Key points are not available for this paper at this time.
Observations at sites in eastern North America show a strong correlation between O 3 and CO concentrations in summer, with a consistent slope ΔO 3 /ΔCO ≈ 0.3. Observations in the aged Denver plume at Niwot Ridge, Colorado, also show a strong correlation but with ΔO 3 /ΔCO = 0.15. These data offer a sensitive test for evaluating the ability of photochemical models to simulate production of O 3 over North America and its export to the global atmosphere. Application to the Harvard/Goddard Institute for Space Studies three‐dimensional, continental‐scale model shows that the model gives a good simulation of the observed O 3 ‐CO correlations and of the associated ΔO 3 /ΔCO. This successful simulation lends support to model estimates of 6 Gmol d −1 for the net O 3 production in the U.S. boundary layer in summer (corresponding to a net O 3 production efficiency of 5.5, which is the number of O 3 molecules produced per molecule of NO x consumed) and 70% for the fraction of the net production that is exported to the global atmosphere. Export of U.S. pollution appears to make a significant contribution to total tropospheric O 3 over the northern hemisphere in summer. Simple interpretation of observed ΔO 3 /ΔCO as an O 3 /CO anthropogenic enhancement ratio is shown to underestimate substantially anthropogenic O 3 production, because O 3 and CO concentrations are negatively correlated in the absence of photochemistry. It is also shown that concurrent observations of ΔO 3 /ΔCO and ΔO 3 /Δ(NO y ‐NO x ) ratios can be used to impose lower and upper limits on the net O 3 production efficiency.
Chin et al. (Wed,) studied this question.