Comparing emission inventories and model-ready emission datasets between Europe and North America for the AQMEII project

This paper highlights the similarities and differences in how emission inventories and datasets were developed and processed across North America and Europe for the Air Quality Model Evaluation International Initiative (AQMEII) project and then characterizes the emissions for the two domains. We focus specifically on the creation of “model-ready” gridded emission datasets for 2006 across the two continental study domains. The practice of creating and processing the two inventories is discussed with a focus on emission factors, spatial allocation, temporal variability, speciation of PM and VOCs, and the mechanics of distributing the data and supporting emission algorithms to the modeling community. The spatial and temporal distribution on common scales is compared for the pollutants of primary concern: NOx, VOCs, SO2, PM2.5, CO, and NH3. Because of differences of population distribution, emissions across North America tend to be more heterogeneous in spatial coverage than in Europe. The temporal patterns in the estimated emissions are largely the result of assumptions used to characterize human activity, with the exception of “natural” emissions, which are modulated by meteorological variability, and emissions from large electric generating units in the U.S., which have the benefit of continuous emission monitors that provide hourly resolved profiles. Emission estimates in both study domains are challenged by several important but poorly characterized emission source sectors, notably road dust, agricultural operations, biomass burning, and road transport. Finally, this paper provides insight on the strengths and weaknesses of emission inventory preparation practices on both continents. One important outcome of this comparison of 2006 emissions between Europe and North America is the greater understanding provided into how the emission estimates developed for the AQMEII project impact regional air quality model performance.