Influence of long-range-transported pollution on the annual and diurnal cycles of carbon monoxide and ozone at Cheeka Peak Observatory

[1] We assess the importance of distant pollution sources on the marine background by combining measurements of carbon monoxide (CO) and ozone (O3) with model simulations from the GEOS-CHEM chemical transport model and two types of back trajectories. Measurements were made over a complete annual cycle from March 2001 to May 2002 at Cheeka Peak Observatory (CPO) in Washington State. Data from an earlier campaign (March–April 1997 and 1998) were also incorporated. The seasonal cycles of CO and O3 show a spring maximum and a summer minimum, consistent with other remote sites in the marine boundary layer. European and Asian pollution emission sources of CO, parameterized by GEOS-CHEM, were grouped into one category called ‘‘LRT,’’ representing ‘‘long-range transport.’’ LRT values produced good correlations with measured CO (monthly averaged r = 0.60, 6-hour averages) and were found to comprise between 15 and 60% of total CO, with higher values in winter/spring and lower values in the summer/fall. CO and O3 observations were classified on the basis of these LRT values, creating two general categories: ‘‘more polluted’’ (LRT > 75th percentile from monthly distribution) and ‘‘less polluted’’ (LRT < 25th percentile). The difference between these two categories is a measure of the net Asian enhancement above the background and reaches a maximum in the spring of 30-ppbv CO and in the spring-fall of 7-ppbv O3. The Asian O3 enhancement peaked at an average of 4 ppbv from April to October, which is the time of year of greatest urban air quality concerns in North America. Numerous long-range transport ‘‘events’’ (defined by at least 24 continuous hours of LRT > 75% percentile) were observed throughout the year. Ozone behavior was quite complicated during these events, producing no consistent enhancements. Classification of the measurements with LRT values from GEOS-CHEM is compared to back trajectories, both isentropic and kinematic. Isentropic trajectories agree reasonably well with GEOS-CHEM between February and May in capturing the CO enhancements in air masses influenced by recent emissions but were found to be less useful from June to January. HYSPLIT kinematic trajectories were slightly better at capturing the Asian pollution events than isentropic trajectories but still missed an event during the summer completely that was verified by GEOS-CHEM. Possible reasons for poor trajectory performance in the summer and fall are discussed. Long-range transport appears not to affect the magnitude of the diurnal variability of O3 measurements at CPO, which reflect daytime production of 2–4 ppbv/d. Rather, ship traffic in the vicinity of CPO, which supplies NOx through combustion, likely plays a dominant role.