Mathematical Modeling of Photochemical Air Pollution

A model is presented for predicting the dynamic behavior of both inert and chemically reacting air pollutants in an urban atmosphere. Pollutants of interest include carbon monoxide, nitric oxide, nitrogen dioxide, ozone, and high and low reactivity hydrocarbons. The model is developed for a general urban area and then applied to the Los Angeles airshed. To evaluate the model, six smoggy days occurring in Los Angeles in 1969 are simulated followed by a comparison of the predictions with the numerous air quality measurements reported by the local air pollution control agencies. In addition, since evaluation of emission control strategies is an important potential use of the model, simulation results are also given demonstrating the possible impact on air quality in the Los Angeles area resulting from implementation of the U.S. Environmental Protection Agency's control strategy for the South Coast Air Basin proposed in July of 1973. The governing equations of the model are based on the general mass conservation relationships for a chemically reactive species in a three-dimensional, turbulent atmospheric boundary layer. These equations are solved independently of the coupled Navier-Stokes and energy equations; meteorological parameters in the model are estimated from measured wind and mixing depth data. Reaction rate expressions for the reacting species are derived from a concise, fifteen step kinetic mechanism describing the important chemical reactions that take place in the atmosphere. A detailed discussion is included of the treatment of the winds and temperature inversion in the Los Angeles area. Since pollutant emissions are an essential input to the model, a general model and comprehensive inventory of contaminant emissions for Los Angeles is presented. Considered in this study are the spatial and temporal distribution of emissions from motor vehicles, aircraft, and fixed sources (including power plants, refineries, etc.). Because of the nonlinear nature of the chemical kinetics, the governing equations themselves are nonlinear. Thus, a finite difference scheme based on the method of fractional steps is developed to solve the equations. A detailed exposition of all difference equations and their manner of solution is given. As mentioned previously, an evaluation study of the model is performed by making extensive comparisons of predictions and measurements for Los Angeles. It is found that local pollutant sources near monitoring stations can hamper the comparison of spatially averaged predictions with point measurements. In an effort to account for sub-grid scale effects, a "microscale" model is developed to predict the concentration elevation observed at a monitoring site caused by local sources, such as traffic on nearby streets and freeways. In general, it is found that the model predicts pollutant concentrations reasonably well considering the complexity of the problem and the uncertainty in many of the parameters in the model. Finally, possible future applications of the model are discussed, including the short range forecast of pollutant concentrations (say up to a few days in the future) and the evaluation of the impact of regional planning decisions and emission control strategies on air quality. To gain experience in applying the model to evaluate emission control strategies, simulations of the Los Angeles region are performed using an emission inventory based on control measures proposed by EPA for the region in 1977. The model results indicate that ozone concentrations will be reduced. substantially from 1969 levels; however, definitive statements with regard to air quality in 1977 must await further refinement of the model and better understanding of the effects of the various proposed control strategies on pollutant emissions.