Methodological studies and applications of several new receptor models for air quality management

Two new receptor models, three-mode factor analysis (TMFA) and direct trilinear decomposition (DTD), have been developed for the purpose of air quality management. Another method, potential source contribution function (PSCF), has been applied to study source locations of pollutants in southern Ontario, Canada.PSCF is a probability function based on the air parcel trajectory data coupled with information regarding the nature of the contaminants measured in that air parcel. Regions with high PSCF values have a higher probability of contributing pollutants to the measured concentrations at the receptor site. The application of PSCF method to both precipitation and particulate samples collected by the network of Acidic Precipitation in Ontario Study (APIOS) produced a visually dramatic presentation of source locations and their seasonal patterns of acidic species in these samples.Both TMFA and DTD are three-mode models that analyze three-way air quality data sets. TMFA is a three-mode counterpart of traditional two-mode factor analysis. With TMFA, the three-way observed data set is decomposed into three two-way factor loading matrices and one three-way core matrix. A few causal factors in each mode can be identified by examining the corresponding loading matrix. The factors in the mode of chemical species are identified as sources, the factors in the mode of sites as "pollution zones", and the factors in the mode of sampling periods as "regimes". The interactions of sources, zones, and regimes are indicated by squares of the entries of the core matrix.DTD is a quantitative receptor model. The three-way data set is decomposed into three two-way matrices. The column vectors of each of the matrices are called "source profiles", "emission patterns", and "site coefficients", respectively. Particulate sources are identified by examining both their source profiles and emission patterns. Emission patterns and site coefficients can then be used to produce estimates of mass contributions of sources to the samples collected at every site in every period. Since DTD identifies sources based on variations in two dimensions, it has a higher capability to distinguish two sources that have similar chemical compositions.U of I OnlyETDs are only available to UIUC Users without author permissio