| Air quality prediction is important for pollution control and environmental management. It is also useful for assessing environmental effectiveness of regional economic development. Air quality prediction models usually play an important role in developing effective air pollution control plans.; In this dissertation, multi-box models were developed for air quality prediction in an urban area. Multi-box models improve upon the conventional single-box models through consideration of more details in spatial variations of source distributions and meteorological conditions, as well as their impacts on the ambient environment. The models can also reflect physical and chemical removal mechanisms and effects of wind-direction variations, with four wind-direction groups being introduced into its computational framework. Information related to emission sources and meteorological conditions in a number of temporal and spatial units was used as inputs for formulating the prediction models. Data from an airport's meteorological station were collected and analyzed for estimating mixing heights of the study area.; A multi-source and multi-grid Gaussian (MSMGG) model, based on the Gaussian plume model, was also developed to effectively simulate the spatial distribution of pollution impacts from multi-point source emissions. The MSMGG model was then incorporated within the three dimensional multi-box (3DMB) modeling framework. This hybrid 3DMB and MSMGG modeling approach can not only effectively simulate impacts of both low- and high-stack sources, but also reflect more details of spatial variations in source distributions and meteorological conditions.; The developed models were employed for predicting monthly average SO2 concentrations in different zones of the Shijiazhuang City, China. The prediction results were verified through two monitoring programs where SO2 concentrations at both ground and upper levels were detected. Compared to simple 3DMB model. The hybrid 3DMB-MSMGG model can simultaneously attain higher prediction accuracy and produce more informative modeling outputs. The modeling results can provide useful bases for the planning of regional air quality management system. |
