Airborne Particulate Matter: Sources, Atmospheric Processes and Health
Case Studies of Source Apportionment from North America
Published:18 Aug 2016
Special Collection: 2016 ebook collection , ECCC Environmental eBooks 1968-2022
Philip K. Hopke, 2016. "Case Studies of Source Apportionment from North America", Airborne Particulate Matter: Sources, Atmospheric Processes and Health, R M Harrison, R E Hester, Xavier Querol
Download citation file:
An important aspect of air quality planning is the identification of air pollution sources and their importance in contributing to the observed ambient conservations. Since the 1960s, there have been efforts to use the measured ambient concentrations and what is known about the nature of source emissions. The methods have been formalized into a set of techniques termed receptor models and they have been extensively applied to a variety of air quality problems. This chapter outlines the history of the application of source apportionment tools. A number of studies are highlighted that have been important in the development or adoption of source apportionment into air quality strategy development. For example, an early application of the chemical mass balance model in Portland, OR, led to improvements in their deterministic dispersion model and enabled it to more accurately reflect the source/receptor relationships in this city. Positive matrix factorization developed in the early 1990s has now become the most widely used receptor model and provides a flexible approach to apportion pollution sources using only the ambient data. Such applications include conventional composition data, volatile organic compounds, particle size distribution data, and high time resolved data from systems like aerosol mass spectrometers or rotating drum impactors. PMF can now also incorporate external information like known source profiles. It is possible to develop conceptual models that align with the nature of the data such as composition as a function of particle size and composition or composition as a function of location and time across a large-scale monitoring network. Illustrative examples of this variety of applications are presented.