This paper presents a bottom-up methodology based on the local emission factors, complemented with the widely used emission factors of COPERT model and near-real-time (NRT) traffic data on road segments to develop a high temporal/spatial resolution vehicle emission inventory (HTSVE) for the urban Beijing area. The results show that the vehicle emissions of NO_x, CO, HC and PM were 10.54 × 10⁴, 42.51 × 10⁴ and 2.13 × 10⁴ and 0.41 × 10⁴ Mg respectively.

The aerosol adjoint module of the atmospheric chemical modeling system GRAPES–CUACE was developed, tested for its correctness, and used in a receptor–source sensitivity test. The results showed that controlling critical emission sources during critical time intervals on the basis of adjoint sensitivity analysis is much more efficient than controlling administrative specified regions during an experiential time period.

An ensemble optimal interpolation (EnOI) data assimilation technique is used to investigate the possibility of optimally recovering the spatially resolved emissions bias of BC. The inversed emission over China in January is 240.1 Gg, and annual emission is about 2539 Gg. Even though only monthly mean BC measurements are employed to inverse the emissions, the accuracy of the daily model simulation improves. We finds that EnOI is a useful and computation-free method to make top-down estimation.

Based on a vehicle emission inventory with high temporal–spatial resolution established by bottom-up methodology, this paper evaluates the performance of the CUACE model and investigates the vehicle emission contribution (VEC) to ambient NO₂ and PM_2.5 with an emission source sensitivity method. With good performance of pollutant concentrations, numerical simulation revealed that the mean VEC is 55.4 and 48.5 % for NO₂ and 5.4 and 10.5 % for PM_2.5 in July and December 2013 respectively.

No obvious changes were found in annual mean concentrations of major chemical components and PM10 in 2013, relative to 2012. But wintertime mass were quite different; approximately 60% of the winter mass increase from 2012 to 2013 can be attributed to severe meteorological conditions in the HBP area, and mass of chemical components exhibited a decline during 2006 to 2010, and then a rise till 2013. Coal-combustion was still the largest anthropogenic source of aerosol pollution in 2013 in China.

A comprehensive aerosol-cloud-precipitation interaction (ACI) scheme from emissions to precipitation has been developed under the CMA chemical weather modeling system GRAPES/CUACE. The ACI for January 2013 has been studied using this model. The interactive microphysical properties of clouds improve the precipitation, showing 24 to 48 % enhancements of threat score for 6 h precipitation in all regions and reduction of the regional mean bias of temperature by 3 °C in certain precipitation events.

Concentrations of PM10, PM2.5 and PM1 were monitored at 24 stations of CAWNET from 2006 to 2014. The average level of PM concentrations and relationships were investigated. Seasonal, interannual, diurnal variations of the PM were revealed. The meteorological factors effects on the PM were discussed. The highest particulate matter (PM) concentrations were observed at the stations of Xian, Zhengzhou and Gucheng, in Guanzhong and the Hua Bei Plain (HBP).

A new model, PLAM/h, has been developed and used in near-real-time air quality forecasts by considering both meteorology and pollutant emissions, based on the two-dimensional probability density function diagnosis model for emissions. The results show that combining the influence of regular meteorological conditions and emission factors together in the PLAM/h parameterization scheme is very effective in improving the forecasting ability for fog-haze weather in North China.

Solar radiation reaching the ground decreases about 15% in China 3JNS Region and by 20 to 25% in the region with the highest AOD.Aerosol cools the PBL atmosphere but warms the atmosphere above it leading to a more stable atmospheric, which causes a decrease in about 52% of turbulence diffusion and in about 33% of the PBL height, resulting in a positive feedback on the PM2.5 concentration within the PBL and the surface as well as the haze formation.