A Thesis Presented To The Academic Degree Committee Of Nanjing University For Applying The Master’s Degree In Environmental Engineering

As the input to air quality modeling system, emission inventory is not only influential on the reliability of modeling results, but also an important basis for policy making abatement of air pollutant emissions. With the downscaled regional emission inventory directly applied, air quality simulation may not reflect the accurate characteristics of pollution at small spatial scale. Therefore, there is an urgent need to establish high-resolution emission inventories based on detailed information of local emission sources and to evaluate the reliability of emission inventories effectively for atmospheric chemistry research and pollution control management.In this study, a comprehensive database of pollution sources in Jiangsu was established based on the information compiled and corrected from various data sources including Environmental Statistics, Pollution Source Census and field survey on large emitters. With emission factors by sector incorprated, anthropogenic emission inventory of Jiangsu province was developed using a bottom-up method. The total emissions of SO2, NOx, CO, TSP (total suspended particulate), PM10, PM2.5, EC (elemental carbon), OC (organic carbon), CO2, NH3 and VOC (volatile organic compounds) were calculated as 1141,1642,7680,2606,1394,941, 57,140,862458,1101 and 1747 Gg in 2012, respectively. Emissions from point sources were estimated to account for 83.9%,71.2%,55.2%,82.7%,75.0%,63.8%,40.7%,30.9%,84.4%, 1.9% and 37.0% of total for those pollutants, respectively.The reliability and improvement of provincial emission inventory was evaluated by satellite observation and air quality simulation. In Jiangsu domain, the correlation coefficient between NOx emissions estimated at provincial level and summer tropospheric NO2 vertical column density (VCD) derived from Ozone Monitoring Instrument (OMI) was 0.534, slightly higher than that between national emission inventory and NO2-VCD (0.531). The results of correlation analysis indicate that the spatial distribution of provincial emission inventory was more consistent with that of NO2-VCD compared to national inventory. In particular, provincial emission inventory improved the emission estimation for small industrial pollution sources.Air quality modeling system, Models-3/Community Multi-scale Air Quality (CMAQ), was developed for southern Jiangsu, and national, regional and provincial emission inventories were respectively applied to evaluate the improved simulation from the provincial inventory. The normalized mean bias (NMB) of SO2, NO2, O3 and PM2.5 between ground observation and simulations from provincial, national and regional emission inventories were -43.64%--9.97%、-65.55%～48.45% and -53.93% -74.08%, respectively, and the normalized mean errors (NME) were 33.22%～47.49%、35.99%～76.53% and 43.45%～95.04%, respectively. Better simulation performance could be achieved when the improved provincial inventory prior to national or regional ones was used in the high-resolution air quality modeling.Based on the brute force method, the contributions of power, iron & steel, and cement plants to ground PM2.5 concentrations in Nanjing was quantified through air quality modeling using the provincial inventory. The dominating factors of ozone formation in the simulation domain were identified as well. The results could provide preliminary basis for air quality control in southern Jiangsu. The contributions of power, iron & steel, and cement plants to PM2.5 average concentrations in Nanjing were estimated at 6.0%,25.5% and 15.9%, respectively, indicating that the impact of emissions from iron & steel plants was significantly larger than those from power and cement plants. The ozone formation mechanism was VOC-limited in the southern Jiangsu and downtown Shanghai. Therefore, VOC emission control should be more effective in ozone concentration abatement.