Understanding Emission Trends, Regional Distribution Differences, and Synergistic Emission Effects in the Transportation Sector in Terms of Social Factors and Energy Consumption

Abstract

China’s transportation sector plays a significant role in reducing carbon dioxide (CO₂) and air pollution. Previous studies have predominantly utilized scenario analysis to forecast emissions for the next 30 to 50 years based on coefficients from a base year. To elucidate the current state of gas emissions in the transportation sector, this study employed panel data for 10 types of gas emissions from 2001 to 2020, analyzing their emission characteristics, tendencies, and synergistic effects. Utilizing the Kaya equation and the logarithmic mean division index (LMDI) decomposition method, we developed a model of pollutant emissions that considers the synergistic effects, pollution emission intensity, energy mix, energy consumption intensity, and population. The results show that all pollutants in the transportation sector decreased except for NH₃ and CO₂. There was a synergistic effect between air pollutants and CO₂ emissions, but the reduction was not significant. From 2013 to 2020, the transportation sector shifted from a high emission intensity with low synergy to a low emission intensity with high synergy. The results indicate that off-road mobile vehicles, on-road diesel vehicles, and motorcycles became the main source of emissions from transportation in certain provinces, and a key area requiring attention in policy development. Gasoline consumption was identified as the primary contributor to the significant increase in synergistic emission variability in the transportation sector. These results provide policymakers with practical ways to optimize emission reduction pathways.

1. Introduction

The Sustainable Development Goals of Good Health and Well-being (SDG 3) and Climate Action (SDG 13) are relevant to sustainable development. The Intergovernmental Panel on Climate Change (IPCC) proposed a comprehensive framework for climate change, wherein socio–economic development has caused the accumulation of greenhouse gas and aerosol emissions, contributed to atmospheric pollution, reinforced the greenhouse effect, and altered the future climate. The transportation sector ranks third among China’s major sources of greenhouse gas emissions, following closely behind the power generation industry [1,2,3]. Therefore, to support and guarantee the attainment of the country’s 2030 emission reduction goal, the transportation industry will need to undertake the arduous task of emission reduction in the coming decades. At the same time, the use of large amounts of fuel by the transportation sector has brought about an increase in air pollutants such as PM, CO, NO_X, VOC, and SO₂.

Air pollution and CO₂ emissions share similar root sources [4,5], which has prompted concerns about how to reduce CO₂ emissions while controlling air pollution. This has become a hot topic of discussion. The China Center for Environmental Politics and Economic Policy Research (CCEEPR) provides a definition of the twofold beneficial effects of air pollution reduction and carbon emission reduction [6]. Firstly, in the process of controlling greenhouse gas emissions, other local pollutant emissions (e.g., SO₂, NO_X, CO, VOC, and PM) can be decreased. Secondly, in the process of controlling local pollutant emissions and carrying out ecological construction, CO₂ and other pollutant emissions can be reduced or absorbed. Consequently, the feasibility of reducing CO₂ emissions while alleviating air pollution and fostering sustainable development, the presence of synergistic emission effects in policy implementation and the factors influencing these effects, and strategies to enhance policy efficiency have attracted considerable academic interest. Given the significant health risks linked to PM_2.5 and NO₂ emissions from the transportation sector [7], the particularly urgent issue of PM_2.5 and O₃ emissions in China’s air pollution landscape [8], the role of NOx as a major precursor to atmospheric ozone and PM_2.5 [9,10], and the impact of VOCs (volatile organic compounds) on particulate matter abatement, we therefore extended the study to include CO₂, PM_2.5, PM₁₀, CO, SO₂, NO_X, and VOC in order to analyze the issue from an all-encompassing societal perspective.

This article is structured as follows: In Section 2, we present a review of the previous research. Section 3 details our research method and data sources. In Section 4, we summarize the results (Figure 1); Section 4.1 and Section 4.2 include an analysis of the pollution sources, emission trends, and LMDI decomposition results; Section 4.3 and Section 4.4 delve into the synergistic effect analysis and regional disparities. In Section 5, we discuss the homogeneity and heterogeneity factors, as well as propose some policy recommendations. Finally, Section 5 concludes by summarizing our findings.

2. Literature Review

2.1. Studies on Emission Trends and Synergistic Emission Effects

Dong’s [11] study filled a gap in the understanding of the synergistic effects of air pollution and carbon emission reductions across all provinces in China. It assessed provincial CO₂ emission reductions from an economic policy perspective and predicted synergistic reduction benefits. Li [12] used data from 18 manufacturing sectors and demonstrated, from the demand theory perspective, that CO₂ and SO₂ have a substitution effect. However, this study did not fully consider the potential impacts of future technological advancements and policy changes. Subsequent scholars began to assess the impact of one aspect, either CO₂ or air pollution, on the other. Yang [13] investigated the air quality benefits of China’s goal to peak carbon emissions by 2030, noting reductions in SO₂, PM_2.5, and NO_X emissions across agriculture, electricity, industry, transportation, and construction sectors. Lu [14] assessed the National Air Pollution Prevention and Control Action Plan’s impact on major pollutants in the Jing-Jin-Ji region, finding that PM_2.5 reduction had the most significant synergistic effect on CO₂ reduction in Beijing–Tianjin–Hebei. However, both studies may have overestimated policy impacts. According to Chen [15], the synergistic effect of carbon emissions reduction surpasses that of pollution reduction and disaggregates the air pollutants. They found that the synergistic emission reduction effect of SO₂ was greater than that of PM_2.5.

Recent discussions on synergistic mechanisms and social cost estimates have increased. Wang [16] examined the collaborative governance of climate change and air pollution, finding that reducing carbon emissions by 1000 tons in China’s industrial sector also reduced air pollutants by 1 ton. However, the study neglected the adaptability of governance mechanisms under varying policies. Yuan [17] examined China’s energy transition’s combined impact on air quality and carbon emissions. Using a 2025 scenario analysis, the study simulated pollutant concentration changes, emphasizing end-of-pipe control measures for pollutants while overlooking socio–economic costs. Shu [18] examined the influence of the “Three-Year Action Plan” in China’s “2+26” cities (Beijing, Tianjin, and 26 other prefecture-level cities in the Jing-Jin-Ji area) on reducing air pollution and mitigating climate change. The study found that nearly half of these counties increased their CO₂ emissions while reducing their air pollutant emissions. However, it may have overestimated policy impacts and overlooked economic, social, and technological changes. Shu’s [19] study investigated feasible pathways to achieve PM_2.5 air quality standards and their synergistic benefits for CO₂ emission reductions by 2030 in northern Chinese cities. Zha [20] developed a system to evaluate the efficiency of urban pollution and carbon emission reduction governance. The results showed a 38.07% increase in synergistic governance efficiency among Chinese cities from 2006 to 2020. However, the study did not examine the impact of specific industries on pollution and carbon emissions. Xian [3] identified synergistic effects between pollution and carbon reduction policies in China’s electricity, industry, transportation, residential, and agricultural sectors. The study showed that current emission reduction policies significantly reduced major air pollutant emissions and slowed CO₂ emission growth. These synergistic effects varied across sectors, but regional heterogeneity was not deeply explored. Xu et al. [21] demonstrated a clear and strengthening synergistic effect between CO₂ and air pollutant emissions in China’s industrial sector. Synergies between coal-related carbon emissions and air pollutants are increasing, whereas those between gas-related emissions and pollutants are declining. However, the study overlooked regional differences in industrial structure, energy consumption, and pollution control affecting these synergies.

2.2. Status of Research on Synergistic Effects in the Transportation Sector

Alimujiang [22] (2020) highlighted the advantages of electric private cars, taxis, and buses in reducing carbon dioxide and air pollutant emissions in Shanghai. However, the study did not thoroughly evaluate the economic costs of widespread electric vehicle adoption, including infrastructure construction and maintenance expenses. Jiao [2] quantitatively analyzed the synergistic effects of CO₂ and air pollutants in Guangzhou’s transportation sector, evaluating emissions across various urban transportation subsectors. However, the study inadequately addressed regional differences in transportation structure, energy consumption, and pollution control measures. Duan [23] assessed the traditional strategies of Chongqing’s road traffic industry, finding that changing transportation trends effectively reduced air pollutants and CO₂ emissions in the short term. The study highlighted the critical role of clean electricity production for electric vehicle development in the medium-to-long term. Guo [24] analyzed and forecasted energy consumption and emissions in the Beijing–Tianjin–Hebei region, concluding that optimizing the energy structure most effectively reduces emissions in the road traffic and aviation sectors. Dong [25] used the LEAP framework and advanced scenarios to estimate the impact of CO₂ emission reductions on air pollution in Zhengzhou’s road traffic. In Fu’s study [26], a framework for scenario analysis was constructed, showing that reducing air pollution and greenhouse gases through the development of compact cities can be achieved by 2050 with the widespread use of new energy vehicles and decarbonization of energy sources. These studies are all based on model-driven scenario analyses, which are beneficial in providing policymakers with potential outcomes under different policy choices, thereby facilitating more informed decision-making. However, this approach is overly reliant on model calculations and policy assumptions, neglecting inconsistencies with actual policy developments, thereby affecting the applicability of the analysis results. Therefore, we believe that modeling analysis based on historical data is essential.

2.3. Research Status Summary

The aforementioned studies generally show that efforts to reduce carbon emissions often also help reduce air pollution, and vice versa. However, the effects can vary depending on the region, industry, and type of pollutant or emission being studied. Almost all studies on synergistic emission reduction in the transportation sector rely heavily on calculations and assumptions, potentially overlooking real-world policy developments. This may affect the applicability of their results. Two studies have specifically focused on synergistic emissions reductions in the transportation sector and have not employed scenario analyses. A study [27] focused on the impact of carbon markets and environmental regulations on synergistic emission reductions in the transportation sector, with the disadvantage that it did not analyze the impact of factors such as energy mix. Another study [28] developed a co-benefit impact model for reducing carbon emissions and air pollutants. According to that study, the primary driving factors behind this synergistic effect are energy efficiency and industrial structure. Furthermore, transportation synergies are typically greater in areas with stricter environmental regulations. However, there are still some shortcomings. For example, while the most serious air pollution in China is PM and O₃ pollution [29], the dependent variable used in the model was only the emission reduction of SO₂, and other pollutants were not considered. They utilized the IPCC model to determine CO₂ emissions, which, as a “top-down” model, incorporates a multitude of uncertainties related to carbon emissions, as identified by Loo [30].

Therefore, our study utilized a “bottom-up” model database to calculate the synergistic effect between carbon dioxide and air pollutants in transportation. This study also incorporated the following seven factors that affect emissions into one model and quantified each factor: air pollution emissions that lead to carbon dioxide emissions, carbon dioxide emissions accompanied by air pollutant emissions, pollution emission intensity in the transportation sector, energy mix in the transportation sector, energy consumption intensity in the transportation sector, economic growth, and population. The innovation of this study is the inclusion of the transportation sector’s pollution emission intensity, an indicator that allows for the status of vehicle and fuel upgrades in 30 provinces and cities to be directly compared. The synergistic effects between CO₂ and air pollution are also quantified, which facilitates the analysis of regional differences. The above advances provide a reliable basis for policy implementation.

3. Materials and Methods

3.1. MEIC Calculation

Internationally, there are two ways to measure emissions: one is a top-down model that calculates only according to energy consumption, and the other is a bottom-up model that calculates according to the specifics of each sector, such as the number of vehicles in the transportation sector, the number of passengers, and other degrees of activity. Generally, the bottom-up calculation method is more accurate [12]. We chose the MEIC database (Multi-resolution Emission Inventory for China, http://www.meicmodel.org) as it relies on iterative updates to technology for the estimation of emissions.

The MEIC database is modeled as follows:

$$E_{\mathit{psk}}=A_{\mathit{ps}}\times {\sum }_t(X_{\mathit{pst}}\times {\mathit{EF}}_{\mathit{pskt}}\times {\sum }_a(C_{\mathit{psta}}\times \left(1-{\mu }_{\mathit{ka}}\right)\left)\right)$$

(1)

where p denotes the province, s denotes the emission source, k denotes the air pollutants or CO₂, t denotes the manufacturing technologies, and a denotes the air pollution control technologies. A represents the activity rate, X represents the proportion of a specific manufacturing technology, EF represents the unmitigated emissions, and μ represents the removal efficiency. The determination of the emission factors and the technical explanations have been detailed in two previously published articles [12,31].

3.2. Model Construction

Firstly, the Kaya equation [32] is widely used internationally to estimate CO₂ emissions. Its expression is as follows:

$${CO}_2={\displaystyle \frac{{CO}_2}{Energy}}\times {\displaystyle \frac{Energy}{GDP}}\times {\displaystyle \frac{GDP}{Pop}}\times Pop$$

(2)

As shown in Equation (2), CO₂ emissions are composed of four factors, listed from left to right in sequence: CO₂ per unit of energy consumption, energy consumption per unit of GDP, GDP per capita, and population ($Pop$,

Table 1. The definitions of the driving factors (for Equations (2)–(4)).

Variables	Description
AP	The concentration of air pollutants.
Uco2	CO2 emissions.
EAP	All pollutants’ emissions.
Egaso	Gasoline consumption.
Etotal	Energy consumption of the transportation sector.
G	GDP.
D, Pop	Population.
H	The concentration of pollutants per CO2 emission unit.
C	The CO2 emissions for each unit of pollutant released.
St	The proportion of gasoline used in the overall energy usage of the transportation sector.
So	The amount of energy used per GDP unit in transportation.
E	Per capita GDP.

). Through this equation, carbon emissions are linked to social development factors. This study further extends the application of the Kaya equation to air pollutant emissions.

3.3. Decomposition Analysis

The fundamental concept of exponential decomposition involves initially establishing a control function that connects the aggregate to be decomposed with specific predefined factors of interest. With the control function in place, a variety of decomposition techniques can be employed to measure the effects of these factors’ changes on the aggregate. This approach allows for a clear and systematic analysis of how different elements contribute to the overall aggregate, providing insights into their individual impacts [31]. Therefore, decomposition analysis can help to increase our understanding of the magnitude of the role of different driving factors. Broadly speaking, there are several ways to perform this: index decomposition analysis (IDA), structural decomposition analysis (SDA), arithmetic mean division index decomposition (AMDI), and logarithmic mean division index decomposition (LMDI).

While AMDI has a residual term [33], SDA is more data-demanding and requires input–output tables [34]. However, LMDI only requires time series data. LMDI was proposed by Ang in 1998 and derived from the IDA method, and it has been widely used in energy-related analysis. It has the advantage of having no residual terms and is easy to interpret [35], and it has been applied to the synergistic abatement of air pollution and carbon dioxide in pioneering research [28,36,37,38].

$$AP={\displaystyle \frac{AP}{U_{co2}}}\times {\displaystyle \frac{U_{co2}}{E_{AP}}}\times {\displaystyle \frac{E_{AP}}{E_{gaso}}}\times {\displaystyle \frac{E_{gaso}}{E_{total}}}\times {\displaystyle \frac{E_{total}}{G}}\times {\displaystyle \frac{G}{Pop}}\times Pop=H\times C\times P\times S_t\times S_o\times E\times D$$

(3)

where AP, U_co2, E_AP, E_gaso, E_total, G, and Pop represent the concentration of air pollutants, CO₂ emissions, all pollutants’ emissions, gasoline consumption, energy consumption, GDP, and population, respectively.

H refers to the concentration of pollutants per CO₂ emission unit, signifying how much CO₂ emissions add to the concentration of pollutants. H is the combined effect of CO₂ and air pollutants. C represents the CO₂ emissions for each unit of pollutant released, serving as a measure for the combined reduction of CO₂ and air pollutants.

H and C are regarded as the co-effect of CO₂ and air pollutants collectively. There are several prior studies that quantify the synergistic emission effects of pollutants and CO₂ in a similar way [14,39,40]. P is pollutant emission per unit of gasoline, indicating emission intensity, incorporating vehicle emissions management, fuel upgrades, and vehicle upgrades.

S_t represents the proportion of gasoline used in the overall energy usage of the transportation sector, reflecting the fuel configuration, which is linked to the enhancement of the fuel structure metric in transportation.

S_o refers to the amount of energy used per GDP unit in the transportation sector, reflecting the intensity of energy consumption in this sector.

E is the per capita GDP, signifying the economy’s growth. D symbolizes the population and is indicative of the impact of demographic shifts.

In this study, the variable o represents the starting year, and t represents the ending year. The symbols ∆H, ∆C, ∆S_o, ∆S_t, ∆E, ∆P, and ∆D denote the contribution rates of each factor between the starting year and the ending year, respectively. These rates, from left to right, correspond to the contribution rate of AP per unit of CO₂, the contribution rate of CO₂ per unit of AP, the contribution rate of the energy structure in the transportation sector, the contribution rate of emission intensity in the transportation sector, the contribution rate of per capita GDP, and the contribution rate of population size.

Thus, the LMDI decomposition formula can be expressed as in Equation (4).

Furthermore, the calculation steps for understanding the contribution of each factor for each province are shown below. A positive final decomposition result for a driving factor indicates its contribution to the overall increase or decrease in air pollutant emissions, while a negative result signifies that the factor drives a change opposite to the general trend. In this study, the LMDI analysis results at both the national and regional scales aggregate the contributions from the provincial level.

$$\begin{array}{l}∆AP={AP}^{\mathit{t}}-{AP}^o=∆H+∆C+∆P+∆S_o+∆S_t+∆E+∆D\\ ={\displaystyle \frac{{AP}^t-{AP}^0}{\ln {AP}^t-\ln A{P}^o}}\times \ln \left({\displaystyle \frac{H^t}{H^o}}\right)+{\displaystyle \frac{{AP}^t-{AP}^o}{\ln {AP}^t-\ln {AP}^o}}\times \ln \left({\displaystyle \frac{C^t}{C^o}}\right)+{\displaystyle \frac{{AP}^t-{AP}^0}{\ln {AP}^t-\ln A{P}^o}}\times \ln \left({\displaystyle \frac{P^t}{P^o}}\right)+{\displaystyle \frac{{AP}^t-{AP}^o}{\ln {AP}^t-\ln {AP}^o}}\times \ln \left({\displaystyle \frac{{S_o}^t}{{S_o}^o}}\right)\\ +{\displaystyle \frac{{AP}^t-{AP}^0}{\ln {AP}^t-\ln A{P}^o}}\times \ln \left({\displaystyle \frac{{S_t}^t}{{S_t}^o}}\right)+{\displaystyle \frac{{AP}^t-{AP}^o}{\ln {AP}^t-\ln {AP}^o}}\times \ln \left({\displaystyle \frac{E^t}{E^o}}\right)+{\displaystyle \frac{{AP}^t-{AP}^0}{\ln {AP}^t-\ln A{P}^o}}\times \ln \left({\displaystyle \frac{D^t}{D^o}}\right)\end{array}$$

(4)

3.4. Regional Variability Analysis

There are obvious inter-regional differences in the synergistic emission effects of VOC, NO_X, and CO₂, which may be due to a variety of factors. To assess and compare both inter- and intra-regional disparities, we selected the Theil index as our measurement tool. The Theil index is a special form of the generalized entropy (GE) index system [41]. It is a tool used to measure the level of inequality and is widely used in the analysis of regional environmental load, energy consumption, and other fields [42,43]. This index allows for the examination of pollution variation both between and within regions, enabling broad-scale evaluation of changes in pollution patterns [44]. The Theil index is usually a positive number between 0 and 1, but a negative Theil index is still valid [45].

Based on Shorrocks [46], this study improves the formula for calculating the Theil index from GDP per capita, population density, population size, the total energy consumption of the transportation sector, and the total gasoline consumption of the transportation sector.

$$T_{total}=\sum {\displaystyle \frac{y_m}{Y}}\times lg{\displaystyle \frac{\left(\frac{y_m}{Y}\right)}{\left(\frac{x_m}{X}\right)}}$$

(5)

$$T_{\mathrm{inter}}=\sum {\displaystyle \frac{y_j}{Y}}\times lg{\displaystyle \frac{\left(\frac{y_j}{Y}\right)}{\left(\frac{x_j}{X}\right)}}$$

(6)

$$T_{intra}=\sum {\displaystyle \frac{y_m}{Y}}\times lg{\displaystyle \frac{\left(\frac{y_m}{y_j}\right)}{\left(\frac{x_m}{y_j}\right)}}$$

(7)

In these equations, T_total represents the overall national Theil index and T_inter implies the inter-regional Theil index. T_intra indicates the intra-regional Theil index. y_m is the cumulative amount of synergistic emission reductions in a province, and y_j is the cumulative amount of synergistic emission reductions in a region. Y denotes cumulative, nationally coordinated emission reductions. x_m denotes GDP per capita, population density, population size, the total energy consumption of the transportation sector, and the total gasoline consumption of the transportation sector in a given province. X_j denotes the GDP per capita, population density, population size, the total energy consumption in the transportation sector, and the total gasoline consumption of the transportation sector of one region. X signifies the GDP per capita, population density, population size, the total energy consumption of the transportation sector, and the total gasoline consumption of the transportation sector across China.

3.5. Data

This study collected panel data from 30 Chinese provinces (excluding Tibet, Hong Kong, Macao, and Taiwan). Data on air pollutants, including SO₂, NO_X, CO, VOC, NH₃, PM₁₀, PM_2.5, black carbon (BC), organic carbon (OC), and CO₂, from 2001 to 2020 were taken from the MEIC database. Their emissions were estimated from 1800 source categories across 55 sectors and 42 fossil fuel types. Version 1.4 of the MEIC database was chosen for analysis in this study.

Population and gross domestic product (GDP) data were sourced from the China Statistical Yearbook. Private car ownership data were derived from the Transportation Statistics Yearbook. The total energy consumption and gasoline consumption data for the transportation sector were obtained from the China Energy Statistical Yearbook. Specifically, the total energy and gasoline consumption data for the transportation sector in the China Energy Statistical Yearbook 2020 were calculated after converting all types of energy consumption into standard coal equivalents, using the conversion tables provided in the China Energy Statistical Yearbook 2020. The specific conversion coefficients were as follows: crude oil at 1.4286 kg of standard coal per kilogram, fuel oil at 1.4286 kg of standard coal per kilogram, and gasoline and kerosene at 1.4714 kg of standard coal per kilogram. For individual years with missing data, we used linear interpolation based on the data from the nearest available years.

4. Results

4.1. Emission Trends

The MEIC database classifies the transportation sector into four categories, namely gasoline, diesel, motorcycles, and off-road vehicles, and these categories were used to determine the emissions that were counted. Since there was no significant change in the proportion of each category between 2001 and 2020, this study selected the year 2020 as an optimal example for analysis (Figure 2). The most significant emissions from gasoline, diesel, and off-road vehicles were CO₂, followed by CO and NO_X. In terms of an increasing or decreasing trend for each emission, the transportation sector showed a decreasing trend in all emissions except for CO₂ and NH₃ emissions, which were generally trending upwards (Figure 3). The average CO₂ emissions from the transportation sector were 20% from diesel vehicles, 35% from motorcycles, and 41% from off-road vehicles between 2001 and 2020. Motorcycle emissions accounted for most of the expansion between 2001 and 2020, while emissions from off-road vehicles and diesel vehicles gradually shrank after 2015.

From Table S1 (in the Supplementary Materials) and Figure 3, we can see that, in 2020, off-road vehicles produced 47–48% of PM (PM_2.5 and PM₁₀) emissions, while diesel vehicles produced 39% of PM emissions. Between 2001 and 2020, the percentage of off-road PM_2.5 and PM₁₀ emissions nearly doubled, from 35% to 63% and 36% to 64%, respectively. Regarding the 20-year average, OC emissions totaled 44% from off-road vehicles and 35% from diesel vehicles, while BC emissions totaled 51% from off-road vehicles and 20% from diesel vehicles. Meanwhile, off-road vehicles accounted for 61% and 69% of 2020 emissions for OC and BC, respectively. In 2001, the SO₂ emissions from diesel vehicles and off-road vehicles were 34% and 40%, respectively. In 2020, SO₂ emissions from diesel vehicles reached 18%, while SO₂ emissions from off-road vehicles increased to 63%. In summary, the proportion of off-road vehicles in the six emission categories of CO₂, PM_2.5, PM₁₀, OC, BC, and SO₂ has expanded, and the reduction of off-road vehicle emissions will become a priority in the field of transportation in the future. Based on the 20-year average, gasoline vehicles accounted for 77% of CO emissions, 69% of VOC emissions, and 68% of NH₃ emissions, respectively. This suggests that the management of gasoline vehicles and the continuous upgrading of fuels continue to play an important role in reducing emissions.

This study took the previous year’s data as the base year and calculated the annual rate of change in the data (Figure 4). The results showed a general downward trend in energy consumption in the transportation sector, in gasoline consumption, and in the growth rate of private vehicles. However, the growth rate of private vehicles is still positive, and the total volume is increasing. The growth rates of CO₂ and air pollutants emitted by the transportation sector are negative, which indicates a downward trend in emissions. In terms of the degree of change in the average annual rate, gasoline consumption and total energy consumption in the transportation sector changed dramatically between 2002 and 2008, then stabilized and subsequently showed a decreasing growth rate after 2009. This was due to the establishment of the Ministry of Transport of the People’s Republic of China (MOT) by the government in 2008. The MOT formulated two standards, namely “Fuel Consumption Limits and Measurement Methods for Operating Passenger Vehicles” and “Fuel Consumption Limits and Measurement Methods for Operating Freight Vehicles”. These policies enabled the implementation of the Energy Law of the People’s Republic of China, which was amended and put into effect in 2009. In 2013, China implemented the carbon emissions trading market and the Air Pollution Prevention and Control Action Plan (APAP). This included the strict control of motor vehicles, such as the release of the National IV and National V stages of automotive diesel standards (Table S2 in the Supplementary Materials), accelerating the elimination of yellow-labeled vehicles (vehicles classified as yellow-labeled include gasoline vehicles failing to comply with National I emission criteria and diesel vehicles with emissions falling short of National III standards) and old vehicles. In 2013, seven provinces and 10 cities introduced subsidies for the elimination of yellow-labeled vehicles [47]. As a result, the transportation sector saw a significant reduction in energy consumption, gasoline consumption, and CO₂ emissions. Therefore, when comparing the synergistic effect as well as changes in emission intensity in Section 4.3, this paper chose to compare 2013 and 2020.

4.2. Characterization of Pollutant Emissions from the Transportation Sector

This study calculated the cumulative effect of each factor on pollutant emissions from 2002 to 2020 (

Table 2. Impact on pollutant emissions.

	VOC	NOX	PM2.5	PM10	CO	SO2
H	−0.286	−0.285	−0.044	−0.044	−2.845	−0.009
C	5246.998	2099.827	52,246.265	50,683.282	1055.803	198,666.364
P	−36,091.316	−32,005.966	−3900.931	−3919.003	−251,946.874	−644.893
St	−3.303	−3.303	−3.303	−3.303	−3.303	−3.303
So	4.174	4.174	4.174	4.174	4.174	4.174
E	8.440	8.440	8.440	8.440	8.440	8.440
D	14,331.660	14,331.660	14,331.660	14,331.660	14,331.660	14,331.660

Note: H = ${\displaystyle \frac{AP}{U_{co2}}}$, C = ${\displaystyle \frac{U_{co2}}{E_{AP}}}$, P = ${\displaystyle \frac{E_{AP}}{E_{gaso}}}$, St = ${\displaystyle \frac{E_{gaso}}{E_{total}}}$, So = ${\displaystyle \frac{E_{total}}{G}}$, E = ${\displaystyle \frac{G}{Pop}}$, D = $Pop$.

). H indicates that, for every additional ton of CO₂ emissions, the emissions were reduced by 0.286 tons for VOC, by 0.285 tons for NO_X, by 0.044 tons for PM_2.5, by 0.044 tons for PM₁₀, by 2.845 tons for CO, and by 0.009 tons for SO₂. C indicates that each increase of 1 ton for VOC, NO_X, PM_2.5, PM₁₀, CO, and SO₂ emissions resulted in 5246.998 tons, 2099.827 tons, 52,246.265 tons, 50,683.282 tons, 1055.803 tons, and 198,666.364 tons of CO₂ emissions, respectively. H+C is the synergistic effect of pollutants with CO₂, and there is a significant positive value for the ratio of VOC, PM_2.5, SO₂, and CO₂ emissions. This suggests a significant synergistic effect of CO₂ and air pollution in the transportation sector. P indicates pollutant emissions per unit of gasoline, which gives the emissions intensity and represents the outcome of vehicle emissions management, fuel upgrades, and vehicle upgrades. VOC emissions decreased by 36,091.316 tons, NO_X by 32,005.966 tons, PM_2.5 by 3900.931 tons, PM₁₀ by 39,190.003 tons, CO by 251,946.874 tons, and SO₂ by 644.893 tons between 2020 and 2020. This shows that vehicle management, as well as upgrades for the transportation sector, can help to reduce pollutant emissions. St indicates the share of gasoline in all energy consumption by the transportation sector, which suggests an energy mix in the transportation sector. Changes in the energy mix have served to reduce air pollutants over the last 20 years; however, its reduction value is smaller than that of P (Rows S_t and P in Table 2). So refers to energy consumption in the transportation sector per unit of GDP, indicating the energy consumption intensity in the transportation sector, a factor that contributes to pollutant emissions. E indicates that the cumulative effect of GDP per capita on the emission of each pollutant is 8.44, which is significantly larger than that of GDP. Finally, D indicates that the cumulative effect of population size on the emission of each pollutant is 14,331.66, which is larger than that of the GDP, indicating that the size of the population results in larger pollutant emissions.

4.3. Distribution of Synergistic Effects and Changes

This study analyzed the seven factors listed in Table 2 that affect pollutant emissions. Three of the factors, H, C, and P, demonstrate different regional differences depending on the pollutants. As explained previously, P denotes air pollutants emitted per unit of gasoline and refers to the emission intensity of the transportation sector. Figure 5 displays the emission intensity on the horizontal axis and the synergistic effect of the pollutant with CO₂ on the vertical axis, while the red dots mark the average of the 30 provinces and cities. Xinjiang is not included in the figure because its emission intensity is much higher than the national average. The figure divides the 29 provinces and cities into four quadrants, with the first quadrant indicating a high emission intensity in the transportation sector and a high synergistic effect. The second quadrant indicates that the transportation sector has a low emission intensity and a high synergistic effect. The third quadrant indicates that the transportation sector has a low emission intensity and a low synergistic effect. Quadrant 4 indicates that the transportation sector has a high emission intensity and a low synergistic effect. The overall view of the scatter plot distribution shows a tendency for provinces and cities to move from the third and fourth quadrants (2013) to the second quadrant (2020). The red dots on the graph (representing the averages) clearly exhibit a trend towards the upper left (2020), departing from the lower left (2013). In terms of changes in the distribution of specific provinces and cities, Beijing, Tianjin, and Shanghai clearly fit the overall trend and were expected to move to the fourth quadrant in 2020 (except for Tianjin’s CO). In contrast, Hebei showed a countertrend movement; that is, it moved from the third quadrant or from near the third quadrant to the fourth quadrant. In 2020, Beijing showed a high synergy of air pollutants and CO₂ emissions, while Hebei became the province with the highest pollutant emissions per unit of gasoline in the country. It is also possible to note that Shandong and Liaoning appeared to move in opposite directions. We compared the emission sources of these four provinces and cities and found that Shandong, Liaoning, and Hebei have different pollutant emission structures (Figure 6).

To be specific, more than 60% of PM_2.5 and PM₁₀ emissions originated from off-road traffic, more than 50% of NO_X emissions originated from diesel vehicles, more than 70% of VOC emissions originated from diesel vehicles, more than 60% of SO₂ emissions originated from off-road vehicles, and 70% of CO₂ emissions came from motorcycles and off-road vehicles in Shandong, Liaoning, and Hebei. In addition, 63% of Beijing’s transportation sector’s CO₂ emissions originated from gasoline vehicles. In short, off-road vehicles, on-road diesel vehicles, and motorcycles were the main sources of emissions in these cities. Recently, Hebei, for the first time since the promulgation of the Hebei Provincial Motor Vehicle Pollution Prevention and Control Methods in 2012, introduced motor vehicle-related regulations on 10 April 2020 [48], namely the “Hebei Provincial Motor Vehicle and Non-Road Vehicle Mobile Machinery Emission Pollution Management Regulations” and the “Hebei Provincial Motor Vehicle and Non-Road Vehicle Mobile Machinery Emission Pollution Management Regulations”. Hebei province banned the use of National III and non-National III diesel vehicles at the end of 2020 [49].

4.4. Differential Analysis of the Seven Regions

After analyzing the impact of various factors on pollutant emissions, it was found that there were significant regional differences in synergistic emission reductions. This section focuses on analyzing the reasons for the formation of regional variability.

Based on the Kaya equation and Equation (3) in this study, five factors were selected: population, population density, GDP per capita, total gasoline consumption of the transportation sector, and the total energy consumption of the transportation sector. The inequality index for each region was measured using the Theil index (Table S4 in the Supplementary Materials). In Figure 6, VOC-Pop represents the Theil index of VOC and CO₂ synergistic emission reductions based on the population size, and VOC-Ga represents the Theil index of VOC and CO₂ synergistic emission reductions based on the gasoline consumed in the transportation sector. The results of the calculations show that gasoline consumption in the transportation sector in each region had the highest Theil index. This means that, at this stage, gasoline consumption is the main cause of regional differences in synergistic emissions in the transportation sector in China.

Figure 7 shows the Theil index for each pollutant, calculated based on population and gasoline consumption, except for VOC and NO_X, for which there are large differences in the Thiel index. The inequality indices for the other pollutants showed a smaller difference and even a higher population-based Theil index than gasoline consumption-based Thiel indices between 2001 and 2006. In summary, the inequality indices for the five pollutants combined showed that gasoline consumption in the transportation sector created greater inequality in synergistic emission reductions between regions. Moreover, the study calculated that the gap in synergistic emissions due to gasoline consumption was mainly caused by inter-regional differences (Table S4b in the Supplementary Materials), with intra-regional differences remaining above and below zero—a much smaller value than for the inter-regional Thiel index.

5. Discussion and Policy Recommendations

Seven pollutant emission impact factors were obtained using the Kaya equation and the LMDI method. These factors are listed in descending order according to the magnitude of their impact on reducing pollutant emissions, as follows: pollutant emission intensity in the transportation sector, energy mix in the transportation sector, air pollution accompanied by carbon dioxide emissions, energy consumption intensity in the transportation sector, GDP per capita, total population, and carbon dioxide emissions accompanied by air pollution (Table 2). Based on the results of the above analysis, the following recommendations are made, considering the current transportation situation in China.

5.1. Homogeneity Factor Analysis

Of the seven factors listed above, the sum of air pollution accompanying CO₂ emissions (H in Table 2) and the sum of CO₂ emissions accompanying air pollution (C in Table 2) are thought to have a synergistic effect. Four of the five remaining components (S_t, S_o, E, and D in Table 2) are consistent across regions. In addition to the transportation energy mix (S_t in Table 2), which reduces emissions, the other three factors, energy consumption in the transportation sector (S_o in Table 2), GDP per capita (E in Table 2), and population (D in Table 2), all contribute to emissions. Over the past 20 years, the impact of total population size on transportation pollutant emissions has been significantly larger than that of per capita GDP, the energy mix of the transportation sector, and the intensity of energy consumption in the transportation sector. At the same time, improvements in the transportation system [50] will increase the population of an area. At present, China is still in a period of continuous development of transportation networks. In 2021, the Chinese government formulated an ambitious “14th Five-Year Plan (2021–2025)” for the development of a modern comprehensive transportation system, setting the goals of doubling the mileage of high-speed railroads in operation, covering more than 95% of cities with a population of more than one million, covering more than 98% of cities with a population of more than 200,000 with expressways, and covering approximately 92% of prefecture-level cities using civil transportation airports. Along with the expanding transportation network, the pressure on pollution control and CO₂ emission reductions during the construction and operation of transportation infrastructure has also increased.

Previous research has indicated that the high concentration of the population in urban areas [15] can lead to a decrease in the overall cost of natural monopolies, such as electricity, gas, liquefied petroleum gas, natural gas, and public transportation, thereby encouraging residents to use more clean energy and public transportation services, which, in turn, lowers pollutant emissions and enhances the air quality. However, the current urban population density in China is decreasing [51]. Similarly, CO₂ shows a nonlinear relationship with population density [52], and the effect of carbon emission reduction resulting from increasing urban density is greatest in medium-sized cities (with a population of more than 500,000 and less than 1 million). An increase in urban density to a medium level provides the most significant reduction in carbon emissions, followed by large cities with a population below 5 million. In contrast, megacities with over 5 million inhabitants fail to demonstrate a significant carbon reduction effect due to urban density. Therefore, policymakers should formulate plans to reduce the demand for transportation and thus reduce emissions, considering the different demographics of different provinces and cities.

5.2. Heterogeneity Factor Analysis

Of the seven pollutant emission impact factors obtained from the Kaya equation and the LMDI model, the synergistic factor H+C and the emission intensity of the transportation sector are heterogeneous, and their analysis is presented in Section 5.1. of this article. Section 5.2. attempts to analyze the reasons for the existence of heterogeneity. Based on the above analysis, the following observation was made: At this stage, there is a synergistic effect between CO₂ and air pollution in China’s transportation sector, with Beijing having the best synergistic effect. Figure 6 shows that Beijing’s PM and SO₂ emissions are mainly from off-road transportation (including aviation and railroads), while NO_X, VOC, and CO₂ emissions are mainly from gasoline and diesel vehicles. Clean energy substitutions for road vehicles and aviation, as well as the electrification of railroads, offer the best potential for synergistic emission reductions [53].

This study divided the provinces and cities in China into seven regions (Figure 8), and the synergistic effect is evident in all regions. However, there is a difference in the distribution of the effect, which is mainly due to the difference in the total amount of gasoline consumed in different regions. After comparing the emission sources in the seven regions (Figure 9), it was found that particulate matter emissions from the transportation sector in southern China are mainly from off-road vehicles and diesel vehicles, while more than 50% of the particulate matter emissions in the other six regions of China are from off-road vehicles. This suggests that the management of diesel vehicles, emission monitoring, and energy-upgrading measures need to be strengthened in the three southern provinces of China, namely Guangdong, Guangxi, and Hainan. A case study made a similar suggestion for the transportation sector in Guangdong province [54]. It was emphasized that the most effective approach to achieving long-term synergistic benefits is to optimize the energy mix and replace fossil fuels with electricity, hybrids, and other environmentally friendly energy sources. This approach will lead to significant reductions in both air pollutant emissions and CO₂ emissions.

NO_X emissions in Central China are mainly from off-road vehicles and diesel vehicles (Figure 9); this is very different from other regions, where more than half of the emissions are from diesel vehicles. This suggests that Henan, Hubei, and Hunan need to strengthen the use of gasoline for off-road vehicles and diesel vehicles. Similar results were obtained in a previous study [55] that analyzed emissions in Henan province and identified CO, NO_X, VOC, and CO₂ emissions (compared to Central China in Figure 9) as originating from on-road vehicles and SO₂ emissions as originating from off-road vehicles. Areas with significant levels of NO_X and VOC emissions are typically located near railroad trunk lines and industrial centers that have substantial transportation requirements. Improving transportation systems may result in a synergistic reduction in emissions. It is essential to prioritize the promotion of clean energy and the use of new energy vehicles in the long term.

5.3. Limitations and Future Research

This study used carbon emission and air pollution data from the MEIC database to maintain consistency between the datasets; therefore, the analysis could only evaluate at the provincial level and failed to consider the specifics of transportation at the municipal level.

The results show that pollutant emissions from the transportation sector in China have generally decreased, whereas CO₂ emissions continue to increase. This study examined the synergistic effect between pollutant emissions and CO₂, using pollutant emissions as the disaggregation target, and concluded that population size is a determining factor. However, this study did not examine the synergistic effect of carbon dioxide as a decomposition target.

This study employed a commonly utilized geospatial clustering method to categorize regions into seven major areas. However, pollution emissions from the transportation sector are closely intertwined with the patterns of inter-regional movement of people and goods, the optimization and improvement of transportation networks, and varying transportation structures across different regions, such as the development of public transportation [56] and the growth of electric vehicles [57]. These aspects present a broad horizon for future research endeavors.

6. Conclusions

This study analyzed whether there are synergistic effects of carbon dioxide and air pollutants in the transportation sector and investigated the driving factors behind pollutant emissions, the role played by each factor, and whether there are regional differences in synergistic effects. The main findings are as follows:

• Air pollutants emitted by the transportation sector in China display a downward trend. Total carbon dioxide emissions, on the other hand, have continued to increase. The proportion of emissions from off-road vehicles is also increasing.
• Over the past two decades, a synergistic relationship has emerged between CO₂ emissions and air pollution in China’s transportation sector. Optimizing this sector’s energy mix can significantly lower air pollution, as reflected in emission intensity, which measures the impact of vehicle emission management, fuel quality improvement, and vehicle technology upgrades. Additionally, transportation emissions driven by population growth now exceed those linked to per capita GDP.
• From 2013 to 2020, the transportation sector in China reduced its emission intensity and increased its synergy, with Beijing showcasing the highest synergistic effect. Shandong, Liaoning, and Hebei exhibited varying trends, with 70% of their CO₂ emissions stemming from motorcycles and off-road vehicles compared to 63% of CO₂ emissions originating from gasoline-powered vehicles in Beijing. Regional differences in effect distribution are attributed to variations in gasoline consumption. In southern China, particulate matter emissions primarily originate from off-road and diesel vehicles, while in other regions, off-road vehicles account for over 50% of emissions.
• Henan, Hubei, and Hunan should enhance their management of off-road and diesel vehicles, while Guangdong, Guangxi, and Hainan should focus on controlling diesel vehicle emissions. Policymakers should create development plans according to the population sizes of provinces and cities in order to decrease transportation demand and reduce pollutant emissions.