Study of Mobile Source Pollutants Based on Multi-Source Data Fusion: A Case Study of Zibo City, China

Abstract

The primary pollutants emitted from mobile sources have become one of the main sources of urban air pollution. In this study, the primary pollutants CO, HC, NOx and PM from mobile sources in Zibo city, China are studied. Firstly, the localized mobile-source pollutant emission factors are corrected using vehicle emission experiments; secondly, multi-source data fusion is performed using road network data, road traffic data, air quality and meteorological data; then, the corrected pollutant emission factors and multi-source data are used to construct a localized emission measurement model and dispersion simulation model, visualize the emission distribution and propose residual concentration correction to accurately obtain the pollutant sharing rate. The results show that the pollutant emission trends are highly coupled with the distribution of urban residents’ working hours. Based on the localized dispersion model used to analyze the pollutant traceability at air monitoring stations, the emission sharing rate of NOx is the largest, and according to the analysis of the changing trend of mobile-source emission sharing rate, the mobile source pollutants in Zibo are mainly influenced by pollutant emissions from road motor vehicles. This study provides data support and a theoretical basis for the management of the transportation industry to carry out exhaust gas control of motor vehicles and non-road mobile machinery and to make decisions related to improving the air environment (delineating the scope of low emission zones).

1. Introduction

In recent years, due to rapid global economic growth, rapid population expansion and accelerated urbanization, air pollutant emissions have increased significantly [1], and the frequency of heavy air pollution is high, posing a serious threat to human health [2]. Air pollution can be divided into primary pollution, secondary pollution and combined pollution. Primary pollutants account for about 30–40% of total atmospheric pollutants, and it is only after a certain amount of accumulated primary pollutants are irradiated by the sun and undergo complex photochemical reactions that secondary pollutants are produced in the atmosphere [3]. Therefore, it is crucial to master the emission and dispersion characteristics of primary pollutants, and mobile sources as an important source of primary pollutants are the focus of environmental management authorities’ management [4,5]. Mobile sources include on-road mobile sources (motor vehicles) and non-road mobile sources (construction machinery, etc.). NOx emissions from mobile sources are important precursors of secondary organic aerosols (SOA) [6,7,8]. Mobile source emissions are not only an important source of PM [9,10,11], but also the main contributor of secondary pollutants in the atmosphere. Relevant epidemiological studies show that long-term exposure to PM can cause adverse effects on health, including cardiovascular and respiratory diseases [12,13,14,15], and it is closely related to early death [16], with China accounting for about 40% of premature deaths from air pollution globally [17]. Therefore, it is necessary to quantify the emissions of mobile sources and develop relevant measures to reduce mobile-source emissions.

Atmospheric pollutants are characterized by easy mixing, complex influencing factors and multiple pollution sources, so traceability is difficult and there are fewer studies on mobile-source pollutant emissions and traceability analysis. Many researchers have conducted emission measurements using emission factor models for road mobile sources and road mobile sources, respectively [18,19,20,21,22,23,24,25]. In addition, a dispersion model [26] was used to analyze the pollutants retrospectively.

The basic emission factors are the basis for accurate emission measurement, and there are mainly the following kinds of relevant studies for obtaining the basic emission factors. The first is vehicle testing using chassis dynamometers. The highly controlled and ambient conditions make the relevant measurements highly reproducible, limiting the ability to fully represent real-world emissions [27]. Luján et al. [28] conducted an emission study using PEMS for EURO VI light-duty diesel vehicles under different driving conditions and showed that low-speed acceleration resulted in higher NOx emissions than high-speed acceleration and that pollutant emission factors were affected by road conditions. Triantafyllopoulos et al. [29] investigated CO₂ and NOx emissions from three EURO VI diesel vehicles under controlled and real-road conditions and showed that there were significant differences in emissions between the two cases. Simonen et al. [30] also made a comparison of both cases and the experimental results were different. All of the above studies showed that studies using chassis dynamometers and PEMS to obtain emission factors are limited to a small number of vehicles, provide less comparison of multiple vehicle types and emission categories, and take less account of road conditions.

Emission models are models that describe the relationship between pollutant emissions and influencing factors, focusing on simulating the emission characteristics of mobile sources and substituting the use characteristics of mobile sources themselves to complete the pollutant emission measurement. The current research on emission models is mainly focused on analyzing emission factors of existing models and establishing emission inventories. In terms of analyzing emission factors, Zhao et al. [31] localized the IVE model by using the obtained cab-driving conditions and local parameters, combined with Vehicle Mass Analysis System (VMAS) experiments, and measured the annual emissions of cab pollutants. Perugu [22] used modified Indian driving conditions and local light vehicle specific driving conditions to correct the MOVES model for emission rates using an improved VSP-based MOVES model for modeling light-duty vehicle emissions in India. Ghaffarpasand et al. [32] estimated road mobile-source emission factors for Isfahan region using the IVE motor vehicle emission model and developed the Isfahan 2018 vehicle emission inventory with a bottom-up approach with a resolution of 1 km × 1 km. Wang et al. [33] established a gridded emission inventory of motor vehicle pollutants at 1 km × 1 km resolution in Xi’an city in 2017 based on the MOVES model and ArcGIS technology. The above literature completed the measurement of localized motor vehicle pollutant emissions using existing emission models, but relying on historical statistics such as emission factors and motor vehicle holdings cannot reflect the actual situation of road traffic in a timely manner, and there are still large differences compared with actual road emissions. In the process of establishing emission inventories, most scholars have only studied emission inventories with high spatial resolution; few have studied emission inventories with high temporal resolution. Desouza, C.D. et al. [34] used a portable emission measurement system to measure pollutant emissions from 30 types of non-road mobile machinery with different engine standards in the city of London. Fan et al. [35] used a portable emission measurement system to test 12 types of machinery under different operating conditions to build an emission inventory of construction machinery in Chengdu and analyze the contribution of each type of machinery. Compared with on-road mobile sources, there are fewer studies on emission modeling and localization of non-road mobile sources in China, and fewer studies on city-level real-time emissions.

To accurately analyze pollutant traceability, it is necessary to consider comprehensive dispersion models, which can calculate the spatial and temporal distribution of mobile-source pollutants based on the source intensity and dispersion state. In recent years, scholars have conducted a lot of research on pollutant dispersion models [36,37]. The meteorological environment and wind speed are important influencing factors to be considered [38]. Gao et al. [39] selected Shijiazhuang City, Hebei Province, which is severely affected by NO₂ pollution, as an example, established a NO₂ dispersion model based on a Gaussian plume model, analyzed the relevant factors affecting NO₂ dispersion, combined with the annual wind speed, wind direction, atmospheric stability and other weather conditions in Shijiazhuang City, and simulated the dispersion area and concentration changes. However, the study only simulated the average value of each weather condition, which differs from the actual dispersion situation. Therefore, this study uses multi-source data and a Gaussian plume model to correct the residual concentrations of pollutants at environmental monitoring points and calculate the pollutant sharing ratio, which improves the accuracy and reliability of the calculation.

Based on the current research status, this study takes mobile source pollutants in Zibo City, China as the research object, makes localization corrections to the basic emission factors based on vehicle-borne emission experiments, constructs a localization emission and dispersion model with multi-source data, analyzes the real-time emission characteristics of urban mobile source primary pollutants, and makes corrections to the pollutant concentrations around environmental monitoring stations considering different wind speed scenarios and the actual situation of dispersion in order to accurately analyze the pollutant traceability. In this way, we can carry out research on mobile-source pollutant emission measurement and traceability.

2. Materials and Methods

In order to accurately reflect the impact of mobile-source pollutant emissions under actual road network characteristics and vehicle characteristics, the base emission data of road mobile sources obtained from the actual on-board emission experiments are used as the basis, and the base emission factors are revised; meanwhile, the road network data, road traffic data, meteorological data and air quality data are combined to analyze the mobile source pollutant emissions and dispersion in Zibo City.

2.1. On-Board Emission Experiments

By conducting actual on-board emission experiments to obtain road emission data of driving gas and diesel vehicles, we provide data support for the construction of a mobile-source pollutant emission measurement model. In order to build the actual on-board emission test platform, MAHA Emission tester MET 6.3 portable emission test equipment was used to conduct the on-board emission test. The optional equipment can simultaneously complete exhaust gas, smoke and particulate concentration tests, and calculate CO, CO₂, HC, NO, NO₂ and particulate mass in the exhaust gas in real time. The specific experimental steps are: (1) Platform construction. Firstly, use the network cable to connect the test equipment with the computer, and use the USB interface or Bluetooth to realize the connection between GPS, OBD and the computer, as shown in Figure 1. (2) Data acquisition. Then, the exhaust gas collection pipe in the test equipment is connected to the exhaust pipe of the car, and the collected exhaust gas is detected in real time. The test equipment uses an infrared method to measure CO, CO₂ and HC; the acousto-optical spectroscopy method is used to measure NO, NO₂ and O₂ based on gas absorption spectrum characteristics, and the light scattering method is used to measure particulate matter. (3) Data matching. Emission data collected by the test equipment in real time, real-time vehicle OBD status and GPS driving track data are stored in the emission test system (PEMS, Portable Emission Measurement System) through the computer terminal, and the software automatically matches the exhaust emission data with the OBD data GPS vehicle operation data.

In addition, in order to obtain more comprehensive and representative emission data, the on-board emission test also needs to consider the influence of the experimental vehicle, the experimental road, the experimental period and other factors. Experimental design: (1) In order to ensure that the emission characteristics of urban motor vehicles are representative, gasoline vehicles with displacement below 1.0 L and above 2.0 L are divided into five zones, diesel vehicles with displacement below 2.0 L and above 4.0 L are divided into five zones and at least five brands of vehicles are selected for each zone in principle. (2) In order to analyze the impact of different road types on emissions, according to the classification of urban roads in China, they are divided into four categories: expressways, main roads, secondary roads and branch roads. The experiments were selected as experimental objects in the six districts and counties of the city in the frequency of use of roads. Details of the experimental vehicles and road sections can be found in Appendix A. (3) The experimental periods were selected as morning peak (7:00 a.m.–9:00 a.m.), evening peak (4:00 p.m.–8:00 p.m.) and flat peak (9:00 a.m.–5:00 p.m.) to obtain the effect of traffic flow on vehicle exhaust emissions. The total experiment time exceeds 500 h.

Based on the above on-board emission experiments, the database required for the emission model is constructed to provide data support for the construction of the mobile-source pollutant emission measurement model.

2.2. Multi-Source Data Fusion

In order to construct the mobile-source pollutant emission measurement and dispersion model, this study combines multi-source data such as road network data, road traffic data, meteorological data and air quality data based on the actual vehicle emission experimental data, and uses them to build the mobile-source localized emission model and dispersion model under different wind speed scenarios. The sources of multi-source data acquisition are as follows: (1) Road network data. Accurate measurement of mobile-source pollutant emissions requires consideration of the effects of different road lengths and road types on mobile-source emission performance. In this study, the map data of the study area was obtained through the OpenStreetMap website, and in order to facilitate the reading of the data by the GIS platform, the map data format obtained was converted to shp format using a converter, and the urban road type, road length and other road network data are obtained after importing GIS, as shown in Figure 2. (2) Road traffic data. In order to conduct traceability analysis of mobile-source pollutants, the road traffic data needed are the traffic flow of each vehicle model, fuel type, emission stage, average speed, etc. In this study, the relevant road traffic data are obtained through the Zibo City Big Data Center, that is, time by time vehicle type, fuel type and emission stage vehicle flow information under different road types. (3) Meteorological data. Meteorological information such as wind speed and direction is an important factor affecting pollutant dispersion, so meteorological data and air quality data are the data basis for pollutant dispersion simulation and traceability analysis. In this study, meteorological data are obtained from the Zibo Meteorological Station, which are mainly hourly data of each administrative region throughout the day, including wind speed, wind direction and cloud volume. (4) Air quality data. In order to conduct accurate traceability analysis on the range of air monitoring sites, the location information and real-time monitoring data of air monitoring sites need to be obtained. Therefore, the air quality data in this study come from the data at the air monitoring stations measured by the urban air quality monitoring system, including the concentrations of air pollutants such as PM_2.5, SO₂, NO₂, CO and the air quality index (AQI). The multi-source data types and their sources are shown in

Table 1. Multi-source data types and their sources.

Data Type	Data Information	Data Source
Road Network Data	Zibo city basic road map, Zibo city different road types distribution, road length	OpenStreetMap
Road Traffic Data	Mobile source traffic volumes and travel speeds by vehicle type;	Zibo Large Data Center
	Vehicle operating conditions data for different road classes;
	Operating vehicle GPS data;
	Vehicle data for different vehicle types, vehicle age, fuel type, emission phase, etc.
Meteorological Data	Hour-by-hour meteorological information on wind speed, wind direction, cloudiness and pressure in Zibo City	Zibo Meteorological Center
Air Quality Data	Location of air quality monitoring stations, real-time monitoring concentrations of pollutants, Air Quality Index (AQI)	Zibo Ecological Environment Monitoring Center

below.

Road network data, road traffic data, meteorological data and air quality data need to be fused and synchronized by time and road location coordinates. The data fusion process is: (1) Data pre-processing. In order to eliminate the impact of different types of data dimensions such as location data, traffic flow data, meteorological data and air quality data on data analysis, and to ensure the reliability of the training model results, the data are first normalized, and for targeting missing data, the DBN-LMBP algorithm is used to supplement the missing values; some of the data after pre-processing are shown in Figure 3. (2) Format transformation. Geometric calibration is performed on the pre-processed data, and all data formats are converted into image formats to ensure that all data are integrated with each other for display in the same interface using the GIS platform. (3) Data matching. Road traffic data are fused with road network data through time and road name to realize real-time traffic flow, speed, vehicle operation and other information of different road types integrate into road network data to obtain hourly traffic volume data, hourly traffic speed, mechanical operation and other data of different road sections; in addition, meteorological data and air quality data are fused with road network data through location coordinates, and air monitoring site monitoring data are matched with location information to obtain hourly weather and air quality data. (4) Data visualization. Using React, a web development framework based on JavaScript language, a visualization interface is developed to visualize emission maps, holdings of each vehicle, pollutant concentrations and pollutant traceability analysis. The integration process is shown in Figure 4 below.

2.3. Mobile Source Pollutant Emission Measurement Model

The construction of emission measurement models is the basis for mobile-source pollutant emission measurement. Therefore, in order to measure the pollutant emissions from urban mobile sources, localized emission measurement models are constructed for on-road mobile sources and non-road mobile sources, respectively, in this study.

2.3.1. Road Mobile Source Pollutant Emission Measurement Model

The emission factors can reflect the emission level of a certain vehicle type in a region and are the basis of emission measurement. To ensure the accuracy of the emission model, this study establishes a motor vehicle emission model based on localized emission factors and real-time traffic flow information. Firstly, localized motor vehicle emission factor data were obtained and corrected based on actual on-board emission experiments and multi-source data [40], and then road network road mobile source emission measurements were completed using road section traffic flow and other data. The mathematical expression for calculating pollutant emissions from motor vehicles [41] is as follows:

$$E_i=E{F}_a\cdot V_a\cdot L_i$$

(1)

where E_i is the vehicle pollutant emissions on link i, g/h; EF_a is the pollutant emission factor of vehicle type a, g/km; V_a is the road traffic flow of vehicle type a and veh/h; L_i is the length of link i, km.

The traffic flow of the road sections involved in the calculation model is obtained from the urban traffic big data center. Due to the lack of a road mobile-source information database and other factors, it is difficult to accurately calculate the traffic volume of road sections based on fuel types and emission standards. Therefore, in the process of measuring road section emissions, the traffic volume is further counted by vehicle type based on the proportion of motor vehicle ownership with different fuel types and emission standards in the city, so that traffic volume data by vehicle type, emission standard and fuel type can be obtained for each road section on a time-by-time basis.

2.3.2. Non-Road Mobile Source Pollutant Emission Measurement Model

Urban non-road mobile sources are mainly construction machinery, including forklifts, excavators, loaders, etc. For the pollutant emission measurement of construction machinery, according to the Technical Guide for the Preparation of Non-road Mobile Pollutant Source Emission Inventory (for Trial Implementation) (the Guide) issued by the Ministry of Environmental Protection of China [25], the power-based emission model is adopted to improve the time accuracy to every hour, and the pollutants are calculated according to the online number, average power and activity level of different machinery; the mathematical expression is as follows:

$$E={\displaystyle \sum _{i=1}^n{\displaystyle \sum _{j=1}^m{\displaystyle \sum _{k=1}^p{P}_{i,j,k}\times G_{i,j,k}\times L{F}_{i,j,k}\times h{r}_{i,j,k}\times E{F}_{i,j,k}}}}$$

(2)

where E is the pollutant emission of non-road mobile machinery, g; P_{i, j, k} are the online quantity of non-road mobile machinery in i model, j power section and k emission stage, set; G_i,j,k are the average rated net power of non-road mobile machinery in i model, j power section and k emission stage, kW/set; LF_i,j,k are load factors of non-road mobile machinery in i model, j power section and k emission stage; hr_i,j,k are online time of non-road mobile machinery in i model, j power section and k emission stage, h; EF_i,j,k are emission factors of non-road mobile machinery in i model, j power section and k emission stage, g/kWh; n is the number of types of non-road mobile machinery, species; m is the number of power sections of non-road mobile machinery, piece; p is the number of emission stages of non-road mobile machinery, piece.

The non-road mobile-source pollutant emission measurement model based on the power method involves parameters that need to be determined, such as ownership, power distribution, emission stages and annual use hours of each type of machinery, which are obtained through the non-road mobile-source registration filing information of Zibo Ecological Environment Bureau, and the emission factors and load factors refer to the default values in the Guide.

2.4. Mobile Source Pollutant Dispersion Model

2.4.1. Dispersion Model under Different Wind Speed Conditions

Considering the influence of the meteorological environment and other factors on the diffusion and migration of pollutants, the Gaussian plume model is used to complete the simulation of the diffusion of mobile source pollutants in the road network for the actual urban traffic road mobile-source exhaust diffusion and different wind speed scenarios, and to conduct traceability analysis around air monitoring stations.

According to Huang et al. [42], wind speed, wind direction and atmospheric stability all affect the diffusion of pollutants, according to which the diffusion of pollutants when different meteorological conditions change is analyzed. In general, wind speed affects the dilution capacity of pollutants; the higher the wind speed, the stronger the atmospheric dilution capacity and the lower the pollutant concentration, and the wind speed has a negative correlation with the pollutant concentration. Wind direction affects the direction of horizontal migration and diffusion of pollutants, and heavy pollution areas generally occur in the downward direction; according to the Pasquill method [43], atmospheric stability is divided into six grades from A to F representing strong instability to stability, and when the atmospheric complex is unstable, atmospheric convection is strong and pollutants are easy to diffuse; when the atmospheric complex is stable, the atmospheric turbulence is suppressed and pollutants are not easily diffused and diluted. In addition, temperature and cloudiness will affect the atmospheric stability grade. Based on the above factors, the dispersion model under different wind speed conditions is established.

(1) Diffusion model with wind. When the wind speed is greater than or equal to 2 m/s, the mobile source emissions are linear. Assuming that the direction of the line source is the positive x direction, the mathematical expression for calculating the diffusion concentration of mobile source pollutants at the receiver point (x, y, z) in the downwind direction [44] is as follows:

$$C\left(x,y,z\right)=\frac{Q}{2U\pi {\sigma }_y{\sigma }_z}\exp \left(-\frac{y^2}{2{\sigma }_y^2}\right)\cdot \left\{\exp \left[-\frac{{\left(z-H\right)}^2}{2{\sigma }_z^2}\right]+\exp \left[-\frac{{\left(z+H\right)}^2}{2{\sigma }_z^2}\right]\right\}$$

(3)

where C(x, y, z) is the diffusion concentration of mobile source pollutants at the receptor point (x, y, z) in the downwind direction, mg/m³; σ_y is the diffusion parameter in the horizontal direction, m; σ_z is the diffusion parameter in the vertical direction, m; Q is the emission source intensity of pollutants from mobile sources, g/km; U is the average wind speed at the height of the discharge outlet, m/s; H is the effective source height, m; x is the downwind distance, m; y is the distance across the wind, m; z is the height of receptor point, m.

(2) Diffusion model under low wind and static wind conditions. When the wind speed is less than 0.5 m/s, it is a static wind condition; when the wind speed is between 0.5 m/s and 2 m/s, it is a small wind condition. This situation requires the use of the moving smoke mass model in the Gaussian model; then, the mathematical expression for calculating the diffusion concentration of pollutants from mobile sources at the receptor point (x, y, z) in the downwind direction [45] is as follows:

$$\begin{array}{l}C\left(x,y,z,H\right)=\\ {\displaystyle \underset{0}{\overset{\infty }{\int }}\frac{Q}{{\left(2\pi \right)}^{\frac{3}{2}}{\sigma }_x{\sigma }_y{\sigma }_z}}\exp \left[-\frac{{\left(x-UT\right)}^2}{2{\sigma }_x^2}\right]\exp \left(-\frac{y^2}{2{\sigma }_y^2}\right)\left\{\exp \left[-\frac{{\left(z-H\right)}^2}{2{\sigma }_z^2}\right]+\exp \left[-\frac{{\left(z+H\right)}^2}{2{\sigma }_z^2}\right]\right\}dT\end{array}$$

(4)

where T is the operation time of the puff, that is, the difference between the time when the puff is released and the time when the concentration value is calculated; σ_x is the diffusion parameter of the puff in the x direction.

The parameters involved in the above-constructed diffusion model based on Gaussian diffusion theory that need to be determined are the diffusion parameters. According to the national standard GB/T13201-91, the diffusion parameters σ_y, σ_z diffusion parameters influenced by the downwind distance x are obtained by using the diffusion parameter method of Briggs [46,47].

2.4.2. Measurement of Mobile-Source Dispersion Concentration at Air Monitoring Sites

In order to conduct traceability analysis at air monitoring sites, the dispersion concentration of mobile-source pollutants at air monitoring sites needs to be measured by using dispersion models under different wind speed scenarios. Under windy conditions, the diffusion of pollutants is influenced by the wind direction and wind speed in the range of 22.5° downwind, so a 45° sector downwind of the pollution source is selected as the diffusion area of pollutants in this study. After several concentration measurement experiments, under the condition that the average wind speed is 3.4 m/s, the diffusion concentration is close to zero at a distance of about 3 km from the pollution source, so under windy conditions, the pollutant source is the center of the circle, with 3 km as the radius, downwind of the 45° circular angle parallelogram of the fan as the diffusion range for calculation. In low-wind and static-wind conditions, pollutants generally spread slowly in all directions, and the diffusion concentration of the pollution source in all directions is basically the same, so the diffusion range is considered circular; the diffusion range is related to the effective source height, so the radius is n times the effective source height. In summary, for the traceability analysis of pollutants from mobile sources, a circle with a radius of 3 km is used as the pollutant dispersion concentration calculation area with a typical urban air monitoring station as the center of the circle; the corresponding dispersion model is used to calculate the dispersion concentration upwind of the air monitoring site so as to complete the traceability analysis of pollutant concentrations at the air monitoring site. The dispersion schematic is shown in Figure 5 below.

In the actual dispersion situation, meteorological conditions have a certain influence on the residual concentration of pollutants. Therefore, the actual diffusion concentrations of pollutants at air monitoring stations are corrected. (1) In the case of windy conditions before and after the change in meteorological conditions, the difference in wind speed is not large, and the residual pollutant dispersion in the next period is the same as the source strength of the pollutant in the previous period; in this study, the wind speed before and after the change in meteorological conditions is set to u₁. At this time, the diffusion distance of pollutants with the change in wind direction is the same for both time periods, x1 = x2, and the dilution effect of wind speed on the concentration of pollutants cannot be taken into account with constant wind speed, so the result of the concentration needs to be multiplied by the dilution factor ε. Then, the residual concentration of pollutants at any receptor point under windy conditions, C′ is as in expressions (7) and (8), is shown. (2) In the presence of static wind conditions before and after the change of meteorological conditions, it is assumed that the residual pollutant does not migrate in the horizontal direction; the diffusion impact area of the pollutant remains unchanged. If u₁ and u₂ are not zero at the same time, the residual concentration C is shown in expression (9). The wind speed is zero before and after the change in meteorological conditions. Then, the residual concentration C′ is shown in expression (10), assuming that the pollutant neither migrates nor is diluted. Then, the sharing rate of mobile source pollutant b around the air monitoring station is shown in expression (11).

$$\epsilon =\frac{u_1}{u_1+u_2}$$

(5)

$$C^{\prime }=\epsilon C\left(f\left(x\right),y,z\right)=\frac{u_1}{u_1+u_2}C\left(\frac{x}{\cos \left(\raisebox{1ex}{$\alpha $}\!\left/ \!\raisebox{-1ex}{$2$}\right.\right)},y,z\right)$$

(6)

$$C^{\prime }=\frac{1}{u_1+u_2}C$$

(7)

$$C^{\prime }=C$$

(8)

$$r_b=\frac{C_b+{C_b}^{\prime }}{C_{Mb}}$$

(9)

where α is the angle of wind direction change, and C is the pollutant diffusion concentration in the previous period; r_b is the share of pollutant b mobile source emissions; C_b is the diffusion concentration of pollutant b moving pollution source; C_b′ is the residual concentration of pollutant b; C_Mb is the monitoring concentration of pollutant b at the space monitoring station.

3. Results

Based on the emission measurement model and dispersion model, we analyze the emission and dispersion characteristics of urban mobile sources, including the spatial and temporal distribution characteristics of pollutant emissions and traceability analysis, so as to screen out high-emission road sections and areas and formulate precise corresponding emission-reduction countermeasures.

3.1. Spatial and Temporal Emission Characteristics of Pollutants

From Figure 6 and Figure 7, it can be seen that the spatial and temporal distribution of pollutant emissions from mobile sources in the city is characterized as follows: (1) From the distribution of pollutant emissions hour by hour, the changes in traffic flow and pollutant emissions from different road types coincide with the overall trend of changes in pollutant emissions, showing the characteristics of “double peak M”, with peak traffic flow and pollutant emissions occurring in the morning peak (7:00–9:00) and evening peak (16:00–18:00). (2) In terms of the spatial distribution of total daily average emissions, the city’s mobile sources emit the most NOx and CO. (3) In addition, during the peak hours, the roads with higher emissions are mainly distributed in the northeast part of the city. Northeast of the city and during working hours, pollutant emissions are closely related to non-road mobile machinery in industrial areas, mainly concentrated in the central part of the city.

The reason analysis includes: (1) The influence of vehicle operation characteristics on emission factors; the traffic flow of road sections during peak hours is large, the vehicle driving speed is relatively reduced, the vehicle idling situation is correspondingly aggravated and the pollutant emission factor is high [48], so the pollutant emission in this condition is serious. (2) The trend of pollutant emission from mobile sources and the distribution of urban residents’ working hours is highly coupled, and during the 11:00 a.m. working-day time, pollutant emissions are closely related to non-road mobile machinery in industrial parks, mainly concentrated in logistics parks, resettlement projects and industrial parks in urban areas. (3) Related to the traffic flow structure of road network: the dense road network and high traffic volume in the northeastern part of the city, especially the large proportion of light and medium-sized passenger vehicles throughout the day. There are also transit highways running through the region, and the models driving on transit highways are mainly heavy trucks, which are generally diesel-powered and have low emission standards, high average trip mileage, high single-vehicle emission factors, and high degradation, and contribute more to NOx and PM [4]. Thus, it is clear that reasonable control of road motor vehicle traffic, improvement of emission standards for road mobile sources and non-road mobile sources, and control of heavy trucks are important ways to improve urban air quality.

3.2. Traceability Analysis

Figure 8 shows the time-to-time sharing rate trends of four mobile sources at the People’s Park site, the highest AQI monitoring site in the study area on a certain day. As can be seen from the figures, the emission sharing rate of NOx is the largest, with a single-day average sharing rate of about 54%; followed by CO, with a single-day average sharing rate of 49.2%; the pollutant with the smallest emission sharing rate is PM₁₀, with a single-day average sharing rate of 43.5%. Therefore, according to the emission characteristics of mobile sources, priority should be given to the control of heavy-duty diesel trucks with high contribution to NOx emissions and non-road mobile machinery with low emission stages. The trend of mobile-source emission sharing rates at this air monitoring site is similar to the time-by-time emission characteristics of mobile-source pollutants, which also shows a “bimodal M” shape, indicating that mobile-source pollutants in Zibo are mainly influenced by the pollutant emissions from road motor vehicles.

In order to verify the accuracy of the results of this study on mobile-source pollutant dispersion concentration measurements and traceability analysis at air monitoring stations, Pearson’s correlation analysis was conducted at several moments in the vicinity of the monitoring stations within 3 km using the traffic volume from the road traffic data mentioned in Section 2.2 and the contribution of mobile source pollutants at air monitoring stations. As shown in

Table 2. Pearson correlation analysis between traffic volume and pollutant share.

			CO Share Ratio	HC Share Ratio	NOx Share Ratio	PM Share Ratio
Pearson correlation			0.882 **	0.894 **	0.882 **	0.911 **
Sig.(0.01)			0.000	0.000	0.000	0.000
bootstrap sampling b	95% confidence interval	lower limit	0.777	0.794	0.771	0.820
		superior limit	0.945	0.954	0.950	0.964

** At 0.01 level, the correlation is significant. ^b Unless otherwise stated, self-service sampling results are based on 1000 self-service sampling samples.

, the results show that the correlation between traffic flow and CO share was 0.872, the correlation with HC share was 0.873, the correlation with NOx share was 0.882, and the correlation with PM share was 0.911 at 95% confidence interval, indicating that urban road motor vehicles were the main contributors of mobile-source pollutants. The emission measurement and dispersion simulation model constructed in this study can accurately reflect the traceability of pollutants at urban air monitoring sites.

4. Conclusions

This study takes mobile-source pollutants in Zibo, China as the research object, based on the combination of emission and dispersion localization models and existing models. The pollutant emission characteristics and traceability are considered in multiple factors, and the above analysis of pollutants is integrated to provide references for accurate regulation.

(1)

Emission experiments were based on vehicle-borne emission tests on typical roads and typical time periods in the city, combined with multi-source data such as road network data and road traffic data in Zibo; an urban localized mobile-source emission measurement model was constructed, meteorological data and air quality data were integrated, and a mobile-source dispersion model for different wind speed scenarios was constructed.

(2)

The spatial and temporal distribution characteristics of mobile-source pollutant emissions in Zibo City: (1) From the perspective of temporal distribution, the trend of mobile-source pollutant emissions is highly coupled with the distribution of urban residents’ working hours, and the change in traffic flow and the trend of daily average pollutant emissions are both “bimodal M” characteristics, with the morning peak and evening peak occurring at 7:00–9:00 and 16:00–18:00, respectively. (2) In terms of spatial distribution, during the peak hours, the road sections with higher emissions are mainly distributed in the denser areas of residential and commercial areas, and during the working hours, the pollutant emissions are mainly closely related to the non-road mobile machinery in the urban industrial areas.

(3)

Traceability characteristics of pollutants in Zibo City: Combined with meteorological factors, the actual dispersion concentration of pollutants at air monitoring stations is corrected based on the estimation method of residual concentration distribution of pollutants, and the results show that the emission-sharing rate of NOx is the largest, and the trend of the mobile-source emission-sharing rate at air monitoring stations shows a “bimodal M” shape, which is similar to the trend of traffic flow and the time-by-time emission characteristics of mobile sources, indicating that mobile-source pollutants in Zibo City are mainly influenced by the pollutant emissions from road motor vehicles.

(4)

For the mobile-source pollutant emission characteristics and traceability analysis in Zibo City, it is clear that heavy-duty diesel vehicles and light buses are the key vehicles for emission management in the morning peak hours. In addition, the industrial areas where non-road mobile machinery gather are the key regulatory areas for pollution prevention and control management. Subsequent studies will consider actual on-board emission experiments of non-road mobile machinery to establish a more accurate database of emission factors. The study provides data support and a theoretical basis for transportation industry management departments to carry out exhaust gas treatment of motor vehicles and non-road mobile machinery, and to make relevant decisions (delineating the scope of low emission zones) to improve the air environment.