Abstract
The purpose of this investigation is to understand the characteristics and the sources of TSP, PM2.5~10 and PM2.5 in Hsinchu Science Based Industrial Park, which is a typical high-tech related industrial area. An intensive sampling program has been conducted in this work. Airborne particulate samples were collected at different air quality monitoring stations from September 15 to 25 in 1999. The collected filter samples were then analyzed to determine the chemical compositions of various particulates. The results showed that the most abundant species in TSP were SO42-, NO3-, Cl-, NH4+, Al, Fe, and Ca, and the most abundant species in PM2.5 and PM2.5~10 were Fe, K, Ca, Al and Mg respectively. It was observed that the concentrations of SO42-,NO3-,Cl-,As and Se in the park are much higher than that of other area, which may be mostly contributed from the semi-conductor manufacture process. Various mathematical models has been utilized to identify the sources and extent of pollution in the area. The ISCST3 (Industrial Source Complex Short Term) dispersion model was employed to simulate the TSP and PM10 concentration in the air and compared with the measured values. Enrichment factor analysis indicated that the TSP in the area were mainly from marine and other industry sources. The VARIMAX-rotated principal component analysis identified the TSP sources included soil, road dust, construction activity, traffic and sea-salt spray respectively. On the other hand, the PM2.5~10 sources were soil, road-dust and construction activity, and the PM2.5 sources were traffic and sulfate. The results of Chemical Mass Balance (CMB) receptor model showed that source contributions to TSP at three different stations were soil (18~45%), road-dust (1~17%), nitrate (6~9%), marine spray (4~7%) and diesel (2~8%). It was found that, 37~52% unknown contribution can still be obtained from a CMB receptor model or a factor analysis. Obviously, the establishment of an comprehensive source profile for high-tech related industry is necessitated to apportion the contributions of different sources.