The development and implementation of a mesoscale modeling system for simulating the meteorology and air quality situation in the Mae Moh Valley

The Mae Moh Valley region of northern Thailand experiences frequent pollutant fumigation events during the annual cool season. The onset and magnitude of these events are driven by the synoptic scale and mesoscale conditions that develop over the Valley during the cool season. A conclusion from previous studies conducted in the Valley was that to properly predict the onset and magnitude of the fumigation events, a three dimensional wind field generated using a mesoscale meteorological model needed to be used in a mesoscale transport and dispersion model. The results of the previous studies led to the modeling analysis presented in this dissertation.; The research hypothesis was that it would be possible to develop a mesoscale dispersion modeling system that could simulate the Valley fumigation events. The null hypothesis was that the air dispersion modeling system could not simulate the fumigation events. The Penn State/NCAR Mesoscale Meteorological Model, Version 5 (MM5) was used to generate the mesoscale meteorological parameters used as input to the United States Environmental Protection Agency (USEPA) CALMET/CALPUFF mesoscale dispersion and transport model. These models were used to forecast one fumigation event observed in 1997. Three model scenarios were considered for the fumigation event. The differences in the model scenarios were a function of modifications to terrain and horizontal and vertical grid resolutions used by MM5. Output from the combined MM5/CALMET/CALPUFF was compared with observations to document the modeling system's strengths and limitations.; The results from this study indicated that MM5 appeared to be capable of simulating the temperature profile required to produce a fumigation event in the Valley. However, due to errors in the input meteorological data, MM5 was not capable of forecasting the light and variable wind conditions present within the Valley prior to and during the fumigation events. These wind field errors contributed to errors in the CALMET and CALPUFF calculations. The theoretical procedure for this modeling system appears to be correct. However, more research is needed to determine what corrections need to be made to the input data and models to improve the fumigation simulation within the Valley.