Near-roadway air quality modelling requires detailed inputs about traffic, vehicle emissions, and atmospheric dispersion in order to capture microscopic features of each phenomena. However, these inputs are often a major source of uncertainty in resultant air quality estimates. We aim to evaluate the impact of different traffic behavior patterns on vehicle emissions and the subsequent effect of these emissions on air quality estimates derived from dispersion models. Two approaches were developed, an “Aggregate” approach, where the behavior of ensembles of vehicles were simulated, and an “Instantaneous” approach, where individual vehicles were tracked through the network. We also developed a dynamic integrated approach to include traffic patterns, vehicle-induced emissions, and air dispersion, and compared the contribution of each to air quality simulations. Our results indicate that the Aggregate approach tends to overestimate emissions. Validation of NO₂ concentrations generated from the dispersion model against measurements revealed a correlation of 61% and 79% for the Aggregate and Instantaneous approaches, respectively. Compared to the more commonly used Aggregate approach, modelling the individual behavior of vehicles provides a marked improvement on emissions and subsequent air quality estimates. Development of correction coefficients could improve Aggregate emission models based on the Instantaneous approach under different traffic situations.

近巷空气质量建模需要有关交通，车辆排放和大气扩散的详细输入，以便捕获每种现象的微观特征。但是，这些输入通常是导致空气质量估算不确定性的主要来源。我们旨在评估不同交通行为模式对车辆排放的影响，以及这些排放对从弥散模型得出的空气质量估算的后续影响。开发了两种方法，一种是“集合”方法，用于模拟车辆集合的行为；另一种是“瞬时”方法，其中通过网络跟踪单个车辆。我们还开发了一种动态集成方法，包括交通模式，车辆诱发的排放物和空气扩散，并比较了它们对空气质量模拟的贡献。我们的结果表明，总体方法倾向于高估排放量。验证NO从弥散模型相对于测量值生成的_2个浓度分别表明聚集方法和瞬时方法的相关性分别为61％和79％。与更常用的汇总方法相比，对车辆的单个行为进行建模可以显着改善排放量和随后的空气质量估算。修正系数的开发可以在不同交通情况下基于瞬时方法改进总排放模型。