Evaluation Of PM2.5 In Chicago ByChemical Mass Balance And Positive MatrixFactorization Models

Abstract

Chemical Mass Balance (CMB) and Positive Matrix Factorization (PMF) models were used to analyze fine particulate matter data from two sites within the city of Chicago. Measurements of metals, organic and elemental carbon, sulfate, nitrate, and gaseous criteria pollutants from the PM2.5 speciation network were evaluated. CMB and PMF results were both strongly influenced by the measurement uncertainty. Variables with a high percentage of measurements below the detection limit were heavily down-weighted so that the models would not be overly influenced by low or unknown concentrations. Variables that were usually above the detection limit were weighted by the root mean square average of 10 % of the measured concentration and the corresponding detection limit. The analysis yielded a nine source CMB and a 10 factor PMF solution for the Chicago sites. Sources represented by the factors were identified using established source profiles from literature and mass to mass ratios of species. The sources identified included secondary sulfate and nitrate, motor vehicles, coal-fired utilities, vegetative burning, wind blown dust, salt used to de-ice roadways and steel production. CMB and PMF predictions for source contribution and composition were compared and contrasted. The two models provided remarkably consistent results. The estimated daily contributions from each source revealed seasonal patterns which also aided in source identification.