Quantification of the effects of molecular marker oxidation on source apportionment estimates for motor vehicles

Molecular markers are individual organic compounds used in receptor models to apportion fine particulate matter to sources. These models currently assume that molecular markers are chemically stable; however, recent laboratory experiments suggest they may be significantly oxidized on atmospherically relevant time scales. To investigate the effects of photo-oxidation, we extended a 3-D chemical transport model (PMCAMx) to simulate norhopane concentrations over the eastern United States during July 2001. Norhopane is an important molecular marker for motor vehicle exhaust. We examined eight different simulation scenarios, using different combinations of reaction rates and source profiles. The simulations including norhopane oxidation better reproduced the observed spatial patterns of norhopane concentrations than the non-reactive cases. Chemical mass balance (CMB) analysis was performed using the PMCAMx-predicted motor vehicle norhopane and elemental carbon (EC) concentrations to quantify the bias caused by oxidation on source apportionment estimates. Norhopane oxidation caused CMB to underestimate total vehicle OC by 10–50%, with larger biases in rural areas. This underestimation was largely due to changes in the amount of OC apportioned to gasoline vehicles which was reduced by as much as 100%. The OC apportioned to diesel vehicle emissions was relatively insensitive to norhopane reaction. Therefore, oxidation can substantially alter CMB estimates regarding the relative importance of gasoline and diesel vehicle emissions.