Next Article in Journal
Design of Step Voltage Regulator Based on IGBT
Previous Article in Journal
LCL Trap Filter Analysis with a PFC Isolated Ćuk Converter Using SiC MOSFET for DCM
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Influencing Factors and Their Spatial–Temporal Heterogeneity of Urban Transport Carbon Emissions in China

1
School of Transportation Engineering, Chang’an University, Xi’an 710064, China
2
School of Economics and Management, Chang’an University, Xi’an 710064, China
*
Author to whom correspondence should be addressed.
Energies 2024, 17(3), 756; https://doi.org/10.3390/en17030756
Submission received: 3 January 2024 / Revised: 31 January 2024 / Accepted: 2 February 2024 / Published: 5 February 2024
(This article belongs to the Section B3: Carbon Emission and Utilization)

Abstract

:
Based on the panel data of China’s 284 prefecture-level cities from 2006 to 2020, this study employs spatial econometric and geographically weighted regression models to systematically analyze the influencing factors and their spatial–temporal heterogeneity of urban transport carbon emissions. The findings reveal the following: (1) GDP per capita, population, urban road area, and private car per capita are important factors causing the increase in urban transport carbon emissions, while the improvement of urban density, public transportation effectiveness, and government environmental protection can mitigate emissions and promote low-carbon development in urban transportation. (2) The worsening impact of GDP per capita on urban transport carbon emissions shows a decreasing trend over time, forming a spatial gradient pattern of gradually increasing from southwest to northeast. However, a similar effect of population increase during the research period, which currently displays an increasing spatial differentiation from north to south in sequence. (3) As another key deteriorating urban transport carbon emission, the influencing degree of private car per capita has gradually decreased from 2006 to 2020 and represented certain spatial gradient patterns. (4) Although the urban road area is favorable to urban transport carbon reduction in the early stage, it gradually begins to change in an unfavorable direction. The urban density is the contrary, i.e., the increase in that begins to play a positive role in promoting the development of low-carbon transportation among more cities. In addition, the influence coefficient of the former also presents an increasing distribution characteristic from south to north. (5) The reduction effect of public transportation effectiveness and government environmental protection on transport carbon emissions are both gradually prominent, where the former also shows space inertia of “increasing gradient from north to south and from north to northeast”.

1. Introduction

In recent years, climate warming, sea level rise, and extreme climate events, caused by the continuous deterioration of global carbon emissions, have seriously threatened the sustainable development of national economies and the survival of human beings. This issue has attracted wide attention from all countries. As the basic facilitator of economic and social production activities, transportation has become the third-largest industry emitting carbon, following electric power and industrial production [1]. According to the International Energy Agency (IEA), the carbon emissions from the transportation industry accounted for as much as a quarter of the world’s total carbon emissions in 2020, with China’s transportation contributing about 10% of that total. It is evident that transportation carbon emissions are expected to continue worsening in the foreseeable future, in line with the economic and social development of various countries. In response to this, the 20th National Congress of the Communist Party of China made it clear that it is imperative to “actively and steadily promote carbon peak neutrality, promote clean, low-carbon, and efficient energy use, and foster clean and low-carbon transformation in various sectors, including industry, construction, and transportation.” Therefore, the swift and effective reduction in transportation carbon emissions has become an important issue that urgently needs to be addressed.
Obviously, one of the strategic directions to solve this problem is to identify the key factors that drive or affect transport carbon emissions. In this regard, numerous previous scholars have conducted extensive analysis and research. Using the LMDI decomposition method, Lee et al. [2] found that the main factors affecting carbon emissions in the Asia–Pacific region were population growth and per capita GDP. Xu et al. [3] used a non-parametric additive regression model to analyze the key influencing factors of CO2 emissions in China’s transportation industry and found that economic growth has a nonlinear impact on transportation CO2 emissions, which conforms to the hypothesis of the environmental Kuznets curve (EKC) developed by Grossman and Kruger [4]. Bai et al. [5] concluded that the reduction in energy intensity and transportation intensity are the key factors in slowing down transportation carbon emissions. However, the increase in per capita wealth is a decisive factor in the growth of transport carbon emissions. Zhu et al. [6] also drew a similar conclusion to Bai et al. [5], but also found that population size significantly increased the road transport carbon emissions, and energy intensity and transport intensity had different driving effects in different areas. For transport carbon emissions in Pakistan, Rasool et al. [7] found that rising oil prices and economic growth could reduce transport CO2 during 1971–2014, while increasing energy intensity, population, and road infrastructure worsen CO2 emissions. The empirical study of Amin et al. [8] shows that the increase in renewable energy consumption effectively reduces transportation carbon emissions. Based on the two-level econometric model, Wang et al. [9] found that transportation carbon emissions are mainly affected by factors such as economic development, transportation structure, energy efficiency of transportation equipment, transportation organization, and infrastructure density. Zhao et al. [10] explored the impact factors of traffic carbon emissions in China’s central regions through the geographical detector method and found that the impact factors of transportation carbon emissions are the gross regional product and the second industry gross domestic product. Based on the data from China’s Guangdong Province from 2001 to 2020, Tang et al. [11] adopted the Tapio decoupling model and the logarithmic mean differentiation index decomposition method and found that the effect of income and urbanization was the main factor promoting the increase in carbon emissions, while the effect of energy intensity was the main factor reducing CO2.
In addition, numerous scholars have discussed the influencing factors of transportation carbon emissions from the perspective of efficiency and its decoupling from the economy. For example, Cui et al. [12] utilized a Tobit regression model to identify the important influencing factors of transportation carbon efficiency and found that the influence of structural factors is relatively small when compared to the technical and management factors. Jiang et al. [13] analyzed the key driving factors for the decoupling of transport-related CO2 and transport turnover, revealing that transport energy efficiency has the most significant effect in accelerating the decoupling between them, while the effect of energy structure is not conducive to the development of decoupling. However, it is regrettable that these works do not pay attention to the differences in influencing factors between regions and over time. Nowadays, a few current studies have also paid attention to this aspect. To the best of our knowledge, using the LMDI model, Zhu et al. [14] analyzed and compared the impact degree of transportation carbon emissions in different regions from the perspectives of economy, population, energy intensity, and industrial scale based on the data of Chinese provinces from 1997 to 2017. Zhang [15] found that the main factors affecting CO2 emissions in the transport sector are the same from the perspective of time and space. Xu and Xu [16] empirically found that the intensity of transportation activities, urbanization, technology, industrial structure, and per capita GDP are important factors affecting the CO2 of the transport industry in various provinces in China. Focusing on urbanization, Lv et al. [17] constructed the geographically weighted regression model (GWR)—the stochastic impact by regression on population, affluence, and technology (STIRPAT) model—and empirically found that urbanization only had a significant positive impact on the carbon emissions of road and air transportation in some provinces. However, such studies generally focus on the provincial level, lacking research focusing on the influencing factors of urban full-caliber transportation carbon emissions.
In general, previous studies have extensively discussed the influencing factors of transportation carbon emissions. However, empirical analysis in China has primarily focused on the national and provincial levels. Only a small number of scholars have explored the Beijing–Tianjin–Hebei city clusters [18,19]. Additionally, there is a lack of literature that regards cities as research objects. Nevertheless, China’s cities are crucial administrative units for the proposal and implementation of numerous energy conservation and emission reduction policies. The main source of transportation carbon emissions is the energy consumption of urban private cars and urban freight. Consequently, the implementation of urban transport carbon reduction work is generally considered as the key battleground for “carbon reduction.” Furthermore, there are close spatial linkages and significant spatial heterogeneity among different regions in China, which may lead to spatial differences in the influencing factors of urban transport carbon emissions. Additionally, economic and social development is a dynamic process, so the influencing degree of various factors on traffic carbon emissions may also have a time evolution law. However, previous literature has paid little attention to the spatio-temporal heterogeneity of influencing factors for urban transport carbon emissions.
In this study, the spatial econometric model (SDM) and the GWR model are utilized to examine the influencing factors of carbon emissions related to urban transport and their spatio-temporal heterogeneity. The analysis focuses on 284 prefecture-level cities in China, with the objective of providing a theoretical foundation for the government to develop targeted emission reduction policies and low-carbon development strategies for industries.

2. Research Methods and Data Sources

2.1. Research Methods

2.1.1. Spatial Econometric Model

Transportation plays a crucial role in connecting the economic and trade flows of various cities, creating a close spatial correlation between urban transport carbon emissions. Traditional ordinary least squares (OLS) only provide an “average” or “global” estimation of parameters, failing to reflect the spatial instability of parameters in different regions [20]. Hence, this paper employs a spatial econometric model for empirical research. Spatial econometric models commonly used include the SDM, spatial lag model (SLM), and spatial error model (SEM). The SDM describes both spatial substantial correlation and spatial disturbance correlation, combining the features of both SLM and SEM. Consequently, this model is adopted for analysis in this paper, and its specific formula is as follows:
Y i = α w i j Y i + α k X i , k + β k w i j X i , k + μ i ,   μ i = γ w i j μ i + ε
where i and j are cities; Y i denotes the urban transport carbon emissions; X i , k is the k t h influencing factor; w i j is the spatial weight matrix; w i j Y i and w i j X i , k are the spatial lag term of the Y i and X i , k respectively, and w i j μ i denotes the spatial error effect; α represents the spatial autoregressive coefficient, and γ is the spatial error coefficient; α k and β k denote the corresponding estimated coefficients; μ and ε are both the random error term. Moreover, Model (1) is SLM when α 0 ,   β 0 ,   γ = 0 , which mainly discusses whether each variable has a d iffusion spillover effect in a certain region. When α = 0 ,   β 0 ,   γ = 0 , Model (1) is SEM, indicating that the interregional spillover is the result of random impact.

2.1.2. GWR Model

Although SDM can identify the impact of various factors on urban transport carbon emissions when considering spatial factors, it is unable to identify the spatial–temporal heterogeneity of the influencing factors among cities. Therefore, this paper further constructs a GWR model for in-depth analysis. The specific formula for the GWR model is as follows:
Y i = β 0 ( u i , v i ) + k β k ( u i , v i ) X i k + ε i ;   β ( u i , v i ) = X T w ( u i , v i ) X 1 X T w ( u i , v i ) Y
where ( u i , v i ) is the spatial latitude and longitude coordinates of city i; β 0 is the fixed effect intercept at the position; X i k is the influencing factor, while β k is the corresponding regression coefficient; ε i is the random error term; w ( u i , v i ) is the spatial weight matrix, and its specific formula is as follows:
w i j = exp d i j b 2
where b denotes the bandwidth and signifies the attenuation coefficient accounting for distance and weight. The Akaike Information Criterion correction (AICc) is employed to prevent overfitting of the model, consistent with previous research [21], thereby determining the optimal bandwidth.

2.2. Variable Selection and Data Source

Referencing prior studies [3,6,11,22,23,24,25], and considering the current state of the transport industry, this paper selects GDP per capita (PGDP), population (POP), private cars per capita (PRC), urban road area (URA), urban density (URD), public transportation effectiveness (PUB), and government environmental protection (GEP) as the key factors affecting urban transport carbon emissions. The definition or measurement of each index is displayed in Table 1.
Considering the availability of data, this paper selects data from 284 prefecture-level cities in China from 2006 to 2020 as research samples. For the data source, the urban transportation carbon emission data referenced in this paper are obtained from the website of MEICModel (http://meicmodel.org.cn/, accessed on 13 October 2023) developed by the Department of Earth System Science at Tsinghua University, which provides high-resolution mapping of transportation carbon emissions with the space resolution of 0.25° × 0.25°, including the CO2 from road, rail, water, and air transport. This is a full range of transport emissions that includes carbon emissions from private cars, which have rarely been included in previous studies. The relevant data involved in other variables are all derived from the China City Statistical Yearbook and EPS database. Table 2 lists the descriptive statistics of all indicators.
Additionally, this paper conducts a multicollinearity test for the influencing factors, the results of which can be seen in Table A1 of the Appendix A. As can be seen from Table A1, the tolerance of each factor is greater than 0.1, and the VIF value is all less than 10, indicating that there is no multicollinearity in the selection of variables adopted in this paper. Meanwhile, to eliminate the impact of dimension on empirical results, all variables are treated logarithmically.

3. Spatial Econometric Analysis of Influencing Factors of Urban Transport Carbon Emissions

3.1. Model Selection

Urban transport emissions exhibit spatial correlation characteristics, and the necessity of considering spatial effects in empirical research should first be validated through the application of Moran’s I index. By calculating Moran’s I index for urban transport carbon emissions from 2006 to 2020, the results are outlined in Table 3. It is evident that Moran’s I has consistently passed the 1% significance level test for each year, signifying a substantial spatial correlation among urban transport carbon emissions in China, thereby manifesting a noteworthy positive spillover effect and an aggregation trend. Consequently, the use of a spatial econometric model in the analysis is warranted.
To identify a suitable spatial measurement model, the Lagrange multiplier (LM) test and likelihood ratio (LR) test are sequentially performed, and the results are presented in Table 4. It is evident from Table 4 that both the LM test and robust LM test reject the null hypothesis at a 1% significance level, signifying the simultaneous existence of substantial spatial correlation and spatial disturbance. This implies that the SDM may be superior to the SEM and SLM for the analysis of influencing factors of urban transport carbon emissions in this paper. Moreover, the LR test also rejects the null hypothesis at a 1% significance level, indicating that the SDM model will not degrade into the SEM and SLM models, hence establishing the superiority of the SDM model over the SEM and SLM models. Additionally, the Hausman test passes the 1% significance level, validating the adoption of the SDM fixed-effect model for the subsequent analysis.

3.2. Analysis of Empirical Results

Based on Model (1), this paper utilizes the SDM fixed effect model to empirically examine the impact of various factors on urban transport carbon emissions, as detailed in Table 5. For the purpose of comparison, conventional OLS panel regression is then conducted. As depicted in Table 5, the R2 of the OLS panel regression model is 0.611, while that of the SDM model is 0.618. This demonstrates a significant improvement compared to the results of OLS, indicating the superiority of the SDM model over traditional OLS. Furthermore, the spatial autocorrelation coefficient (rho) is significantly positive, signifying the positive impact of urban transport carbon emissions in neighboring cities on those in local cities and positive spatial spillover effects in transport carbon emissions between cities.
(1) GDP per capita (PGDP): GDP per capita exhibits a significant positive correlation with urban transport carbon emissions, as evidenced by the corresponding regression coefficient of 0.356, surpassing the 1% significance level. This finding aligns with the research outcomes of Wang et al. [26] and Zhu et al. [14], reflecting that economic development is the primary driver of the escalation in urban transport carbon emissions. The ongoing economic growth contributes to increased tourism and visitation demands, thereby accelerating the flow of goods and people, consequently leading to a corresponding exacerbation in urban transport carbon emissions. Additionally, the negative coefficient of W*LNPGDP (−0.326) suggests that the GDP per capita of neighboring cities has a mitigating effect on local transport carbon emissions. This phenomenon may be attributed to the “connected vessels” effect, wherein neighboring areas attract local investment and population, resulting in a reduction of local urban transport carbon emissions.
(2) Population (POP): A 1% increase in POP results in a corresponding 0.333% increase in urban transport carbon emissions, underscoring the significance of POP as a factor contributing to the escalation of urban transport carbon emissions. Clearly, population growth leads to a heightened need for transportation; consequently, yielding increases carbon emissions. Simultaneously, the increase in POP in neighboring cities significantly reduces the transport carbon emissions of local cities, as demonstrated by the significant negative coefficient of W*LNPOP (−0.603). The reason for this phenomenon is that there is a significant competitive relationship between Chinese cities, and some cities have even introduced talent attraction policies to accelerate urban population agglomeration. This will obviously cause the population to gather in the surrounding cities through the “siphon effect”, showing the increasing population of the surrounding cities significantly, while the local population decreases, and then the carbon emissions generate decrease.
(3) Private cars per capita (PRC): Table 5 demonstrates that the coefficients of LNPRC and W*LNPRC are both significantly positive, indicating that private cars in local and surrounding cities both exacerbate transport carbon emissions in the local area. This is mainly due to private cars being predominantly fueled by gasoline and diesel during the research period.
(4) Urban road area (URA): According to the coefficients of LNURA and W*LNURA (0.121 and 0.0784) in Table 5, which pass the significance test at the 1% level, urban road area exhibits a positive correlation with urban transport carbon emissions. In other words, the enhancement of local and surrounding cities’ infrastructure exacerbates the escalation of local transport carbon emissions. This may be attributed to the construction of infrastructure stimulating the aggrandization of the transportation demand, consequently resulting in worsening carbon emissions.
(5) Urban density (URD): Table 5 shows that the coefficients of LNURD and W*LNURD are −0.0403 and −0.023, respectively, indicating that urban density can reduce urban transport carbon emissions. High urban density often means developed and efficient public transport networks, as well as compact urban layouts, indicating that citizens can meet the needs of people through close employment, shopping, and leisure facilities and then effectively reduce per capita car use, thereby reducing fuel consumption and the corresponding greenhouse gas emissions.
(6) Public transportation effectiveness (PUB): The improvement of public transport efficiency has a significant positive effect on reducing urban transport carbon emissions, with the corresponding regression coefficient being significantly negative (−0.0196). In China, since the 1990s, many cities have begun to use clean energy buses and gradually turned from natural gas vehicles to electric automobile, which has low emissions and large capacity. The higher the efficiency of public transport, the more convenient it is for people to travel by, which will attract more citizens to abandon private cars and use public traffic, thus contributing to a significant reduction in urban transport carbon emissions. Meanwhile, the surrounding areas’ public transportation effectiveness is also conducive to the development of local low-carbon transportation, demonstrated by the significantly negative coefficient of W*LNPUB (−0.0613). This may be related to the “demonstration effect” of regional public transport development.
(7) Government environmental protection (GEP): According to Table 5, the coefficients of LNGEP and W*LNGEP are both negative (−0.0561 and −0.0678) and have passed the 1% level significance test, indicating that government environmental protection is negatively correlated with urban transportation carbon emissions. In other words, government environmental protection in local and surrounding cities reduces the local transport carbon emissions. The greater the government’s awareness and intensity of environmental protection, the stricter the implementation of relevant energy-saving and emission reduction policies, such as clean-energy vehicle subsidy policies, which can effectively slow down the deterioration of transport carbon emissions.

4. Spatio-Temporal Heterogeneity Analysis of Influencing Factors

Although the above SDM reflects the impact of various factors on urban transport carbon emissions, it does not reflect the heterogeneity of different cities, as well as the change tendency of the influencing degree. Therefore, this paper adopts the GWR model, as shown in Model (2), to discuss the spatio-temporal heterogeneity of influencing factors. Building on this foundation, in an effort to better illustrate the heterogeneity, this paper utilizes ArcGIS to visually display the coefficients estimated by the GWR model (due to space constraints and the legibility of the visualization map, only the spatial distribution of regression coefficients for 2006 and 2020 are shown in the analysis below).

4.1. GDP per Capita (PGDP)

Figure 1 illustrates the spatial distribution of the regression coefficient of PGDP in China. As shown in the figure, over time, the positive coefficient of PGDP shows a characteristic of continuous decreasing, where the value ranges from 0.0428 to 0.1703 in 2006 while it ranges between [0.0214, 0.0838], [0.0122, 0.0610], and [0.0072, 0.0473] in 2010, 2015 and 2020 in sequence, indicating that the degree of positive influence of PGDP on urban transport carbon emissions gradually diminishes. This observation partially confirms the existence of the EKC theory, suggesting that the relationship between economic development and the transport of CO2 may follow an inverted “U” shape. In addition, it can be found that the impact of PGDP on urban transportation carbon emissions has formed a spatial inertia, reflecting a spatial distribution feature with a gradient increase from northeast to southwest.

4.2. Population (POP)

The regression coefficients of POP on urban transport carbon emissions are displayed in Figure 2. It can be observed that the positive coefficient of POP increased from 2006 to 2020, indicating that the impact becomes progressively more significant with the increase in POP in most cities. Spatially, the influence of POP on urban transport carbon emissions exhibits a gradient pattern of gradual increase from northeast to southwest during the early period, while it has shifted to a gradually increasing distribution from north to south in the past several years. In addition, it should be noted that the negative influence coefficient of POP is generally located in China’s Northeast cities, which is closely related to the perennial population outflow in Northeast regions.

4.3. Private Cars per Capita (PRC)

From the perspective of private cars per capita, as illustrated in Figure 3, the impact coefficients of PRC influencing urban transport carbon emissions overall are positive among all cities, revealing a carbon emission increasing effect of PRC. However, the regression coefficients become progressively smaller from 2006 to 2020, where the value intervals are [3.3577,7.4689], [1.4698,4.3957], [0.6285,2.7690], and [0.1558,1.5467] in 2006, 2010, 2015, and 2020, respectively. This finding suggests that the worsening effect of the PRC on transport carbon emissions gradually becomes less pronounced in China. This trend is closely related to the continuous promotion and application of new energy vehicles in the private car sector.
In terms of spatial features, from 2006 to 2010, the influence of the PRC on urban transport carbon emissions generally presents a gradually increasing gradient pattern from south to north. This suggests that compared with the northern region, private cars per capita in the southern cities have a smaller positive impact on transport carbon emissions. Since 2010, the influence degree of PRC has shown the following spatial distribution characteristics: With cities along the Yellow Sea and Bohai Sea as the core, a gradual gradient decreases to the peripheral cities. In other words, in the cities along with the Yellow Sea and Bohai Sea, the deteriorating effect of PRC on carbon emissions is highest, which declines with the gradual spread of cities to the periphery in sequence.

4.4. Urban Road Area (URA)

The spatial distribution of the regression coefficients of urban road areas on urban transport carbon emissions is displayed in Figure 4. In 2006, the coefficient of URA is basically negative, however, in 2020, that turns basically positive, indicating that the effect of URA on urban transport carbon emissions has changed from inhibiting to promoting. The reason for this phenomenon may be as follows: URA has a dual impact on transport carbon emissions, that is, on the one hand, it can slow down traffic congestion and then reduce transport carbon emissions; On the other hand, it leads to more traffic demand and increase transport carbon emissions. At present, the latter influence dominates. Regarding spatial influence, from 2006 to 2020, the influence coefficients of URA basically show a spatial distribution pattern of decreasing from south to north.

4.5. Urban Density (URD)

Figure 5 displays the spatial distribution of the regression coefficients that urban density influences on urban transport carbon emissions. The impact of urban density on transportation carbon emissions is significantly different in different cities, with some showing positive effects and others showing negative influence, and the latter includes a gradual increase in the number of cities. In 2006, the development of URD in Jilin and Heilongjiang provinces reduced the transportation carbon emissions, while other regions deteriorated. The reason for this phenomenon may be that the cities in the northeastern regions are small in scale and have high coverage of public transport. By 2020, the URD’s coefficient of more cities in the northern region begins to turn from positive to negative, including the Beijing–Tianjin–Hebei urban agglomeration, some cities in the Shandong Peninsula urban agglomeration, and Baotou–Hohhot–Erdos–Yulin urban agglomeration.

4.6. Public Transportation Effectiveness (PUB)

From the perspective of public transportation effectiveness, as illustrated in Figure 6, the impact of PUB on urban transport carbon emissions was positive in most cities in 2006. However, the coefficient is small, indicating a slight worsening effect of PUB on carbon emissions. In 2020, the influence of PUB has all consistently become negative in China, showing an improved effect of low-carbon transport, which also reveals that the development of city public transport has played an increasingly important role in curbing transport carbon emissions. Spatially, since 2010, the regression coefficients of China cites’ PUB have generally exhibited a gradient-increasing distribution pattern from the north to south and from the north to northeast regions, which has also formed space inertia.

4.7. Government Environmental Protection (GEP)

Figure 7 displays the spatial distribution of regression coefficients of GEP in 2006 and 2020. In 2006, the impact of GEP on transport carbon emissions in China’s western regions was negative, while that in other cities was positive. However, the coefficients are all small, revealing that the effect of GEP is small. With time elapsing, the GEP coefficients in most cities begin to turn from positive to negative, and the absolute values of most coefficients have increased compared with that in 2006. Those conclusions reveal that the GEP plays an increasingly important role in curbing the carbon emission of urban transportation. In addition, from the perspective of space, although there are significant differences in the influence of GEP among different cities, the spatial distribution pattern of its influence has not formed inertia.

5. Conclusions and Policy Recommendations

How to promote the reduction in urban transport carbon emissions and achieve the “double carbon” goal as soon as possible has become an important topic of widespread concern in China. By constructing an SDM model and GWR model and utilizing city-level panel data from 2006 to 2020, this study explores the influencing factors and their spatiotemporal differentiation of carbon emissions from urban transportation in China. The findings indicate the following:
(1) The SDM model demonstrates that GDP per capita, population, urban road area, and private car per capita are significant factors contributing to the worsening in urban transportation carbon emissions, while public transportation effectiveness, urban density, and government environmental protection are key pathways for reducing urban carbon emissions of the transport industry.
(2) The estimation results of the GWR model show that the adverse effect of GDP per capita on transport carbon emissions decreases year by year during the study period and shows the spatial distribution pattern of the gradient increases from northeast to southwest. In contrast, the worsening effect of POP on transport carbon emissions (in most cities) has shown an increasing trend from 2006 to 2020 and currently displays a gradually increasing spatial differentiation from north to south. Meanwhile, private cars per capita have a worsening impact on urban transport carbon emissions, but the degree of this effect gradually decreases from 2006 to 2020, which also generally forms a spatial heterogeneity characteristic of “taking the cities along the Yellow Sea and Bohai Sea as the core and gradually decreasing to the surrounding cities”. At the same time, the effect of urban road areas on transportation carbon emissions shifts from negative to positive, with a distribution pattern from south to north. Although urban density increases transportation carbon emissions in some cities, this impact is turning from positive to negative in more cities, contributing to emission reduction. In addition, the reduction effect of public transportation effectiveness is gradually prominent, forming a space inertia of “increasing gradient from north to south and from north to northeast”. During the study period, government environmental protection plays an increasingly important role in reducing carbon emissions from urban transport.
Based on the aforementioned conclusions, the following policy recommendations can be put forward:
(1) Pay attention to the urban economy, and population development evolution, combined with reality, and scientifically and reasonably promote the decoupling development of economies, populations, and urban transport carbon emissions. Combined with the urban development and population increasing stage, a simulation model of urban transport carbon emissions should be established to simulate and warn the long-term development trend of urban transport carbon emissions, and corresponding economic development and population policies should be proposed.
(2) Steadily promote and accelerate the development of public transportation. The development and improvement of public transport and its efficiency have gradually become the main way to reduce the carbon emission of urban transport. Therefore, the government should accelerate the development of bus, rail transit, taxi, and online car modes, use bus lanes and other various strategies to improve the efficiency of public transport, and then accelerate the construction of low-carbon transport cities.
(3) Focus on improving the efficiency of urban transport as the goal to build and improve transport infrastructure for the development of low-carbon transport escort. Urban road areas can not only lead to larger transportation demand and then increase carbon emissions but also improve transportation efficiency by opening roadblocks and enhancing accessibility, thereby reducing transportation carbon emissions. Therefore, it is necessary for cities to have clear goals and further accelerate the construction of transportation infrastructure with low-carbon transportation as the goal.
(4) Forward-looking development and reasonable planning of urban layout to create a compact city. Urban density has gradually become an important means to improve carbon emission reduction in the field of transportation. So, the government should aim to develop efficient cities and strive to build close living circles to reduce the use of private cars to promote carbon emission reduction in transportation.
(5) Improve and strengthen environmental protection measures and popularize clean-energy vehicles. Based on the actual situation of the region, the government should improve energy conservation and emission reduction policies in the region, especially in the field of transportation, such as new energy subsidy policies, the elimination and replacement of traffic vehicles, the update of emission standards, etc., and provide vigorously support the emission reduction work in transportation with the help of fiscal policies, taxation, and administrative penalties. In addition, striving to promote the popularization and application of new energy vehicles to accelerate the construction of low-carbon transportation.
(6) Adapt to local conditions, make key breakthroughs, and formulate targeted urban transportation carbon reduction strategies. Due to the obvious spatial and temporal heterogeneity of influencing factors in Chinese cities, the government should fully combine the characteristics of urban reality and key influencing factors when implementing emission reduction policies and implement characteristic policies and measures that fit the regional reality.

Author Contributions

Conceptualization, P.Z. and B.S.T.; methodology, S.Z.; validation, P.Z. and B.S.T.; formal analysis, P.Z. and S.Z.; data curation, S.Z.; writing—original draft preparation, P.Z. and B.S.T.; writing—review and editing, S.Z. and Q.Y.; visualization, B.S.T.; supervision, Q.Y. All authors have read and agreed to the published version of the manuscript.

Funding

This research is supported by the Innovation Capability Support Program of Shaanxi (2023-CX-TD-11; 2024ZC-YBXM-158), Shaanxi Outstanding Youth Foundation (2021JC-27), and the China Postdoctoral Science Foundation (2023M730361).

Data Availability Statement

The data presented in this study are available upon request from the corresponding author. The data are not publicly available due to the results are generated by model construction.

Conflicts of Interest

The authors declare no conflicts of interest.

Appendix A

Table A1. Test for multi-collinearity.
Table A1. Test for multi-collinearity.
PGDPPOPPRCURAURDPUBGEP
Tolerance0.24640.94140.34540.23880.31690.97070.3103
Variance inflation factor (VIF)4.06001.06002.90004.19003.16001.03003.2200

References

  1. Li, W.; Li, H.; Zhang, H.; Sun, S. The analysis of CO2 emissions and reduction potential in China’s transport sector. Math. Probl. Eng. 2016, 2016, 1043717. [Google Scholar] [CrossRef]
  2. Lee, K.; Oh, W. Analysis of CO2 emissions in APEC countries: A time-series and a cross-sectional decomposition using the log mean Divisia method. Energy Policy 2006, 34, 2779–2787. [Google Scholar] [CrossRef]
  3. Xu, B.; Lin, B. Factors affecting carbon dioxide (CO2) emissions in China’s transport sector: A dynamic nonparametric additive regression model. J. Clean. Prod. 2015, 101, 311–322. [Google Scholar] [CrossRef]
  4. Grossman, G.M.; Kruger, A.B. Environmental Impacts of a North America World Free Trade Agreement; NBER Working Paper; NBER: Cambridge, MA, USA, 1991; p. 3914. [Google Scholar]
  5. Bai, C.; Chen, Y.; Yi, X.; Feng, C. Decoupling and decomposition analysis of transportation carbon emissions at the provincial level in China: Perspective from the 11th and 12th Five-Year Plan periods. Environ. Sci. Pollut. Res. 2019, 26, 15039–15056. [Google Scholar] [CrossRef] [PubMed]
  6. Zhu, C.; Du, W. A research on driving factors of carbon emissions of road transportation industry in six Asia-Pacific countries based on the LMDI decomposition method. Energies 2019, 12, 4152. [Google Scholar] [CrossRef]
  7. Rasool, Y.; Zaidi, S.A.H.; Zafar, M.W. Determinants of carbon emissions in Pakistan’s transport sector. Environ. Sci. Pollut. Res. 2019, 26, 22907–22921. [Google Scholar] [CrossRef] [PubMed]
  8. Amin, A.; Altinoz, B.; Dogan, E. Analyzing the determinants of carbon emissions from transportation in European countries: The role of renewable energy and urbanization. Clean Technol. Environ. Policy 2020, 22, 1725–1734. [Google Scholar] [CrossRef]
  9. Wang, J.T.; Ma, X.M. Research on factors affecting carbon emissions from transportation in China—An analysis based on two-level econometric model. J. Peking Univ. 2021, 57, 1133–1142. [Google Scholar] [CrossRef]
  10. Zhao, Q.; Zhou, Y.L.; Fang, Q.S. Analysis of spatial and temporal evolution of carbon emissions and its influencing factors in the central region. J. Environ. Sci. 2023, 43, 354–364. [Google Scholar] [CrossRef]
  11. Tang, Y.; Jiang, H. Analysis of the decoupling effect and driving factors of carbon emissions from the transportation sector in Guangdong Province. Sci. Rep. 2023, 13, 18744. [Google Scholar] [CrossRef]
  12. Cui, Q.; Li, Y. An empirical study on the influencing factors of transportation carbon efficiency: Evidences from fifteen countries. Appl. Energy 2015, 141, 209–217. [Google Scholar] [CrossRef]
  13. Jiang, X.T.; Su, M.; Li, R. Investigating the factors influencing the decoupling of transport-related carbon emissions from turnover volume in China. Sustainability 2018, 10, 3034. [Google Scholar] [CrossRef]
  14. Zhu, C.; Wang, M.; Yang, Y. Analysis of the influencing factors of regional carbon emissions in the Chinese transportation industry. Energies 2020, 13, 1100. [Google Scholar] [CrossRef]
  15. Zhang, K.; Liu, X.; Yao, J. Identifying the driving forces of CO2 emissions of China’s transport sector from temporal and spatial decomposition perspectives. Environ. Sci. Pollut. Res. Int. 2019, 26, 17383–17406. [Google Scholar] [CrossRef]
  16. Xu, X.; Xu, H. The Driving Factors of Carbon Emissions in China’s Transportation Sector: A Spatial Analysis. Front. Energy Res. 2021, 9, 664046. [Google Scholar] [CrossRef]
  17. Lv, Q.; Liu, H.; Yang, D.; Liu, H. Effects of urbanization on freight transport carbon emissions in China: Common characteristics and regional disparity. J. Clean. Prod. 2019, 211, 481–489. [Google Scholar] [CrossRef]
  18. Guo, M.; Meng, J. Exploring the driving factors of carbon dioxide emission from transport sector in Beijing-Tianjin-Hebei region. J. Clean. Prod. 2019, 226, 692–705. [Google Scholar] [CrossRef]
  19. Li, J.; Jing, M.T.; Yuan, Q.M. Research on regional transportation carbon emission measurement and driving factors under green development—Taking Beijing-Tianjin-Hebei as an example. Arid. Zone Resour. Environ. 2018, 32, 36–42. [Google Scholar] [CrossRef]
  20. Wu, Y.M.; Li, J.X. Analysis of provincial industrial total factor productivity based on geographically weighted regression model. Econ. Geogr. 2006, 5, 748–752. [Google Scholar]
  21. Bowman, A.W. An alternative method of cross-validation for the smoothing of density estimates. Biometrika 1984, 71, 353–360. [Google Scholar] [CrossRef]
  22. Jung, M.C.; Kang, M.; Kim, S. Does polycentric development produce less transportation carbon emissions? Evidence from urban form identified by night-time lights across US metropolitan areas. Urban Clim. 2022, 44, 101223. [Google Scholar] [CrossRef]
  23. Kenworthy, J.R.; Laube, F.B. Patterns of automobile dependence in cities: An international overview of key physical and economic dimensions with some implications for urban policy. Transp. Res. Part A Policy Pract. 1999, 33, 691–723. [Google Scholar] [CrossRef]
  24. Liddle, B. Urban density and climate change: A STIRPAT analysis using city-level data. J. Transp. Geogr. 2013, 28, 22–29. [Google Scholar] [CrossRef]
  25. Liddle, B. Urban transport pollution: Revisiting the environmental Kuznets curve. Int. J. Sustain. Transp. 2015, 9, 502–508. [Google Scholar] [CrossRef]
  26. Wang, W.W.; Zhang, M.; Zhou, M. Using LMDI method to analyze transport sector CO2 emissions in China. Energy 2011, 36, 5909–5915. [Google Scholar] [CrossRef]
Figure 1. Spatial distribution of regression coefficients of PGDP in 2006 and 2020.
Figure 1. Spatial distribution of regression coefficients of PGDP in 2006 and 2020.
Energies 17 00756 g001
Figure 2. Spatial distribution of regression coefficients of POP in 2006 and 2020.
Figure 2. Spatial distribution of regression coefficients of POP in 2006 and 2020.
Energies 17 00756 g002
Figure 3. Spatial distribution of regression coefficients of PRC in 2006 and 2020.
Figure 3. Spatial distribution of regression coefficients of PRC in 2006 and 2020.
Energies 17 00756 g003
Figure 4. Spatial distribution of regression coefficients of URA in 2006 and 2020.
Figure 4. Spatial distribution of regression coefficients of URA in 2006 and 2020.
Energies 17 00756 g004
Figure 5. Spatial distribution of regression coefficients of URD in 2006 and 2020.
Figure 5. Spatial distribution of regression coefficients of URD in 2006 and 2020.
Energies 17 00756 g005
Figure 6. Spatial distribution of regression coefficients of PUB in 2006 and 2020.
Figure 6. Spatial distribution of regression coefficients of PUB in 2006 and 2020.
Energies 17 00756 g006
Figure 7. Spatial distribution of regression coefficients of GEP in 2006 and 2020.
Figure 7. Spatial distribution of regression coefficients of GEP in 2006 and 2020.
Energies 17 00756 g007
Table 1. Definition of the variable adopted in the paper.
Table 1. Definition of the variable adopted in the paper.
VariableDefinition or MeasurementUnit
Urban transport carbon emission (TCE)CO2 emission of urban transport industry (including road, rail, water, and air traffic) measured by MEIC model divided by the total population.Tons per capita
GDP per capita (PGDP)GDP divided by the total population.CNY per 10,000 people
Population (POP)Urban total population number at the end of year.10,000 people
Private cars per capita (PRC)Total number of private cars divided by the total population in a city.Vehicles per 10,000 people
Urban road area (URA)Urban road area divided by the total population.m2 per capita
Urban density (URD)Urban total population divided by built-up area.10,000 people per km2
Public transportation effectiveness (PUB)The ratio of bus passenger traffic to the number of buses in operation.Ten thousand passengers per car
Government environmental protection (GEP)Ratio of urban green land area to the total population.Hectare per 10,000 people
Table 2. The descriptive statistics of all variables adopted in the paper.
Table 2. The descriptive statistics of all variables adopted in the paper.
VariableObsMeanStd.dev.MinMax
TCE42600.59100.42400.04064.8900
PGDP42604.83605.00100.276050.6300
POP4260440.7000313.400016.41003416
PRC42600.09020.09570.00061.1340
URA42604.67006.16500.003675.9100
URD42600.35400.42500.02228.7490
PUB42600.00140.00099.78 × 10−60.0244
GEP42600.00180.00347.93 × 10−60.0481
Table 3. Results of Moran’s index measurement.
Table 3. Results of Moran’s index measurement.
YearMoran’s IE(I)SD(I)Z-Statisticp-Value
20060.171 ***−0.0040.0374.7240.000
20070.164 ***−0.0040.0384.4250.000
20080.186 ***−0.0040.0384.9590.000
20090.178 ***−0.0040.0304.7240.000
20100.208 ***−0.0040.0385.4930.000
20110.226 ***−0.0040.0395.8880.000
20120.223 ***−0.0040.0395.8020.000
20130.223 ***−0.0040.0405.7240.000
20140.217 ***−0.0040.0405.5730.000
20150.197 ***−0.0040.0405.0680.000
20160.213 ***−0.0040.0405.4460.000
20170.229 ***−0.0040.0405.8240.000
20180.235 ***−0.0040.0405.9420.000
20190.235 ***−0.0040.0405.9330.000
20200.254 ***−0.0040.0406.3980.000
Note: *** denotes the 1% significance levels.
Table 4. LM and LR tests.
Table 4. LM and LR tests.
TestTest Statisticsp-Value
LM testLMSpatial error1281.5010 ***0.0000
Spatial lag407.1780 ***0.0000
Robust LMSpatial error934.6620 ***0.0000
Spatial lag60.3390 ***0.0000
LR testH0: SAR model224.6100 ***0.0000
H0: SEM model114.6400 ***0.0000
Note: *** denotes the 1% significance levels.
Table 5. Estimation results of SDM.
Table 5. Estimation results of SDM.
VariableSDMOLS
CoefficientVariableCoefficient
LNPGDP0.3560 ***W*LNPGDP−0.3260 ***0.1580 ***
(0.0161)(0.0248)(0.0166)
LNPOP0.3330 ***W*LNPOP−0.6030 ***0.0321
(0.0155)(0.0457)(0.0100)
LNPRC0.2820 ***W*LNPRC0.0964 ***0.2210 ***
(0.0111)(0.0152)(0.0101)
LNURA0.1210 ***W*LNURA0.0784 ***0.1820 ***
(0.0136)(0.0257)(0.0163)
LNURD−0.0403 **W*LNURD−0.02300.1510 ***
(0.0174)(0.0317)(0.0205)
LNPUB−0.0196 **W*LNPUB−0.0613 ***0.0080
(0.0086)(0.0146)(0.0098)
LNGEP−0.0561 ***W*LNGEP−0.0678 ***−0.0861 ***
(0.0113)(0.0219)(0.0137)
rho0.4630 ***
(0.0159)
Sigma2_e0.1050 ***
(0.0023)
No.4260 4260
R20.6180 0.6110
Note: *** and ** denotes the 1% and 5% significance levels.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Zhao, P.; Tian, B.S.; Yang, Q.; Zhang, S. Influencing Factors and Their Spatial–Temporal Heterogeneity of Urban Transport Carbon Emissions in China. Energies 2024, 17, 756. https://doi.org/10.3390/en17030756

AMA Style

Zhao P, Tian BS, Yang Q, Zhang S. Influencing Factors and Their Spatial–Temporal Heterogeneity of Urban Transport Carbon Emissions in China. Energies. 2024; 17(3):756. https://doi.org/10.3390/en17030756

Chicago/Turabian Style

Zhao, Peng, Bei Si Tian, Qi Yang, and Shuai Zhang. 2024. "Influencing Factors and Their Spatial–Temporal Heterogeneity of Urban Transport Carbon Emissions in China" Energies 17, no. 3: 756. https://doi.org/10.3390/en17030756

APA Style

Zhao, P., Tian, B. S., Yang, Q., & Zhang, S. (2024). Influencing Factors and Their Spatial–Temporal Heterogeneity of Urban Transport Carbon Emissions in China. Energies, 17(3), 756. https://doi.org/10.3390/en17030756

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop