Implementation of process analysis in a three-dimensional air quality model

Gridded, regional air quality models are used extensively in the development and evaluation of air quality regulations and in reconciling theoretical predictions of atmospheric behavior to observational data. Therefore, understanding the response of air quality models to changes in emissions or atmospheric conditions is important both for improving understanding of atmospheric processes and for designing effective air quality policies. Diagnosing model behavior is difficult because regional air quality models generally only output data on environmental state variables such as concentrations, and not on the rates of the processes that control the state variables. This thesis describes the development and application of Process Analysis Post Processing Tools (PAPPTs) for photochemical grid models; these tools provide dynamic information on environmental processes, such as horizontal and vertical pollutant fluxes crossing cell boundaries, chemical production and consumption rates, emission rates, and deposition rates, as well as initial and final concentrations. The PAPPTs were implemented in a regulatory photochemical grid model called the Comprehensive Air Quality Model with extensions vii (CAMx). The use of the tools was demonstrated through the analysis of physical and chemical processes in regulatory models in central California and southeast Texas. The PAPPTs revealed the processes that contributed to model under-prediction of ozone generation in the Fresno area, and the processes that determined the impact of industrial emission events on ozone formation processes in the Houston-Galveston area. The PAPPTs also were used to develop a sub-domain model to evaluate stochastic emissions, to examine the impact of wild fire plumes on ozone formation processes, and to evaluate the isoprene emission inventory in southeast Texas. These case studies demonstrate the utility and versatility of the PAPPTs.