A Simulation Approach to Assess Air Pollution from Road Transport

A hybrid deterministic/statistical modeling approach is used to predict air pollution levels generated from motor vehicle emissions. The inclusion of the effects of source emissions and meteorology on ambient concentrations are allowed for by the deterministic models, but are applied only within their range of greatest reliability. Two deterministic models are used in separate simulation exercises: the ATDL model to obtain an areal indication of pollution levels and the GM model to provide information on concentration reduction with distance from the roadway. Extrapolation of the reliable output data obtained from the deterministic models is allowed by the use of the statistical model to predict the maximum levels of pollution concentration. Monte Carlo simulation is included to show the effects of system uncertainty with respect to emissions. Variability in meteorology is allowed by running the model separately with a time-series input of wind speed for different calendar years. The methodology is demonstrated with an example in an Australian town that is being subjected to increasing heavy vehicle traffic due to mining operations in the area.