Author/Authors :
vanderschelden, graham washington state university - laboratory for atmospheric research - department of civil and environmental engineering, Washington, USA , foy, benjamin saint louis university - department of earth and atmospheric sciences, Missouri, USA , herring, courtney washington state university - laboratory for atmospheric research - department of civil and environmental engineering, Washington, USA , kaspari, susan central washington university - department of geological sciences, Washington, USA , vanreken, tim washington state university - laboratory for atmospheric research - department of civil and environmental engineering, Washington, USA , jobson, bertram washington state university - laboratory for atmospheric research - department of civil and environmental engineering, Washington, USA
Abstract :
A multiple linear regression (MLR) chemical mass balance model was applied to data collected during an air quality field experiment in Yakima, WA, during January 2013 to determine the relative contribution of residential wood combustion (RWC) and vehicle emissions to ambient pollutant levels. Acetonitrile was used as a chemical tracer for wood burning and nitrogen oxides (NOx) as a chemical tracer for mobile sources. RWC was found to be a substantial source of gas phase air toxics in wintertime. The MLR model found RWC primarily responsible for emissions of formaldehyde (73%), acetaldehyde (69%), and black carbon (55%) and mobile sources primarily responsible for emissions of carbon monoxide (CO; 83%), toluene (81%), C2-alkylbenzenes (81%), and benzene (64%). When compared with the Environmental Protection Agency’s 2011 winter emission inventory, the MLR results suggest that the contribution of RWC to CO emissions was underestimated in the inventory by a factor of 2. Emission ratios to NOx from the MLR model agreed to within 25% with wintertime emission ratios predicted from the Motor Vehicle Emissions Simulator (MOVES) 2010b emission model for Yakima County for all pollutants modeled except for CO, C2- alkylbenzenes, and black carbon. The MLR model results suggest that MOVES was overpredicting mobile source emissions of CO relative to NOx by a factor of 1.33 and black carbon relative to NOx by about a factor of 3.
Keywords :
chemical mass , balance model , residential wood combustion , vehicle emission
