CAMx4 Model Performance for Ozone in the UpperMidwest over the Summers of 2001, 2002, and 2003

摘要

The United States Environmental Protection Agency (EPA) recently promulgateddesignations for the 8-hour ozone National Ambient Air Quality Standard on April 15,2004 (69 CFR 23858, April 30, 2004). State Implementation Plans will include resultsfrom 3-D Eulerian photochemical transport models. In the past, these models wereapplied to short episodes with a few days used to “spin-up” the model simulation toreduce the impact of initial concentrations. Advancement in computer technology nowallows for the application of photochemical models over large domains and entiresummer seasons. This approach will better account for regional transport and minimizeany influence from initial concentrations.Several photochemical models treat the physical processes and chemistry that formozone. The Comprehensive Air Quality Model with Extensions (CAMx4) is applied toJune through August for 2001, 2002, and 2003. The photochemical modeling domainconsists of a 36 km grid over the Central and Eastern United States and a two-way nested12 km grid over the Upper Midwest.Operational evaluation for ozone modeling purposes includes evaluating model estimatesagainst observed 8 hr averaged ozone. The summers of 2001, 2002, and 2003 will beused for ozone strategy development to capture the variety of high ozone episodes thatoccur across the Upper Midwest.The modeling system predicts the spatial patterns of ozone concentrations over the LakeMichigan and Ohio Valley region quite well for all 3 summer seasons. Interestingregional spatial biases are also noted in the modeling, particularly in the Chicago area.The modeling system tends to over-predict low concentrations of 8 hr ozone and slightlyunder-predict the highest concentrations of 8 hr ozone in the Upper Midwest. Thevariability in performance from one summer to the next is minimal. Performancedegradation in particular regions of the domain seem more influenced by static inputs,such as emissions, than the changing meteorological inputs. Overall, the modeling systemperforms within EPA guidelines for the metrics used to characterize performance.