Abstract
The inter-comparison of chemical transport models (CTMs) of four research groups was conducted for O3 and PM2.5 simulations in the Kanto area in summer 2007. Three groups used CMAQ v4.6 and the other employed CMAQ v4.5; each group used different emissions data. The results clearly showed that all the CTMs reproduced well the temporal variation of O3 (r > 0.5 at all four sites), but all the CTMs underestimated PM2.5 concentrations and did not reproduce the PM2.5 temporal variation (r <0.5) at suburban sites. Among the PM2.5 species, the concentration and temporal variation of SO42- was predicted well by all the CTMs (within a factor of 2 and r > 0.5 in most cases). By contrast, the elemental carbon (EC) and organic aerosol (OA) concentrations were underestimated by all the CTMs. The predicted NO3- concentrations showed a high variability (up to a factor of 10) among the four CTMs. Differences in the predicted concentrations of the primary products were explained well by the differences in their emission rates. However, differences in the secondary products, including O3, TNO3 (= HNO3+ NO3-), and the secondary organic aerosols, were not explained by the differences in the precursors' emission rates or in boundary conditions.
