colleagues [15] found that the total GHG emission in China from food consumption reached 468 Mt in 2016. The result obtained in the present paper indicates that such an emission reached 452 Mt in 2017, which is different from the result obtained by Cao and his colleagues due to different accounting methods and data. In addition, the GHG emission caused by food consumption experienced an increasing trend both in the total amount and in the consumption of different types of food. The main reason for this lies in the continuous improvement of the income and living standards of the residents. Overall, both GHG emissions from the consumption of imported food and the consumption of domestic food experienced increasing trends for both urban and rural residents, reflecting increasing demands of the Chinese people on various types of food. However, the gap between urban and rural residents has been increased, mainly affected by the increasing income gap between urban and rural residents [51]. Of the different types of food consumptions, the change of GHG emission caused by other processed food consumption is the most obvious. Such other processed food types include vegetables, fruits, dairy products, and other types of food, which are more carbon intensive. In particular, the improvement of living standards induced more demands on such food. The GHG emission from processed meat and processed fishery food increased significantly due to the adjustment of diet structure of the residents.

5.2 Effects of food consumption on GHG emissions from different sectors

As the top emission sector, the agricultural sector is the upstream sector for food system and relies on a large amount of fossil fuels and different raw materials to support its production, leading to higher CO₂ (such as the combustion of fossil fuels), methane (such as cattle husbandry), and NO₂ emissions (such as nitrogen fertilizer). Different from the large farming and highly mechanized modern agricultural production mode in western countries, due to the large size of China’s population and land use, agricultural production in many provinces (autonomous regions/municipalities) in China is still traditional. Compared with intensive modern agriculture, this traditional agricultural production mode leads to more energy consumption and GHG emissions. The processing of food from agriculture and food product and manufacture sector has the second highest GHG emission because this sector consumes a large amount of fossil fuel to support its production. Although with the active promotion of energy saving and emission reduction technologies, and the overall industrial GHG emission intensity experienced a decreasing trend [52], the total GHG emission is still increasing due to its large economy scale. In this regard, the manufacture of chemical raw material and chemical product sector generated a higher GHG emission because the production of fertilizers is energy intensive and China consumed the most chemical fertilizers in the world. The annual use of chemical fertilizers in China accounts for 35% of the total in the world, which is the sum of the that in the United States and India, but the effective utilization rate is only 35%, which is 10%–20% lower than those in developed countries [53]. Although the central government issued an action plan for zero growth of fertilizer use by 2020 in 2015 [54], the continuously increasing food demand and the ineffective enforcement of related environmental protection regulations led to a more fertilizer consumption and a higher corresponding GHG emission from the manufacture of chemical raw materials and chemical products sector. The transport, storage, and postal services sector is also energy intensive. With the improvement of income and living standards of the residents in China, the demand for different types of vegetables, fruits and dairy products and the requirement of preservation are soaring, leading to a more energy consumption in this sector. Therefore, this sector experienced the most significant GHG emission growth in all the sectors.

5.3 GHG emissions from food consumption at a provincial level and inter-provincial transfer

With the implementation of the reform and opening-up policy and the construction of the Yangtze River Economic Delta, the eastern coastal provinces (autonomous regions/municipalities) and a few central provinces (autonomous regions/municipalities) significantly improved their economies, leading to a higher income and living standard of residents than other provinces (autonomous regions/municipalities) in China. Meanwhile, due to resource endowments, imbalanced economic development, and geographical positions, the eastern provinces (autonomous regions/municipalities) in China have a much higher population density, followed by the central regions, while few residents live in the west of China. Such regional population disparity led to significant GHG emission differences caused by food consumption between different provinces (autonomous regions/municipalities). In terms of the differences between urban and rural residents, the GHG emission from food consumption of urban residents is higher than that of rural residents in all the studied provinces (autonomous regions/municipalities) in China. Especially, due to the high urbanization rate in Shanghai, Beijing, and Tianjin Municipalities, the gaps between urban and rural residents are the largest in these cities. Due to significant differences of natural resource endowments, food culture and economic conditions [8], domestic food trade activities occurred between different provinces (autonomous regions/municipalities), leading to a large amount of interprovincial transfer of such GHG emissions.

5.4 Research limitations

Several research limitations exist and should be highlighted. First, due to the IO method and the difficulty in obtaining accurate data, including food sources, energy intensity and energy mix of food production in that export country, this study used import assumptions for GHG emission accounting [39], no further decomposition analysis on GHG emissions from imported food trade is conducted. Besides, different sectoral aggregations will lead to different results [55]. Moreover, the international feedback effect was not considered when using the MRIO table to calculate the GHG emissions at the provincial level [56]. Furthermore, energy consumption from cooking and refrigerators of residents is not considered due to the lack of such data. The citizens in China use various energy sources for their cooking, including liquified natural gas, coal, wood, and electricity. It is extremely difficult to get such data due to the large size of the country and the relatively small figures compared with other energy consumption. However, it is believed that these limitations do not significantly influence the research results.

6 Policy implications

The per capita GHG emission in China caused by food consumption increased from 122.7 kg CO₂e in 2002 to 325.4 kg CO₂e in 2017. Although it is still much lower than those in the developed countries, such as 1310 kg CO₂e·cap^–1 in the United States [57] and 1790 kg CO₂e·cap^–1 in Denmark [58], due to the large population, the rapid economic growth, the continuous adjustment of food structure of the residents, and the industrial transformation of the food industry, it is urgent to identify effective mitigation pathways to reduce the total GHG emission from food consumption. By considering the realities in China, especially regional disparity, several policy recommendations are proposed.

First, the central government should implement appropriate policies in support of low carbon food. Such food contains less carbohydrates and can be made with more advanced low carbon technologies [59]. Especially, it is encouraged to feature such industries with low emissions and less power consumption in its entire life cycles, including its production, transportation and consumption stages. Therefore, it is critical for the central government to establish national low carbon food standards, labels, and certify low carbon food enterprises. Also, low carbon food markets should be supported by both the central government and local governments so that low carbon food enterprises can improve their competitiveness. Besides, research and development activities should be supported so that adequate funds can be used for developing more low carbon food types and more energy efficient food production technologies. In addition, several economic instruments can be applied to promote low carbon food. For instance, carbon emission tax can greatly incorporate environmental impacts of meat products into their prices so that such impacts can be internalized. Similarly, financial subsidies can be used to support those low carbon food enterprises since their production costs may be higher than other food producers. Further, new regulations should be released to guide both food producers and consumers, such as the prohibition of once-off plastics food utensils, the avoidance of the overuse of food packaging materials, and more serious punishment on wasting food [60]. Of course, considering significant regional disparities, regional cooperation should be encouraged so that more developed eastern provinces (autonomous regions/municipalities) can transfer their advanced technologies and management experiences to their western counterparts.

Moreover, at an enterprise level, more efforts should be made by all the relevant food companies to actively seek to apply more energy efficient technologies and equipment so that their overall GHG emissions can be reduced. In this regard, energy audit is crucial so that potential mitigation opportunities can be identified. Also, energy structure optimization is effective. This means that all food enterprises should consider the potential application of renewable and clean energy, such as solar power, wind power, geothermal power, biogas, and hydro-power, by considering the local realities. Especially, since many food companies may need heat for their operation, it is encouraged for these companies to seek potential synergies from their neighboring firms so that energy cascading opportunities can be found. In addition, since agricultural sector has the highest GHG emission of all the relevant sectors, it is crucial to mitigate its corresponding emission. The agricultural sector in China consumes a great number of fertilizers. Such an amount has tripled from 1980 to 2010 and contributed to 7% of national GHG emission in 2010 [61]. Thus, the agricultural sector should avoid the overuse of fertilizer and increase the use of organic fertilizer. Modern agricultural activities, such as water-saving irrigation (drip or spray irrigation) and hydroponics, should be widely promoted [62]. Another prominent problem in the food industry is food packaging. To meet with various demands from different consumers, excessive packaging materials have been consumed. Some packing materials are in high quality, but are used for only once, leading to more packaging wastes. Consequently, all the food producers should avoid the overuse of packaging materials. Similarly, the food transportation and storage sector has been playing a key role in keeping food products fresh. Many food delivery companies have to purchase refrigerated trucks, which are very emission-intensive. To address this issue, Van et al. [63] suggested that reducing the transportation distance and increasing the storage and trading volume could make the food system more sustainable. Furthermore, other measures are also effective and should be implemented, such as increasing the use of clean energy vehicles, selecting a reasonable storage address, planning the optimal transportation route, and reducing the demand for refrigeration in the transportation process.

Furthermore, it is necessary for all the stakeholders to change their food consumption behaviors so that they can fully engage in such mitigation efforts. Several scholars found that consumption behaviors can reduce food related carbon emissions of one ordinary family [64–66]. According to a survey conducted by China Agricultural University, at least 10% of the meals were wasted in restaurants [67]. The amount of protein and fat wasted by the catering industry in China is as high as 8 million tons and 3 million tons, respectively [68]. Another report prepared by the Chinese Academy of Agricultural Sciences suggests that the annual waste of food products is 61.92 million tons, including12.12 million tons of meat and 8.24 million tons of aquatic products [69]. Consequently, healthy food consumption behaviors should be promoted, such as diversifying food selections, increasing the consumption of plant-based food, reducing meat and milk products consumption, avoiding overeating, purchasing local food, and caring food expiration dates, etc. In this regard, capacity-building activities should be encouraged to improve the awareness of the general public, such as internet, TV and radio promotions, regular workshops, pamphlets, and billboards.

7 Conclusions

China is the most populous country in the world. Due to rapid industrialization and urbanization, food consumption structure of the residents in China has experienced profound changes in the past decades. Under such a circumstance, this study applied both the SRIO and MRIO methods to investigate the trends of GHG emission from food consumption at both national and provincial levels, to uncover the emission differences between urban and rural residents, the sectoral contributions, as well as the GHG emission differences between food production and consumption perspectives.

First, the results demonstrate that the total GHG emission from food consumption increased by 287% in 2002–2017. The fastest growing GHG emission is from meat products consumption, followed by other foods which include vegetables, fruits and dairy products. Although the GHG emission from food consumption has stably increased, such a GHG emission from urban residents increased faster than that from rural residents. Secondly, the GHG emission from food consumption in China mainly comes from the agricultural production, food processing and manufacturing, packaging, transportation sectors. Among the GHG emissions of various sectors, the most obvious change is from the sector of food transportation, storage and postal services. From a regional point of view, the per capita GHG emissions from food consumption increased in almost all the provinces (autonomous regions/municipalities) except Shandong Province and Jiangxi Province in 2010–2012. And the top five provinces (autonomous regions/municipalities) with the largest GHG emissions from food consumption in 2012 are Shandong Province, Hubei Province, Guangdong Province, Zhejiang Province, and Jiangsu Province. The top three provinces (autonomous regions/municipalities) with the largest gaps in total GHG emissions from food consumption between urban and rural residents are Shanghai Municipality, Tianjin Municipality, and Beijing Municipality. In terms of per capita GHG emissions, the top two regions are Inner Mongolia Autonomous Region and Xinjiang Autonomous Region, while for the GHG emission gaps between per capita of urban and rural residents, the top two regions are Shandong Province and Hubei Province. In addition, due to the differences of natural resources and industrial construction in different provinces (autonomous regions/municipalities), food trade has induced significant GHG emission transfers among different provinces (autonomous regions/municipalities) in China. The main food export provinces (autonomous regions/municipalities) are Inner Mongolia Autonomous Region, Jilin Province and Heilongjiang Province, while the main import provinces (autonomous regions/municipalities) are Beijing Municipality, Shanghai Municipality and Jiangsu Provinces.

These research findings provide valuable policy implications to mitigate food-related GHG emissions. Policy recommendations are proposed for governments, enterprises and the general public. These policies should be carefully adopted by considering the local realities since different provinces (autonomous regions/municipalities) are facing different challenges.

In general, the key research findings of this study can help understand the trend of GHG emissions from food consumption in China at national, provincial, and sectoral levels so that appropriate mitigation policies can be made by decision-makers. Such methods can be shared by other countries with similar challenges so that they can initiate their own mitigation efforts by considering the local conditions.

References

[1] Jiang H, Geng Y, Tian X, Uncovering CO₂ emission drivers under regional industrial transfer in China’s Yangtze River Economic Belt: a multi-layer LMDI decomposition analysis. Frontiers in Energy, 2020, online, doi:10.1007/s11708-020-0706-z

[2] Song X, Geng Y, Li K, Does environmental infrastructure investment contribute to emissions reduction? A case of China. Frontiers in Energy, 2020, 14(1): 57–70

[3] Mi Z, Zhang Y, Guan D, Consumption-based emission accounting for Chinese cities. Applied Energy, 2016, 184: 1073–1081

[4] Zhang X, Geng Y, Shao S, How to achieve China’s CO₂ emission reduction targets by provincial efforts? An analysis based on generalized Divisia index and dynamic scenario simulation. Renewable & Sustainable Energy Reviews, 2020, 127: 109892

[5] Li A, Lin B. Comparing climate policies to reduce carbon emissions in China. Energy Policy, 2013, 60: 667–674

[6] Cao Q, Kang W, Xu S, Estimation and decomposition analysis of carbon emissions from the entire production cycle for Chinese household consumption. Journal of Environmental Management, 2019, 247: 525–537

[7] Zhang Y J, Bian X J, Tan W, The indirect energy consumption and CO₂ emission caused by household consumption in China: an analysis based on the input–output method. Journal of Cleaner Production, 2017, 163: 69–83

[8] Wu S, Lei Y, Li S. CO₂ emissions from household consumption at the provincial level and interprovincial transfer in China. Journal of Cleaner Production, 2019, 210: 93–104

[9] Tian X, Geng Y, Dong H, Regional household carbon footprint in China: a case of Liaoning province. Journal of Cleaner Production, 2016, 114: 401–411

[10] Tian X, Geng Y, Dai H C, The effects of household consumption pattern on regional development: a case study of Shanghai. Energy, 2016, 103: 49–60

[11] Wang X, Chen S. Urban-rural carbon footprint disparity across China from essential household expenditure: survey-based analysis, 2010–2014. Journal of Environmental Management, 2020, 267: 110570

[12] Vermeulen S J, Campbell B M, Ingram J S I. Climate change and food systems. Annual Review of Environment and Resources, 2012, 37(1): 195–222

[13] Xu Z, Sun D W, Zeng X A, Research developments in methods to reduce the carbon footprint of the food system: a review. Critical Reviews in Food Science and Nutrition, 2015, 55(9): 1270–1286

[14] Wu Y, Wang X, Lu F. The carbon footprint of food consumption in Beijing. Acta Ecologica Sinica, 2012, 32(5): 1570–1577

[15] Cao Z, Hao J, Xing H. Spatial-temporal change of Chinese resident food consumption carbon emissions and its driving mechanism. Progress in Geography, 2020, 39(1): 91–99

[16] Zhi J, Gao J. Analysis of carbon emission caused by food consumption in urban and rural inhabitants in China. Progress in Geography, 2009, 3: 429–434

[17] Yue Q, Xu X, Hillier J, Mitigating greenhouse gas emissions in agriculture: from farm production to food consumption. Journal of Cleaner Production, 2017, 149: 1011–1019

[18] Lin B, Xie X. CO₂ emissions of China’s food industry: an input–output approach. Journal of Cleaner Production, 2016, 112: 1410–1421

[19] Kucukvar M, Samadi H. Linking national food production to global supply chain impacts for the energy-climate challenge: the cases of the EU-27 and Turkey. Journal of Cleaner Production, 2015, 108: 395–408

[20] Feng W, Cai B, Zhang B. A bite of China: food consumption and carbon emission from 1992 to 2007. China Economic Review, 2020, 59: 100949

[21] Song F, Reardon T, Tian X, The energy implication of China’s food system transformation. Applied Energy, 2019, 240: 617–629

[22] Su B, Ang B W. Structural decomposition analysis applied to energy and emissions: some methodological developments. Energy Economics, 2012, 34(1): 177–188

[23] Wang X, Huang H, Hong J, A spatiotemporal investigation of energy-driven factors in China: a region-based structural decomposition analysis. Energy, 2020, 207: 118249

[24] Yu M, Zhao X, Gao Y. Factor decomposition of China’s industrial electricity consumption using structural decomposition analysis. Structural Change and Economic Dynamics, 2019, 51: 67–76

[25] Dietzenbacher E, Kulionis V, Capurro F. Measuring the effects of energy transition: a structural decomposition analysis of the change in renewable energy use between 2000 and 2014. Applied Energy, 2020, 258: 114040

[26] Zhang Y. Impact of urban and rural household consumption on carbon emissions in China. Economic Systems Research, 2013, 25(3): 287–299

[27] Su B, Ang B W. Multiplicative structural decomposition analysis of aggregate embodied energy and emission intensities. Energy Economics, 2017, 65: 137–147

[28] Zhang Y J, Bian X J, Tan W, The indirect energy consumption and CO₂ emission caused by household consumption in China: an analysis based on the input–output method. Journal of Cleaner Production, 2017, 163: 69–83

[29] Yu G X, Wang X Q, Wu H J, Analysis of energy consumption and carbon emission of urban residents in Anhui province. Journal of Anhui University of Science and Technology, 2020, 40: 7 (in Chinese)

[30] Chen C. Carbon footprint estimation on food consumption of residences in Lanzhou city. Dissertation for Master’s Degree. Lanzhou: Lanzhou University, 2013

[31] Ding L. Research on spatial differences of residents’ food consumption carbon emissions in Guangdong province. Dissertation for Master’s Degree. Guangzhou: Guangzhou University, 2013

[32] National Development and Reform Commission (NDRC) of China. General Principles for Calculation of the Comprehensive Energy Consumption. Beijing: China Standards Press, 2008

[33] National Development and Reform Commission (NDRC) of China. Guidelines for the preparation of provincial GHG inventories. Beijing, China, 2010 (in Chinese)

[34] Intergovernmental Panel on Climate Change. 2006 IPCC guidelines for national greenhouse gas inventories. Institute for Global Environmental Strategies, 2006

[35] Gao Z, Geng Y, Wu R, China’s CO₂ emissions embodied in fixed capital formation and its spatial distribution. Environmental Science and Pollution Research International, 2020, 27(16): 19970–19990

[36] Peters G P, Hertwich E G. Pollution embodied in trade: the Norwegian case. Global Environmental Change, 2006, 16(4): 379–387

[37] Miller R E, Blair P D. Input-output Analysis: Foundations and Extensions. Cambridge: Cambridge University Press, 2009

[38] United Nations. Handbook of National Accounting: Integrated Environmental and Economic Accounting. Studies in Methods, Series F, No 61, New York, 1993, avabilable at the website of unstats.un.org

[39] Su B, Ang B W. Input–output analysis of CO₂ emissions embodied in trade: competitive versus non-competitive imports. Energy Policy, 2013, 56: 83–87

[40] Weber C L, Peters G P, Guan D, The contribution of Chinese exports to climate change. Energy Policy, 2008, 36(9): 3572–3577

[41] Zhang B, Qiao H, Chen Z M, Growth in embodied energy transfers via China’s domestic trade: evidence from multi-regional input-output analysis. Applied Energy, 2016, 184: 1093–1105

[42] Mi Z, Meng J, Guan D, Chinese CO₂ emission flows have reversed since the global financial crisis. Nature Communications, 2017, 8(1): 1712

[43] Guan D, Peters G P, Weber C L, Journey to world top emitter: an analysis of the driving forces of China’s recent CO₂ emissions surge. Geophysical Research Letters, 2009, 36(4): L04709

[44] Shuang S, Xiu F F. Evolution of final demand pattern, changes in industrial structure and CO₂ emission in China–based on input-output model and SDA method. Journal of Shanxi Finance and Economics University, 2013, 35: 11 (in Chinese)

[45] National Bureau of Statistics of China. Input-Output Table of China 2002, 2007, 2012, 2017. 2017, available at the website of stats.gov.cn

[46] National Bureau of Statistics of China. China Energy Statistical Yearbook. Beijing: China Statistics Press (in Chinese)

[47] Fan Z. Study on the impact of family population on residents’ consumption carbon emissions—analysis of CFPs data. Dissertation for Master’s Degree. Kaifeng: Henan University, 2019

[48] Liu W, Tang Z, Han M. The 2012 China Multi-Regional Input-Output Table of 31 Provincial Units. Beijing: China Statistics Press

[49] National Bureau of Statistics of China. Industrial classification for national economic activities, 2011, available at the website of stats.gov.cn

[50] Zhuang M, Geng Y, Pan H, Ecological and socioeconomic impacts of payments for ecosystem services—a Chinese garlic farm case. Journal of Cleaner Production, 2021, 285: 124866

[51] Qian H. The structural decomposition of income gap sources between urban and rural residents in China. Statistics & Decisions, 2020, 36(20): 76–79

[52] Geng Y, Fujita T, Chiu A, Responding to the Paris Climate Agreement: global climate change mitigation efforts. Frontiers in Energy, 2018, 12(3): 333–337

[53] Deng W F. Overview of China’s fertilizer industry in 2019. 2020, available at the website of leadleo.com

[54] Ministry of Agriculture and Rural Affairs of China. Action Plan for Zero Growth in Fertilizer Use by 2020. 2015, available at the website of moa.gov.cn

[55] Su B, Huang H C, Ang B W, Input–output analysis of CO₂ emissions embodied in trade: the effects of sector aggregation. Energy Economics, 2010, 32(1): 166–175

[56] Su B, Ang B W. Multi-region input–output analysis of CO₂ emissions embodied in trade: the feedback effects. Ecological Economics, 2011, 71: 42–53

[57] Kim B, Neff R. Measurement and communication of greenhouse gas emissions from US food consumption via carbon calculators. Ecological Economics, 2009, 69(1): 186–196

[58] Albert O, Marianne T, Jonathan L, Tracking the carbon emissions of Denmark’s five regions from a producer and consumer perspective. Ecological Economics, 2020, 177: 106778

[59] Yang X, Jia X. Low-carbon economy and low-carbon food. Energy Procedia, 2011, 5: 1099–1103

[60] Xian J, Bian J. China launches clean plate campaign 2.0 as Xi calls for end to food wastage. 2020, available at the website of People’s Daily (in Chinese)

[61] Zhang W F, Dou Z X, He P, New technologies reduce greenhouse gas emissions from nitrogenous fertilizer in China. Proceedings of the National Academy of Sciences of the United States of America, 2013, 110(21): 8375–8380

[62] Zentner R P, Lafond G P, Derksen D A, Effects of tillage method and crop rotation on non-renewable energy use efficiency for a thin Black Chernozem in the Canadian Prairies. Soil & Tillage Research, 2004, 77(2): 125–136

[63] Van Hauwermeiren A, Coene H, Engelen G, Energy lifecycle inputs in food systems: a comparison of local versus mainstream cases. Journal of Environmental Policy and Planning, 2007, 9(1): 31–51

[64] Davis J, Sonesson U, Baumgartner D U, Environmental impact of four meals with different protein sources: case studies in Spain and Sweden. Food Research International, 2010, 43(7): 1874–1884

[65] González A D, Frostell B, Carlsson-Kanyama A. Protein efficiency per unit energy and per unit greenhouse gas emissions: potential contribution of diet choices to climate change mitigation. Food Policy, 2011, 36(5): 562–570

[66] Weber C L, Matthews H S. Food-miles and the relative climate impacts of food choices in the United States. Environmental Science & Technology, 2008, 42(10): 3508–3513

[67] Ding S. The potential of saving food. West China Development, 2013, 03: 52–54 (in Chinese)

[68] Liu Y Q. Do you have cleared your plate today? High School Years, 2013, 13: 4–5 (in Chinese)

[69] Ding S. What is the potential of saving food? Qiushi, 2013, 05: 23–24 (in Chinese)

RIGHTS & PERMISSIONS

Higher Education Press