 |
NARSTO Workshop on Innovative Methods for Emission Inventory Development and Evaluation University of Texas, Austin October 14-17, 2003 |
 |
Using Ambient Measurements To Critically Evaluate The Temporal Trends Of U.S. Carbon Monoxide Emission Inventories
David D. Parrish
Aeronomy Laboratory, National Oceanic and Atmospheric Administration Boulder, CO 80305
Carbon monoxide (CO) plays important roles in the photochemistry of the global troposphere. It provides a large fraction of the “fuel” for the photochemical production of ozone, and its reaction with the hydroxyl radical is the primary sink for that critical species. Global emission inventories of CO are derived from integration of the inventories of the world´s nations (http://arch.rivm.nl/env/int/coredata/edgar/). Comparison of these national inventories presents an enigma. Western Europe, with a population 40% greater than the U.S., reports CO emissions smaller by nearly a factor of 2. Further, Western European emissions reportedly are decreasing by 2.3 %/year, while U.S. emissions are thought to be nearly constant (0.2%/year increase). Given the large resources devoted to catalytic converters on U.S. automobiles (the primary source of CO), the lack of a significant decrease in U.S. emissions is surprising.
The U.S. EPA is a source of contradictory information regarding temporal trends of national CO emissions. One website (http://www.epa.gov/air/aqtrnd01/carbon.html) indicates a 6% increase (0.7%/year) from 1992 to 2001. In contrast, the 2000 U.S. EPA report “National Air Pollutant Emission Trends, 1900 – 1998” indicates a decrease of 1.9%/yr from 1985 to 1998 while the recent update to that report (http://www.epa.gov/ttn/chief/trends/trends01/trends2001.pdf ) shows a 2.9% decrease from 1985 to 2000. Our goal here is to use ambient measurements to discriminate between these conflicting inventories.
Measured ambient levels of CO have decreased at a significantly faster rate than any of the emission inventories. Urban CO levels decreased by 4.9%/year (based on the annual mean of the 2nd highest maximum 8-hr average for about 350 predominately urban and suburban monitoring sites nationwide) from 1985 to 1999. This corresponds to a decrease of more than a factor of 2 over this period. Hallock-Waters et al. [1999] report that rural CO levels in the U.S. Mid-Atlantic region decreased at approximately 5 ppbv/year from 1989 to 1997. This decrease is consistent with the decrease in the urban emissions. These trends in ambient measurements cannot be reconciled with the tabulated emission trends, and indicate errors in the latter. CO emissions, particularly from the mobile sources that dominate urban CO levels, evidently have decreased significantly faster than recognized in emission inventories.
Measured CO and NOx concentrations in U.S. urban areas show that the CO/NOx vehicular emission ratio has decreased at an average annual rate of 7 to 9% from 1987 to 1999 (i.e. a factor of ?3). This rate of decrease is at least twice that derived from mobile source emission inventories. These considerations all indicate that the rate of decrease of U.S CO emissions is underestimated in current emission inventories. We conclude that the control strategies that have been implemented, particularly catalytic converters on automobiles, are significantly more efficient at reducing CO emissions than recognized by the inventories.