Long-range sources of Toronto particulate matter (PM2.5) identified by Aerosol Laser Ablation Mass Spectrometry (LAMS)

An aerosol laser ablation mass spectrometer (LAMS) was operated continuously in downtown Toronto for a period of 11 months. The mass spectra of individual particles were classified according to their chemical composition using neural net-based adaptive resonance theory (ART-2a). The types of particles thereby identified were used as the basis for classifying particles from a 10-day sampling period within the 11-month sampling study. The particles within a class were further segregated by size, and the number of particles analyzed per hour was calculated for each of the resulting composition and sized based classes. ART-2a was used to combine these classes into groups based on similarity in their temporal trends. Two unusual sources were thereby identified: forest fire particles and intercontinental dust. The forest fire particles primarily exhibited potassium and potassium sulphate, while the intercontinental dust exhibited crustal metals, like titanium, aluminum, and iron. Back trajectory analysis of the air masses that transported these particles to Toronto revealed their origins as Northern Canada and Africa. From the chemical composition and size of the particles, a forest fire episode in Western Canada was identified during the 11-month sampling study on 1 June 2002. Other episodes of particles originating from Africa were not observed. These findings demonstrated the use of temporal variations in single particle chemistry to identify sources.