RESUMO
Accurate meteorological fields and applicable air quality models are important ways to optimize air pollution simulations. To improve the accuracy of winter air pollution models in the Sichuan basin, we conducted a meteorological field simulation using 25 sets of parameterized scheme combinations in the Weather Research and Forecasting (WRF) Model. Based on the optimal parameters, the air pollution levels were simulated using AERMOD and CALPUFF models in a local large steel plant, and the data were verified by comparing the data from four National Ambient Air Monitoring Stations (NAAMS). The results indicated that the WRF model parameters had substantial effects on the simulation of the ground wind field, high-altitude wind field, and ground humidity field. In contrast, the parameters had no significant effect on the simulation of the ground temperature field, high-altitude temperature field, and high-altitude humidity field. The combination of the SLAB land surface process scheme and Dudhia shortwave radiation scheme with four boundary layer schemes, namely YSU, ACM2, BouLac, and MRF, could well-simulate the trends of winter surface wind, temperature, and humidity fields in Sichuan basin. The simulation results were analyzed by combining the statistical parameters of high-altitude wind, temperature, and humidity. The group 1 parameter scheme was applicable to simulate the meteorological field of Dazhou. Group 13 and Group 17 parameters were applicable to simulate the meteorological fields in Chengdu during the day and night, respectively. The correlation between CALPUFF simulation and monitoring value was better than that for AERMOD. CALPUFF was more accurate than AERMOD when referring to the monitoring data from NAAMS No.3. In addition, the simulation quality of CALPUFF was slightly better than that of AERMOD with reference to data from NAAMS No.2. Using air pollutant monitoring data from NAAMS as a reference, the simulated results of CALPUFF on NOx and PM10 were improved compared to AERMOD at all four stations. Data from the Q-Q diagram indicated that the simulated results of CALPUFF on SO2, NOx, and PM10 were closer to the monitored values compared to those of AERMOD.