Source Apportionment: Principles and Methods
Published:18 Aug 2016
J. G. Watson, J. C. Chow, L. A. Chen, G. Engling, and X. L. Wang, in Airborne Particulate Matter
Download citation file:
Receptor model source apportionment has been facilitated by the availability of particulate matter (PM) speciation networks that measure elements, ions, and carbon fractions, and the availability of effective variance (EV)- and positive matrix factorization (PMF)-chemical mass balance (CMB) solutions to identify and quantify source contributions. However, receptor modeling software is too often applied without a thorough evaluation of the results. Quantitative source contribution estimates derived from these solutions must be challenged as part of a larger modeling and data analysis effort that supplies a “weight of evidence” for the major contributors. PMF-derived source factors should be compared with measured source profiles to identify potential source mixing within a factor and collinearities among factors. EV-CMB solutions should justify the use of measured profiles from other areas as representing those in the study area. Cost-effective methods exist to obtain more relevant source profiles that better represent the potential contributors. As pollution controls reduce primary emissions, elemental source markers and elemental carbon are becoming less useful for distinguishing among source types. Much more information can be obtained from speciation network filters at minimal additional cost to provide more specific markers related to important source types, such as solid fuel combustion for heating and cooking and secondary organic aerosol contributions. Receptor models have been productive for identifying sources, quantifying their contributions, and justifying regulations for residential wood combustion and cooking emission reduction strategies. When used as complements to source-oriented models and emission inventory development, air quality management practices can more accurately allocate pollution control resources.