Temporal And Spatial Distributions Of Pollutants At Urban Vehicle Emission Hotspots Based On Driving Trajectories

Motor vehicles have become the main source of the urban air pollution.Furthermore,as an emission source closer to the human respiratory zone,they present a more significant impact on the human health.Intersections and congested segments on urban expressways are two typical emission hot-spots for urban roads.The study of vehicle emission mechanisms and spatiotemporal characteristics of pollutants for such hot-spots will enhance to the analysis of the concentration distribution of pollutants at hot-spots so as to develop refined emission control measures for reducing emissions or the exposure of pollutants to human.Existing studies related to emissions at intersections and congested segments have focused on investigating the influencing factors of total emissions often by such emission models based on average speeds or driving cycles.However,few studies have been found in the literature on estimating emissions based on real-world vehicle trajectories,especially on the analysis of spatial and temporal distribution of vehicle emissions at emission hot-spots.In addition,the reliable second-by-second emission data of buses are scare,while the second-by-second emission data on light-duty passenger cars are abundant today,which can better represent the characteristics of traffic flows.Further,a high consistency of change patterns was found between emission rates of vehicle emissions(CO,HC,PM and NOx)and the vehicle specific power(VSP),while a secondary reaction does take place for PM and NOx when they disperse in the air.In this context,the objective of the research in this dissertation is to,by taking CO as the target study pollutant,develop spatiotemporal distribution models of vehicle emissions for intersections and congested segments on expressways based on real-world second-by-second activity data and emission test data for light-duty gasoline passenger cars,realizing a dynamic description of characteristics of pollutants at the hot-spots with a high spatiotemporal resolution.The study in the dissertation includes the following main contents:(1)A quantitative method for estimating emissions at vehicle emission hot-spots based on vehicle trajectories was developed.A huge amount of data were collected including real-world second-by-second activity data of passenger cars on the road network in Beijing,emission test data of light-duty gasoline passenger cars of China V,and the floating car data of weekdays in Beijing.After a quality control of the collected data,the trajectory scenarios and their characteristics were extracted.On this basis,a quantitative method for estimating emissions at vehicle emission hot-spots based on vehicle trajectories was developed.The method includes the analysis on dividing VSP bins,calculation of emission rates for each VSP bin,quantitative method of estimating emissions and emission indicators based on VSP distributions of classified trajectories for emission hot-spots.The emission intensity in the indicators was the output of the spatiotemporal distribution model of emissions.This study proposed a calculation method of the emission intensity based on the traffic density and emission rates.(2)An emission factor prediction model under multiple influencing factors at intersections was developed.For the whole intersection as well as the upstream and the downstream of the intersection,the study analyzed the impact of road types,the number of stops,turning movements,and delays on VSP distributions and emission factors,and then,determined key influencing factors.Further,the emission factor prediction models were developed based on delays and the number of stops(or turning movements)for the whole intersection and the upstream or downstream of the intersection by employing the regression model that consists of both qualitative and quantitative variables.The results showed that the emission factor had a linear relationship with the delay with the R~2 above0.89 and the models passed the test.(3)A spatiotemporal distribution model of vehicle emissions at intersections was developed based on driving trajectories.Through an analysis of trajectory characteristics under different operating conditions at intersections,the trajectories were divided into five categories,the acceleration trajectory,deceleration trajectory,non-stop trajectory,trajectory between two stops,and idle trajectory.Then,a method for estimating the spatial distribution of vehicle emissions at intersections based on classified trajectories was proposed.Further,a traffic density estimation method was proposed based on the classified trajectories,followed by the development of the temporal and spatial distribution model of vehicle emissions at intersections based on the traffic density and the classified trajectory-based average emission rates.The fleet dissipating acceleration zone was found to be the zone with the highest emission intensity at intersections,whose average emission intensity was as twice as that of the idle zone.The highest point of the intersection emission intensity was the vehicle starting segment,with a moving speed equal to the wave speed of the fleet dissipating wave.(4)A spatiotemporal distribution model of vehicle emissions on congested segments of the expressway was developed based on driving trajectories.The congestion propagation and dissipation type on the expressway classified based on the floating car data.The vehicle trajectories of congested segments were divided into four categories,leaving the congestion,entering the congestion,in the congested segment,and in the non-congested segment.The classified trajectories were further grouped based on the VSP distributions and emission characteristics of classified trajectories.Then,a power function relationship between the acceleration and the speed of the acceleration trajectory was developed.On this basis,the spatiotemporal distribution model of vehicle emissions on congested segments of expressways was developed,which described the temporal and spatial distribution characteristics of vehicle emissions on the congested segments.The congested segments were ranked by the emission intensity of the spatial-temporal zone,from highest to lowest,accelerating for leaving the congestion,being in the congestion,in the non-congestion,and decelerating for entering the congestion.The emission intensity decreases gradually as the vehicle speed increases for leaving the congestion.The highest point of the emission intensity was the segment where the vehicle started for acceleration,and the moving speed of the highest point was equal to the wave speed of the bottleneck dissipating wave.For a comparison,the emission estimation method based on average speed was applied.The results showed that the average speed model particularly underestimated the emissions on the downstream road of the bottleneck where vehicles intensively accelerate.(5)In the case study,the AERMOD atmospheric dispersion model was used to simulate the concentration distribution of pollutants for the intersections and congested segments of the expressway.The study analyzed the pollutant concentration distributions and their forming mechanism under the factors such as wind directions,calculation methods of emission intensity,time interval,and traffic volume at intersections.The results showed that the concentration distribution was significantly affected by the calculation method of the emission intensity.Finally,a quantitative method predicting the pedestrian exposure at intersections was proposed.The application examples showed that the traditional calculation method of the emission intensity led to the pedestrian exposures that were underestimated by 8%-31% at intersections.