Comprehensive sensitivity analyses on air quality model performance for PM_2.5 simulation

The urban air quality model inter-comParison study in Japan (UMICS) was conducted in order to improve the Community Multiscale Air Quality model (CMAQ) performance of the for PM_2.5 simulation. UMICS consist of three phases including the first phase focusing on elemental carbon, the second phase focusing on major ionic components (sulfate, nitrate and ammonium) and the third phase focusing on organic aerosol (OA) (UMICS1, 2 and 3). The results of UMICS3 for improvement of substantial OA underestimation were described in this paper. Because primary OA accounted for most of OA in the participating models, changes in volatile organic compounds emissions caused only slight changes in OA concentrations. Meanwhile, additional primary OA emissions because of a large amount of semi-volatile organic compounds and condensable organic compounds emissions substantially increased OA concentrations. The results emphasized the importance of emission sources that were not considered in the existing emission data. In addition, sensitivity analyses on various processes including meteorology and emission were conducted in order to show errors of PM_2.5 simulation originating from model input data and configurations. The results indicated that the importance of model input data is comparable to, or greater than that of model configurations in improvement of model performance for PM_2.5 simulation.