Comparison of several wood smoke markers and source apportionment methods for wood burning particulate mass

Abstract

Residential wood combustion has only recently been recognized as a major contributor to air pollution in Switzerland and in other European countries. A source apportionment method using the aethalometer light absorption parameters was applied to five winter campaigns at three sites in Switzerland: a village with high wood combustion activity in winter, an urban background site and a highway site. The particulate mass from traffic (PM sub(traffic)) and wood burning (PM sub(wb)) emissions obtained with this model compared fairly well with results from the super(14)C source apportionment method. PM sub(wb) from the model was also compared to well known wood smoke markers such as anhydrosugars (levoglucosan and mannosan) and fine mode potassium, as well as to a marker recently suggested from the Aerodyne aerosol mass spectrometer (mass fragment m/z 60). Additionally the anhydrosugars were compared to the super(14)C results and were shown to be comparable to literature values from wood burning emission studies using different types of wood (hardwood, softwood). The levoglucosan to PMwb ratios varied much more strongly between the different campaigns (4-13%) compared to mannosan to PM sub(wb) with a range of 1-1.5%. Possible uncertainty aspects for the various methods and markers are discussed.