A Study On Effects Of Satellite Based AOD Data Assimilation On Haze Numerical Forecast

Nowadays, The frequencies of the occurrence of haze weather in China increases evidently, the duration of haze grows gradually, and the consequences worsen by degrees. It is of great value socially and economically to capture the evolutions and changes of the air pollution by making a fully use of atmospheric chemical model. As the main way to improve the prediction accuracy of atmospheric chemical model, data assimilation method has been widely used. By now, satellite data has significant advantages in terms of spatial extent, temporal and spatial resolution, consistency, and economic benefits. So how to effectively and effectively use the satellite based AOD data to improve the haze numerical forecast is of great significance in research. A WRF-Chem/GSI 3D-VAR assimilation system wasapplied to assimilate MODIS and FY-3A/MERSI AOD separately in haze cases over China, and the model results were compared and analyzed based on satellite data and ground observation data. The main research results and conclusions are as follows:The WRF-Chem managed good simulations of air pollution processes, it is capable of simulating of meteorological processes and chemical processes simultaneously. AOD data assimilation by GSI had positive impacts on the prediction of main factors of air quality in haze weather such as AOD, PM2.5 and PM10, and the improvements got better as the thinning methods of AOD data during data assimilation got thinner in this case.It showed that assimilating AOD data can significantly improve the initial field, as the analysis field after the data assimilation had a more accurate description of the initial field. AOD data assimilation has a positive effect on the subsequent forecast of AOD, PM2.5, PM10, and the strength and distribution of each variable were improved reasonably.From the statistical analysis point of view, the overall effect of the assimilation test is better than the control test, and satellite AOD data assimilation has obvious regional and individual characteristics by distinctive AOD data for the improvement of particulate matter prediction. MODIS assimilation test and FY-3A / MERSI assimilation test showed advantages in PM2.5 and PM10 prediction, respectively.AOD data assimilation can significantly improve the forecast accuracy of haze weather events. MODIS and MERSI assimilation experiments showed some differences in the accuracy of PM2 5 and PM10 prediction. For different satellite AOD data, the assimilation of AOD data had different effects on the distribution and intensity of AOD in the initial field correspondingly, which also led to different results in the numerical forecast. In this case, The MODIS AOD DA experiment proved an advantage in PM2 5 forecast while FY-3A/MERSI in PM10. Overall, the WRF-Chem-GSI assimilation forecast system has broad prospects for air quality forecast and improvement.