Application of optimally scaled target factor analysis for assessing source contribution of ambient PM10.

Speciated coarse particulate matter (PM10) data obtained at three air quality monitoring sites in a highly industrialized area in Spain between 2002 and 2007 were analyzed for assessing source contribution of ambient particulate matter (PM). The source apportionment of PM in this area is an especially difficult task. There are industrial mineral dust emissions that need to be separately quantified from the natural sources of mineral PM. On the other hand, the diversity of industrial processes in the area results in a puzzling industrial emissions scenario. To solve this complex problem, a two-step methodology based on the possibilities of the Multilinear Engine was used. Application of positive matrix factorization to the dataset allowed the identification of nine factors relevant to the study area. This preliminary analysis permitted resolving two mineral factors. As a second step, a target rotation was implemented for transforming the mineral factors into experimentally characterized soil resuspension and industrial clay sources. In addition to improving the physical interpretation of these factors, the target rotation reduced the errors arising from the rotational freedom of the solution and the multicollinearity among sources. In this way, the main primary industrial emissions of PM in the zone were identified by this target factor analysis. A marked decrease was observed between 2002 and 2007 for the contributions of industrial sources coinciding with the implementation of mitigation measures in their processes. This study supports the utility of source apportionment methodologies for quantitatively evaluating the effectiveness of the abatement programs for air quality improvement.