PM(10) source characterization at urban and highway roadside locations.

Levels of PM(10) were measured at two different roadside locations in the Stockholm region in Sweden, one highway south of Stockholm and one urban street canyon in the center of the city. PM(10) samples were taken during six separate campaigns over one full year, and analyzed for 29 metals, in order to help characterize sources of PM(10). Five contributing factors were identified by multivariate receptor modeling using positive matrix factorization. Factors were classified, based on their seasonal variation and published data on metal composition of different sources, as: 1) resuspension; 2) vehicle derived; 3) road salt; 4) regional combustion and 5) long-range transport. Resuspension and long-range transport were shown to be important contributors to the PM(10) levels at both sites. In fact, long-range transport was the main contributor to the PM(10) levels at the highway roadside. The vehicle source was only of major importance at the urban roadside, where it frequently contributed between 10 and 20 microg m(-3). Brake wear was an important component in the vehicle source. Vehicle exhaust was not detected as a separate source and was not identified as a major source for PM(10). To our knowledge, this is the first study identifying brake wear as a major source of PM(10) during urban driving.