Spatial–Temporal Evolution and Factor Decomposition for Ecological Pressure of Carbon Footprint in the One Belt and One Road

Low-carbon and green development is important to promote the sustainable economic and social development of countries along the One Belt and One Road. These countries have distinct differences in their ability to withstand carbon emission pressures and their driving factors, due to different stages of development and economic characteristics. This paper presents a model of ecological pressure of the carbon footprint in energy consumption (EPcfec), defined by three states: light, balanced, and heavy pressure. The EPcfec was calculated from data on 56 countries along the One Belt and One Road from 1994–2014, and analysis of the data’s temporal and spatial evolutionary rules was conducted. Furthermore, we used the LMDI method to extract the driving forces of EPcfec and evaluated the contribution of these factors to the overall region and seven sub-regions. The results showed that EPcfec growth slowed over time, with the value of EPcfec reaching 3190.51 in 2014. Resource-rich countries have a greater value of EPcfec and are mainly distributed in parts of West Asia, North Africa, and Southeast Asia. The per capita export of goods and services, and the population density on productive land contribute to ecological pressure on the carbon footprint. Energy structure, the influence of international trade on GDP, and energy intensity exerted an inhibitory effect on the ecological pressure of the carbon footprint. This paper proposes mitigation measures for optimizing energy structure, improving energy efficiency, developing low energy consumption, and promoting green international trade. Our results provide support for countries along the One Belt and One Road to mitigate ecological pressures resulting from their carbon footprint.


Introduction
Actively responding to global climate change has become an urgent issue for all countries [1]. In December 2015, 196 countries signed the Paris Agreement to jointly promote a plan to curb global warming [2]. China is the first developing country to propose energy-saving and emission reduction targets. It has committed itself to reducing carbon emission intensity in 2020 to 55-60% of the levels in 2005 [3]. In November 2014, the Chinese government strived to realize the target of decreasing carbon emission as soon as possible [4,5] in accordance with the China-US Joint Statement on Climate Change. Many countries have also successively promulgated their own low-carbon action plans. At present, the situation of carbon emissions is still very serious, and is mainly driven by economic sectors with high energy consumption. Global greenhouse gas emission data show that in 2015, global carbon At present, factor decomposition is mostly used in the study of carbon emission driving factors, and a flood of related research has been undertaken in this area since the mid-1980s. There are four main types of methods used. The first category comprises those that focus on identifying the factors affecting carbon emissions through structural decomposition techniques (SDA) under the input-output analysis framework, combined with input-output tables and models [36]. Feng [37] found that the economic growth of the United States during 1997-2007 was the key factor driving the increase of carbon emissions. The change in the energy structure over this period was an important factor in restraining carbon emission growth. Such studies often rely on input-output tables for non-continuous intertemporal analysis, and the types of impact factors are single, which makes it difficult to examine the impact of multiple types of socioeconomic factors. The second category uses the Computable General Equilibrium (CGE) model and scenario analysis to calculate the impact of different socioeconomic factors and their combinations on carbon emissions, in accordance with general equilibrium conditions. This type of research is also dependent on input-output tables, however it is not straightforward to uncover the underlying continuum of a time series [38,39]. The third category comprises the use of econometric models to study the impact of various factors on carbon emissions. There are many such methods, but all of them can reveal only statistical laws, not the intrinsic driving factors [40,41]. The fourth category uses the Kaya identity, along with extensions and improvements. They use factorial decomposition techniques (IDA) to calculate the contribution of different factors. This type of method has a variety of decomposition forms and can be selected based on research needs [42][43][44][45]. The logarithmic mean Divisia index (LMDI) method is such a method. It was proposed by Ang et al. in 1998, who deduced its mathematical properties, as well as summarizing a large number of applications of IDA models. Its advantages include using of diverse forms, simple calculations, and the ability to find intrinsic driving factors [46][47][48][49]. The method is used in many research fields, such as for the decomposition of energy consumption or carbon emission factors. For example, Vinuya et al. [50,51] used the LMDI decomposition model to find that economic and population growth is the largest factor driving carbon emissions. The increase in energy use efficiency and the decline in energy intensity can effectively restrain the increase in carbon emissions due to the impact of GDP and population increase. Ma and Stern improved the IDA model and used the LMDI method to decompose the energy intensity changes from 1980-2003, confirming that technological progress is the most important factor in reducing energy intensity [52]. Zhang et al. used the LMDI method to find that the rapid growth of carbon emissions in China's power generation industry is mainly due to the use of coal, and energy efficiency is also an important contributor [53]. Inglesi-Lotz applied LMDI to study the main factors affecting the changes in CO 2 emissions in South Africa. They found that energy intensity had a negative impact on CO 2 emissions during 2008-2014 [54]. The use of LMDI by Kopidou et al. showed that the two main drivers of industrial CO 2 emissions and employment are economic growth and resource intensity, and that the optimization of the energy structure is conducive to national emission reduction [55].
The One Belt and One Road is a shared, win-win, open economic and trade cooperation initiative, covering many countries and regions. At present, most of the research on One Belt and One Road focuses on strategic significance, institutional design [11,56], and capacity cooperation [57][58][59]. The primary focus has been how to promote the economic development of each country. However, there are several studies on the green and coordinated development of the international regional economy. The related literature focuses on the OECD [60,61], East Asia, ASEAN [62,63], the European Union and its sub-regions [17,64], and a few international regions, meaning less attention is paid to the coordinated development of the regional economy and environment in the One Belt and One Road area. In fact, most of the countries along the "One Belt and One Road" are developing or underdeveloped countries, many of which are not concerned with the problem of low carbon and emission reduction. In addition, the development stages and economic characteristics of the countries along the route are quite different. Therefore, selecting countries along the One Belt and One Road as the target and exploring the carbon footprint ecological pressures of those countries is of great significance to sustainable development; it can also contribute to global emission reduction targets.
In summary, previous research has mainly focused on the measurement and causes of carbon emissions and carbon footprints. There has been little research on ecological pressure caused by carbon footprints, and it does not directly reflect the impact of carbon emissions on the ecological environment.
The LMDI method has been widely used in carbon emission and carbon footprint research. In previous work looking at causal factors, the choice of factors was not varied, with the main focus always being the country. Explanations of changes to ecological pressure resulting from the global carbon footprint were insufficient. In addition, the scope of the countries included was relatively limited, and the differences in economic characteristics and resource endowments among countries are relatively small, which limits the applicability of the research. By selecting countries along the One Belt and One Road with large differences in economic characteristics and resource endowments-based on an ecological-pressure-based model of energy consumption and the carbon footprint-we adopted the LMDI decomposition method to introduce the driving factors that reflect the links between the global carbon footprint and ecological pressure. Factors include the per capita export goods and services trade volume, and the influence of international trade on GDP. Our approach makes the study of the ecological pressures arising from energy consumption as measured by the carbon footprint more straightforward.
The rest of the paper is organized as follows: In Section 2, the concepts of the energy consumption carbon footprint (CFEC) and carbon footprint ecological pressure (EP cfec ) are proposed, and the measurement and factor decomposition model of EP cfec are explained and constructed. Section 3 introduces the research scope, data sources, and regional divisions. Section 4 analyzes the calculation results and spatial evolution of EP cfec from 1994-2014, and uses the LMDI method to decompose EP cfec . The conclusions and recommendations are given in Section 5.

Carbon Footprint Measurement Model of Energy Consumption
The carbon footprint is a quantified value that measures carbon emissions. It is based the concept of an ecological footprint and can directly measure the response of a natural system to carbon emissions from human activities [65]. In this paper, we use a measure of carbon footprint based on energy consumption, mainly related to coal, oil, and natural gas. The carbon footprint is measured using the productive land area needed to store energy consumption (see Equation (1)). According to the data provided by the IPCC (2006), productive land includes forests, grasslands, arable land, gardens, and other agricultural lands, of which forest and grassland carbon reserves account for 93% of the total productive land [66]: where CFEC refers to the carbon footprint of energy consumption (hm 2 ); Q i is the energy consumption of the ith energy (t); and EF i is the energy carbon emission factor (t/tec), for which the study takes the average from several agencies, such as the EIA, as shown in Table 1 [67 -69]. β is the conversion coefficient for converting carbon emissions into land area, for which the adopted WWF's value is 6.49 t/hm 2 [70]. The ecological pressure of the carbon footprint in energy consumption (EP cfec ) refers to the pressure of carbon emissions on the natural ecosystem and is the ratio of the carbon footprint to productive land area. The calculation is as follows: where CFEC is the carbon footprint as described above; and S f , S a , S c , and S p refer to the area of forest, cultivated land, permanent cropland, and permanent ranching land, respectively, with the unit hm 2 .
When EP cfec ∈ (0, 1), the productive land can fully absorb the carbon emissions resulting from energy consumption, and the carbon emissions, therefore, exert little pressure on the ecological environment. The smaller the value of EP cfec , the lower the ecological pressure. This is termed the light pressure state.
When EP cfec = 1, the carbon emissions generated by energy consumption are equal to those of being absorbed by productive land; that is, the productive land has reached its capacity to absorb the carbon emissions from energy consumption. This is the balanced state.
When EP cfec ∈ (1, +∞), the productive land cannot fully absorb the carbon emissions, and the ecological environment faces a large amount of pressure from carbon storage. The larger the value of EP cfec , the greater the ecological pressure on the carbon footprint. At this point, it is in the high pressure state.
A schematic diagram of the three pressure states is shown in Figure 1.

The Econometric Model of EPcfec
The ecological pressure of the carbon footprint in energy consumption (EPcfec) refers to the pressure of carbon emissions on the natural ecosystem and is the ratio of the carbon footprint to productive land area. The calculation is as follows: where CFEC is the carbon footprint as described above; and Sf, Sa, Sc, and Sp refer to the area of forest, cultivated land, permanent cropland, and permanent ranching land, respectively, with the unit hm 2 .
When EPcfec ∈ (0, 1), the productive land can fully absorb the carbon emissions resulting from energy consumption, and the carbon emissions, therefore, exert little pressure on the ecological environment. The smaller the value of EPcfec, the lower the ecological pressure. This is termed the light pressure state.
When EPcfec = 1, the carbon emissions generated by energy consumption are equal to those of being absorbed by productive land; that is, the productive land has reached its capacity to absorb the carbon emissions from energy consumption. This is the balanced state.
When EPcfec ∈ (1, +∞), the productive land cannot fully absorb the carbon emissions, and the ecological environment faces a large amount of pressure from carbon storage. The larger the value of EPcfec, the greater the ecological pressure on the carbon footprint. At this point, it is in the high pressure state.
A schematic diagram of the three pressure states is shown in Figure 1.

Decomposition Method of EPcfec
The paper applies the LMDI method to the decomposition of ecological pressure factors influencing the carbon footprint. The model introduces two factors related to the EPcfec-the

Decomposition Method of EP cfec
The paper applies the LMDI method to the decomposition of ecological pressure factors influencing the carbon footprint. The model introduces two factors related to the EP cfec -the influence of international trade on GDP, per capita exports of goods and service trade-reflecting linked changes in the ecological pressure between countries. We decomposed the driving factors into five categories: energy structure, energy intensity, the influence of international trade on GDP, per capita export goods and services trade volume, and population intensity of productive land (see Equation (3)). The variables are defined in Table 2.
where C is the carbon emission of non-renewable energy as described above; Q is the consumption of non-renewable energy; E is the total energy consumption including renewable energy and non-renewable energy; Y is the gross domestic product; EXP is exports of goods and services; POP is the total population of every country; H is the productive land area; ε means coefficient; EF means carbon emission intensity, which is the ith energy carbon emission coefficient; ES means the energy structure that is the proportion of non-renewable energy in the total energy consumption; EI means the energy intensity that is the ratio of total energy consumption to GDP; YE means the influence of international trade on GDP, which is the ratio of GDP to exports of goods and services; EP means the per capita exports of goods and service trade, which is the ratio of exports of goods and services trade to the population; PH means the population intensity of productive land, which is the ratio of the population to productive land; and ε means the reciprocal of the conversion coefficient β, as shown in Table 2.  (1)) hm 2 /t The contribution of each individual factor to the ecological pressure was calculated by using the LMDI model in Equation (4), and the rate of change of each individual factor was calculated by the decomposition in Equation (5). where: where: In Equations (4) and (5), ∆EP cfec is the change in EP cfec ; ∆EP EF , ∆EP ES , ∆EP EI , ∆EP YE , ∆EP EP , ∆EP PH , and ∆EP ε denote the change in carbon intensity, energy structure, energy intensity, influence of international trade on GDP, per capita exports of goods and service trade, population intensity of productive land, and the conversion coefficient from the 0th to the Tth year, respectively-these are the contribution values; W T is the weight of influence; D represents the change in EP cfec , D EF , D ES , D EI , D YE , D EP , D PH , and D ε are the rates of change in carbon intensity, energy structure, energy intensity, Influence of international trade on GDP, per capita import and export goods and service trade, population intensity of productive land, and the conversion factor from the 0th to the Tth year, respectively. ES 0 , EI 0 , YE 0 , EP 0 , PH 0 and ES T , EI T , YE T , EP T , PH T denote the energy structure, energy intensity, influence of international trade on GDP, per capita export goods and service trade volume, and population intensity of productive land in the 0th year and Tth year, respectively.

Data Sources
According to the bilateral trade agreement of the One Belt and One Road Initiative, combined with the availability of data, 56 countries along the route were selected, as shown in Table 3. Due to the unavailability of data eight countries were excluded: Armenia, Bosnia and Herzegovina, Hungary, Moldova, Montenegro, Nepal, Serbia, and East Timor. The time period used was 1994-2014.  The information analysis center of the Oak Ridge National Laboratory in the United States Carbon Dioxide Information Analysis Center (CDIAC) provides carbon emission data [71]. The primary energy sources for the countries studied mainly comprise coal, oil, and natural gas. Compiling data on energy consumption and carbon emissions shows that the carbon emissions of countries along the One Belt and One Road are on the rise. The highest value of the carbon emissions in 2012 was 486.41 × 10 7 t after which a downward trend occurred, showing that various countries began to focus on green and low-carbon development models. Carbon emissions from the energy consumption of coal, oil, and natural gas are also decreasing. From 2000-2012, the carbon emissions from oil and natural gas production have changed by a small amount, and remained within the range 50 × 10 7 -70 × 10 7 t. Of these, coal combustion is the main source of carbon emissions and is growing rapidly. In 2012, the carbon consumption of coal consumption reached 251.9 × 10 7 t, which was 1.5 times that of oil and natural gas consumption. The trends are shown in Figure 2.
within the range 50 × 10 7 -70 × 10 7 t. Of these, coal combustion is the main source of carbon emissions and is growing rapidly. In 2012, the carbon consumption of coal consumption reached 251.9 × 10 7 t, which was 1.5 times that of oil and natural gas consumption. The trends are shown in Figure 2. Data about the productive land area, population, and GDP was taken from the World Bank public data resource [72]. Taking into account the integrity and continuity of the data, a small amount of data was processed using interpolation. In order to analyze regional differences, 56 countries along the route were divided into seven regions: Mongolia, Russia, Central Asia, West Asia, North Africa, Central and Eastern Europe, Southeast Asia, South Asia, and East Asia ( Table 3). The spatial layout is shown in Figure 3.  Data about the productive land area, population, and GDP was taken from the World Bank public data resource [72]. Taking into account the integrity and continuity of the data, a small amount of data was processed using interpolation. In order to analyze regional differences, 56 countries along the route were divided into seven regions: Mongolia, Russia, Central Asia, West Asia, North Africa, Central and Eastern Europe, Southeast Asia, South Asia, and East Asia ( Table 3). The spatial layout is shown in Figure 3. within the range 50 × 10 7 -70 × 10 7 t. Of these, coal combustion is the main source of carbon emissions and is growing rapidly. In 2012, the carbon consumption of coal consumption reached 251.9 × 10 7 t, which was 1.5 times that of oil and natural gas consumption. The trends are shown in Figure 2. Data about the productive land area, population, and GDP was taken from the World Bank public data resource [72]. Taking into account the integrity and continuity of the data, a small amount of data was processed using interpolation. In order to analyze regional differences, 56 countries along the route were divided into seven regions: Mongolia, Russia, Central Asia, West Asia, North Africa, Central and Eastern Europe, Southeast Asia, South Asia, and East Asia ( Table 3). The spatial layout is shown in Figure 3.

The Carbon Footprint in Energy Consumption
The changes of the carbon footprint in energy consumption (CFEC) derived from Equation (1)

The Carbon Footprint in Energy Consumption
The changes of the carbon footprint in energy consumption (CFEC) derived from Equation (1) for 56 countries, from 1994-2014, are shown in Figure 4.  With the increase of energy consumption, the carbon footprint increased during the period 2001-2012. In 2012, the peak value of 74.9 × 10 7 hm 2 was reached, after which there was a downward trend. This is the same trend as for carbon emissions of countries along the One Belt and One Road during the same period.

Ecological Pressure of the Carbon Footprint in Energy Consumption
The EPcfec values of 56 countries from 1994-2014 were calculated according to Equation (2), and are shown in Tables 4 and 5.   In 2012, the peak value of 74.9 × 10 7 hm 2 was reached, after which there was a downward trend. This is the same trend as for carbon emissions of countries along the One Belt and One Road during the same period.

Ecological Pressure of the Carbon Footprint in Energy Consumption
The EP cfec values of 56 countries from 1994-2014 were calculated according to Equation (2), and are shown in Tables 4 and 5. In general, the EP cfec has increased in all countries. Of these, 44 countries are under light pressure and 12 countries are under high pressure. Some countries are already under excessive high pressure (i.e., the EP cfec of SGP reached 2876.59, which is 50.49 times the average value of 56.97 in 2014). Carbon emissions from energy consumption in some countries, such as Singapore, Bahrain, Qatar, Kuwait, UAE, Israel, and Brunei, have far exceeded the digestion capacity of the productive land. On the one hand, these countries are more dependent on non-renewable energy, and on the other hand, the area of productive land is relatively small. Countries with light pressure can also naturally digest carbon emissions from energy consumption, especially countries with an EP cfec of less than 0.1, such as Kazakhstan, Mongolia, and Saudi Arabia. This is directly related to the large area of productive land in these countries. Therefore, protecting or expanding productive land area has a direct impact on reduction of the EP cfec . Figure 5 shows the spatial evolution of the EP cfec in 56 countries in 1994, 2002, 2008, and 2014. It can be seen that there is a significant difference in the EP cfec among these countries. Mongolia, Kazakhstan, Kyrgyzstan, Tajikistan, Afghanistan, Laos, Cambodia, Myanmar, Yemen, and other countries were under light pressure. The EP cfec is always below 0.05 and the change is small. Although China, Iraq, Jordan, and Thailand, among others, are under light pressure, but with a rapid rate of increase-the EP cfec in these countries has increased from a relatively low level to over 0.5, which is related to the increase in carbon emissions in these countries. Oman, Lebanon, and Malaysia have changed from light pressure to heavy pressure, which is related to the small changes in productive land in these countries and the growth of carbon emissions. In the Czech Republic, Croatia, Slovenia, Russia, and some other countries, the EP cfec has shown a downward trend, and some countries have shifted from a heavy state of pressure to a light pressure state. This is related to the decline in carbon emissions in these countries, mainly because the carbon emissions from coal are decreasing. in productive land in these countries and the growth of carbon emissions. In the Czech Republic, Croatia, Slovenia, Russia, and some other countries, the EPcfec has shown a downward trend, and some countries have shifted from a heavy state of pressure to a light pressure state. This is related to the decline in carbon emissions in these countries, mainly because the carbon emissions from coal are decreasing.   Figure 6 shows the time variation of the EPcfec in seven regions. The areas with greater ecological stress are mainly located in Southeast Asia, West Asia, and North Africa, and the lowest EPcfec value is 6.71. However, in Central Asia, where the level of income is relatively low, and in Mongolia and Russia, where the area of productive land is relatively large, the ecological pressure is relatively low, and the EPcfec value is between 0.05 and 0.17.

Factors Decomposition for the Overall Region
Equations (3)-(5) were used to decompose the ecological pressure of the carbon footprint in energy consumption, and calculate the contribution and rate of change of EPcfec for different factors. Figure 7 shows changes in the contribution value of each factor to EPcfec, and Figure 8 shows the rate of change of each factor. A contribution value greater than 0 indicates that the factor has a promoting effect on the EPcfec, and a value of less than 0 shows an inhibitory effect. The greater the absolute value, the greater the influence.

Factors Decomposition for the Overall Region
Equations (3)-(5) were used to decompose the ecological pressure of the carbon footprint in energy consumption, and calculate the contribution and rate of change of EP cfec for different factors. Figure 7 shows changes in the contribution value of each factor to EP cfec , and Figure 8 shows the rate of change of each factor. A contribution value greater than 0 indicates that the factor has a promoting effect on the EP cfec , and a value of less than 0 shows an inhibitory effect. The greater the absolute value, the greater the influence. Note that in the original statistics for Myanmar, the maximum value of the influence of international trade on GDP is 1005.36 and the minimum is 4.98, which is abnormal. Including the factor analysis of the factors in the overall region, the contribution of other factors to the EPcfec was diluted. Therefore, Myanmar was excluded from this part of the analysis. Note that in the original statistics for Myanmar, the maximum value of the influence of international trade on GDP is 1005.36 and the minimum is 4.98, which is abnormal. Including the factor analysis of the factors in the overall region, the contribution of other factors to the EPcfec was diluted. Therefore, Myanmar was excluded from this part of the analysis. As can be seen from Figure 7, the five factors that contributed to EPcfec in 2014 are ranked as follows: energy intensity, influence of international trade on GDP, energy structure, population intensity of productive land, and per capita exports of goods and service trade. Among them, three factors (energy intensity, influence of international trade on GDP, and energy structure) play a decisive role; two factors (population intensity of productive land, per capita export goods and service trade) play a catalytic role.
(1). Energy intensity (EI): From 1994-2014, EI had an inhibitory effect on the EPcfec gradually increased over time. The contribution in 2014 was −46.64, and the rate of change was 0.55, as shown in Figure 8a. The energy consumption intensity of the countries also decreased significantly over the same period. Countries should continue to implement energy-saving and emission-reduction measures, significantly reduce energy consumption per unit, in order to strengthen the inhibitory effect of energy consumption intensity on the EPcfec.
(2). Influence of international trade on GDP (YE): in 1994-2014, the contribution of this factor was between −12 and 0, and the rate of change was between 0.69 and 1, as shown in Figure 8b. This factor has an inhibitory effect on the EPcfec but shows a slight weakening trend. In the same period, YE gradually decreased and export dependency rose slowly. This shows that international trade has gradually enhanced the economic growth of countries along the One Belt and One Road and has also increased the ecological pressure arising from the carbon footprint. Therefore, countries along the One Belt and One Road urgently need to develop their foreign trade in accordance with low-carbon, environmental protection, and green models.
(3). Energy structure (ES): from 1994-2014, the contribution of the ES to the EPcfec was between −1.3 and 0.36, and the rate of change was between 0.96 and 1.01, as shown in Figure 8c. The impact on the EPcfec fluctuated, but not by much. In general, it showed inhibitory effects and only acted as a promoter in 2008 and 2011 during which time its promotion was small. The ES mainly depends on the proportion of non-renewable energy, such as coal, oil, and natural gas. This proportion has been relatively large in the countries along the One Belt and One Road, and has remained at about 78%. Therefore, there is a very great potential for reducing the ecological pressure of the carbon footprint in energy consumption by adjusting the energy structure.
(4). Population intensity of productive land (PH): This factor has a significant role in increasing the EPcfec and is rising year by year. In 2014, the contribution was 42.86 and the rate of change was 2.57, as shown in Figure 8d. In the same period, the PH increased continuously, and the population intensity in 2014 was 2.57 times that of 1994. Countries along the One Belt and One Road are mostly developing countries and underdeveloped countries, with a large population base and strong growth inertia. It is hard to reduce the population size in the short term. Therefore, the key to reduce the EPcfec is to protect or expand the area of productive land, such as forests and grasslands.
(5). Per capita exports of goods and services (EP): This factor made the largest contribution to the EPcfec. From 1999-2014, the contribution to the EPcfec rose dramatically from 9.72-53.39, an Note that in the original statistics for Myanmar, the maximum value of the influence of international trade on GDP is 1005.36 and the minimum is 4.98, which is abnormal. Including the factor analysis of the factors in the overall region, the contribution of other factors to the EP cfec was diluted. Therefore, Myanmar was excluded from this part of the analysis.
As can be seen from Figure 7, the five factors that contributed to EP cfec in 2014 are ranked as follows: energy intensity, influence of international trade on GDP, energy structure, population intensity of productive land, and per capita exports of goods and service trade. Among them, three factors (energy intensity, influence of international trade on GDP, and energy structure) play a decisive role; two factors (population intensity of productive land, per capita export goods and service trade) play a catalytic role.
(1) Energy intensity (EI): From 1994-2014, EI had an inhibitory effect on the EP cfec gradually increased over time. The contribution in 2014 was −46.64, and the rate of change was 0.55, as shown in Figure 8a. The energy consumption intensity of the countries also decreased significantly over the same period. Countries should continue to implement energy-saving and emission-reduction measures, significantly reduce energy consumption per unit, in order to strengthen the inhibitory effect of energy consumption intensity on the EP cfec .
(2) Influence of international trade on GDP (YE): in 1994-2014, the contribution of this factor was between −12 and 0, and the rate of change was between 0.69 and 1, as shown in Figure 8b. This factor has an inhibitory effect on the EP cfec but shows a slight weakening trend. In the same period, YE gradually decreased and export dependency rose slowly. This shows that international trade has gradually enhanced the economic growth of countries along the One Belt and One Road and has also increased the ecological pressure arising from the carbon footprint. Therefore, countries along the One Belt and One Road urgently need to develop their foreign trade in accordance with low-carbon, environmental protection, and green models.
(3) Energy structure (ES): from 1994-2014, the contribution of the ES to the EP cfec was between −1.3 and 0.36, and the rate of change was between 0.96 and 1.01, as shown in Figure 8c. The impact on the EP cfec fluctuated, but not by much. In general, it showed inhibitory effects and only acted as a promoter in 2008 and 2011 during which time its promotion was small. The ES mainly depends on the proportion of non-renewable energy, such as coal, oil, and natural gas. This proportion has been relatively large in the countries along the One Belt and One Road, and has remained at about 78%. Therefore, there is a very great potential for reducing the ecological pressure of the carbon footprint in energy consumption by adjusting the energy structure.
(4) Population intensity of productive land (PH): This factor has a significant role in increasing the EP cfec and is rising year by year. In 2014, the contribution was 42.86 and the rate of change was 2.57, as shown in Figure 8d. In the same period, the PH increased continuously, and the population intensity in 2014 was 2.57 times that of 1994. Countries along the One Belt and One Road are mostly developing countries and underdeveloped countries, with a large population base and strong growth inertia. It is hard to reduce the population size in the short term. Therefore, the key to reduce the EP cfec is to protect or expand the area of productive land, such as forests and grasslands.
(5) Per capita exports of goods and services (EP): This factor made the largest contribution to the EP cfec . From 1999-2014, the contribution to the EP cfec rose dramatically from 9.72-53.39, an increase of 5.49 times, and the rate of change climbed from 1.35 to 3.24, an increase of 2.4 times, as shown in Figure 8e. In the same period, the EP also experienced rapid growth, which promoted the economic growth of all countries. This shows that the export trade of goods and services in certain countries is in a period of high growth, accompanied by high energy consumption. This is consistent with the YE factor. Therefore, the development of low-carbon and green international trade will be the key to reducing the EP cfec . While actively conducting foreign trade, countries along the One Belt and One Road need to provide new impetus for foreign trade development through means such as technological innovation and structural adjustment.

Decomposition Analysis of Factors in Different Sub-Regions
Changes of the studied ecological pressure drivers across the seven sub-regions are summarized in Table 6. (1) Mongolia and Russia The energy structure, energy intensity and the Influence of international trade on GDP caused depressing effects, among which energy intensity played the greatest role. Per capita export goods and services trade has a significant role in increasing the EP cfec , and contributed the most to the EP cfec . The population intensity of productive land has shown a weak promotion effect, and its contribution gradually decreased.
(2) Central Asia The energy structure, energy intensity and influence of international trade on GDP had a suppressing effect on the EP cfec , among which the energy intensity showed the greatest inhibition. The per capita export goods and services trade caused fluctuations. The ∆EP EP was −0.01 from 1994-1999, showing the inhibition effect. During the period from 2000-2004, it was transformed into a promoting factor. During the period of 2010-2014, the ∆EP EP reached 0.04, and the rate of change was 1.88. The intensity of productive land population always showed a promotive effect, and the effect was greater than that of the per capita export goods and services trade volume.

(3) West Asia and North Africa
The energy structure, energy intensity, and the influence of international trade on GDP inhibited the EP cfec in this region. The inhibitory effect of energy intensity gradually increased, and there was little change in the energy structure. The inhibitory function of YE increased at first and then weakened. The per capita value of the trade in goods and services and the population intensity of productive land played a catalytic role. The effect of the former was greater than that of the latter, and both indicators showed an upward trend over time.

(4) Central and Eastern
The energy structure, the Influence of international trade on GDP and the energy intensity had an inhibitory effect on the EP cfec , with the effect is moving from small to large and increasing year by year. The per capita export goods and service trade volume, and the population intensity of productive land were both promoting effects. The former had a greater effect than the latter, and both indicators are increasing over time.

(5) Southeast
The influence of international trade on GDP had an inhibiting effect in this region. The energy structure has been transformed from inhibition to promotion, and the promotion shows an increasing trend. The energy intensity has been transformed from promotion to inhibition. The per capita exports of goods and services, the trade volume, and the population intensity of productive land still show a positive effect. The role of these factors has increased over time. The rapidly expanding population size of the region is the main reason for the promotion of the population intensity of productive land. The main reason for this is the consumption of non-renewable energy, which is the main reason for the transformation of the energy structure into a promoting effect.

(6) South Asia
Energy intensity always played an inhibitory role, and caused and increasingly strong effect year by year. The overall performance of the energy structure has increased, but the promotion effect has been relatively small, and showed a slight inhibitory effect during 2005-2009. The Influence of international trade on GDP is generally low. From 2010-2014, it began to show a small promoting effect and needs more attention. The per capita export goods and services trade volume, and the population intensity of productive land showed a promoting effect. The former is greater than the latter, and the promotion of both indicators showed an upward trend. The area is also dominated by primary energy consumption, which is the most important reason that the overall performance of the energy structure is the driving force.

(7) Eastern Asia
The energy intensity had a suppressing effect on the EP cfec and the overall trend of suppressing effect enhanced. The influence of international trade on GDP is generally suppressed. The energy structure, the per capita export goods and service trade, and the population intensity of productive land are all contributing factors, and the impact is enhanced year by year. Among them, the per capita export of goods and services trade has the greatest promotion effect, while the energy structure and the productivity of productive land tended to be more moderate, indicating that the energy structure adjustment and productive land protection measures in the region work to good effect.
The driving effects of the various factors in the seven sub-regions are shown in Table 7. √ represents that the factor contributes to the EP cfec ; × represents the factor having an inhibitory effect on the EP cfec ; and is used to show the role of the factor changes over time.

Discussion
From the above research results, we can see that the ecological pressure of the carbon footprint in energy consumption is creating a serious challenge for the countries along the One Belt and One Road. Therefore, all the countries along the One Belt and One Road should pay more attention to the issue of ecological pressure, including setting clear emission reduction targets and emission reduction paths, in order to change the development model, realize green and low-carbon development, and to coordinate the balance between economic growth and environmental protection. In the future, with the aim of reducing the ecological pressure of the carbon footprint, governments should take green international trade and protecting the productive land as the primary methods, while continuing to reduce energy intensity and optimize the energy using structure (i.e., the clean energy occupying the most percentage of all the energy in the development). The green development of international trade requires all countries to strengthen cooperation and coordinate with each other, and fully consider the impact on other countries' environment in terms of export trade and capacity cooperation. These countries should fully consider their own resource endowments and regional advantages, choose a decompression route, focus on developing the domestic economy and international trade, reduce the ecological pressure of the carbon footprint, and provide new impetus for economic growth. For instance, the Mongolia and Russia region should continue to exert the inhibitory effect of the population intensity of the productive area, and Eastern Asia, South Asia, and Southeast Asia should further optimize the energy structure, increase its inhibitive influence, and reduce its effect promoting the EP cefc . Economically underdeveloped regions, such as Central Asia, should learn from the experience of other countries by choosing green low-carbon industries as soon as possible and adopting a sustainable development model.

Conclusions
Utilizing data from the countries along the One Belt and One Road from 1994-2014, we measured the ecological pressure of the carbon footprint in energy consumption (EP cefc ), and revealed the temporal and spatial dynamic changes EP cefc . Furthermore, by using the LMDI method to decompose its driving factors, contributions to ecological pressure arising from the carbon footprint in energy consumption was analyzed for the entire region and in seven sub-regions of the One Belt and One Road.
In general, the value of the EP cefc along the One Belt and One Road increased annually as a whole. From 1994-2014, the value of the EP cefc of 56 countries along the One Belt and One Road increased annually. The growth rate increased at first and then decreased, which indicates that the overall EP cefc of the countries along One Belt and One Road is still high. However, the growth trend has gradually slowed down. In terms of sub-regions, the region with the most ecological stress on the carbon footprint was found to be Southeast Asia, and the lowest was Central Asia. The proportion of carbon footprint ecological pressure in Central and Eastern Europe and Mongolia was gradually decreasing over the study period. The proportion in Southeast Asia, South Asia, and China was gradually increasing, and Central Asia remains basically unchanged. The per capita export of goods and services, and the population density on productive land contribute to ecological pressure on the carbon footprint. Energy structure, the influence of international trade on GDP, and energy intensity exerted an inhibitory effect on the ecological pressure of the carbon footprint. The former had a greater effect than the latter, while the energy structure has less effect, and positive and negative fluctuations. The effect of overall per capita export of goods and services, is greater than the population intensity of productive land.
From the perspective of the seven sub-regions, the effects of various factors on the EP cefc are different, and the influence of each driving factor on different regions is also quite different. Energy intensity shows inhibition in all regions. During the period of 2010-2014, the inhibitory effects were as follows from small to large: Central Asia, Mongolia, Russia, South Asia, East Asia (China), Central and Eastern Europe, West Asia, North Africa and Southeast Asia. The energy structure had both inhibitory and promoting effects, but the effect was always small. Areas where it caused an inhibitory effect include Mongolia, Russia, Central Asia, West Asia, North Africa, and Central and Eastern Europe. Regions where it was a promoter include South Asia and East Asia. The impact of the Influence of international trade on GDP was generally inhibitory effect. The impact of the two factors of per capita exports of goods and services and the population intensity of productive land showed a promoting effect as a whole, and the effect of the former was greater than the latter. However, in the Mongolia-Russia region, the population intensity of productive land shows a slight increase.
This study mainly discusses issues regarding the ecological pressure of the carbon footprint in energy consumption, and other measurements of the ecological pressure of the carbon footprint from various angles, such as transportation and tourism. The conclusions of the paper can only be explained by the carbon emissions generated by energy consumption; however, this has its limitations. The regional division of land was mainly based on geography and not according to economic characteristics. The generality of the analysis conclusions is affected by this. Additionally, the productive land area conversion coefficient directly used the WWF's value to measure the carbon footprint in energy consumption without considering the difference in storage capacity of different types of productive land for carbon emissions.
Author Contributions: Q.S. established the research framework; Y.G. and F.M. jointly established the research model; C.W., B.W., X.W., and W.W. collected the data and carried out the result calculations; Q.S. provided the data acquisition channel; and Q.S. and Y.G. analyzed the results and wrote the paper together.
Funding: This study was financially supported by the National Social Science Foundation of China (grant number 17BJY139).