The Study Of Physics And Chemistry Schemes And Emission Inventory In Air Pollution Forecasting Model

The thesis firstly provided a brief introduction to an air quality forecasting system based on the meso-scale chemical-meteorology model WRF-Chem, and developed a new approach to improve the emission inventory with high temporal and spatial resolution. In an original grid of emission inventory, resolution could be improved by redistributing data based on high resolution population data in counterpart grid. Besides, after dealing with population data with smoothing operator for given times, emission data with high accuracy could be identified, which also kept spatial tendency of population data. For the temporal resolution, hourly change of original data based on4different categories was used in order to identify hourly change of result data.A heavy precipitation process in East China during July23-242009is simulated to conduct a sensitivity experiment with WDM6-and Morrison-schemes in WRF model. The accumulation precipitation, rain rate and the microphysical cloud-parameters are analyzed in two simulations to assess the two microphysics schemes in precipitation prediction. Compared to the observation of the heavy rainfall, the under-prediction of precipitation and the earlier rain start are simulated by WRF-model with both cloud microphysics schemes. However, the precipitation prediction with the Morrison scheme is more reasonable, while the WDM6scheme produced a higher error of more than50%compared to the observation. Due to the definition of raindrop of WDM6scheme, more raindrops in small sizes than the reasonable raindrop size distribution were modeled. Based on a calculation, it is found that the unreasonable definition of raindrops in WDM6scheme might lead to an overestimation of evaporation and an underestimation of sedimentation for the raindrops, resulting in the under-prediction of accumulated precipitation on the ground. The extended experience also made a conclusion that Morrison Scheme could provided a more available forecasting accuracy.In order to identify sensitivity of emission data redistribution and processes of aerosols feedback, a heavy hazy pollution case, happened around Nanjing in December2013, was simulated and analyzed. Comparing with experience(WHWD) with modified emission inventory, the results showed that experience(WHWDWF) with both modified emission inventory and aerosols feedback processes could improve accuracy of aerosols forecasting in a large extent. For the NO2, experience(WHWDWF) could reflect the heavy pollution during hazy period but overestimate the low level after this case. For the O3, experience(WHWD) could not provide evident modification to final forecast, while experience(WHWDWF) give a little improvement. Besides, in a relatively dry environment, nucleation process of aerosol particles suffered limitation and mass concentration of cloud water in experience(WHWDWF) was obviously lower than experience(WHWD), which also led to a smaller accumulated rainfall on the ground. High aerosol mass concentration in experience(WHWDWF) was due to the limited precipitation, which is more similar to observation data. On the other hand, as a result of relatively strong wet scavenging process, aerosol mass concentration in experience(WHWD) was relatively low.