PM_2.5/PM₁₀ Assimilation And Application Based On The Surface Routine Observations And Satellite AOD

With the rapid development of modern social economy and the influence of human activities,the regional aerosol pollution is increasing year by year.The frequent occurrence of extreme pollution with PM_2.5.5 and PM₁₀0 as the primary pollutant causes the phenomenon of haze weather,which on human health and agricultural production.Based on the meteorological data and the WRF-Chem atmosphere coupling model,a three-dimensional variational assimilation system for the ground aerosol observation data and the satellite AOD product is established in this paper,and using model products to make statistics of background error covariance?BEC?.One control experiments and three assimilation experiments were designed to test the improvement effect of assimilation on aerosol simulation and prediction for a process of air pollution in December 15,2015.Based on the 48 and 24 hour forecast difference field data of the December 2015WRF-Chem model,the NMC method is used to calculate the covariance of the background error.The results show that the standard deviation of background error has obvious spatial distribution characteristics.The greater the aerosol emission rate in the region,the greater the standard deviation of the background error in the region.The Gauss function model is used to fit the horizontal auto-correlation coefficient.With the increase of the horizontal distance,the auto-correlation coefficient of each component gradually attenuates.Among them,the component of OIN attenuates fastest and the NO₃ component attenuates the slowest.In the vertical direction,the auto-correlation of each component attenuates rapidly with the increase of the height of the ground.Among them,the OIN component attenuates most quickly,which may be due to its composition in the dust or the dust as the main component,the vertical upward diffusion level is lower.By comparing the standard deviation of the standard deviation of the background error of the aerosol,it is found that the error standard difference between the two components of OIN and NO₃ is the greatest,while the error standard difference between EC and SO₄ is the least.Compared to different regions,with the gradual improvement of the pattern resolution,the standard deviation of the background error is gradually increased,the level correlation scale decreases,and the vertical correlation is weakened.Based on two observation data of ground aerosol concentration and satellite AOD product,a control experiment and three assimilation tests were designed for an aerosol pollution in December 15,2015.They are DA-AOD?only assimilation of satellite observations?,DA-SUR?only assimilation of ground observations?and DA-BOTH?Simultaneous assimilation of ground and satellite observations?.It is found that all assimilation experiments have different degrees of improvement on the initial field compared to the non assimilative control experiments.After data assimilation,the distribution and intensity of the pollution were all changed,and the assimilation was closer to the fact after data assimilation.By comparing with the actual situation,it is found that even the DA-AOD test that assimilates satellite observation data can effectively correct the problem of low concentration of model background field,especially for the large area of PM_2.5.5 in the northeastern part of Shandong Peninsula and the northern part of Jiangsu.DA-BOTH is the best experiment,by comparing with the initial field,the RMSE of PM_2.5.5 was reduced by 40.4?g/m³,the CORR was increased by43.7%;the RMSE of PM_2.5.5 was reduced by 68.5?g/m³,the CORR was increased by56.4%.The 24 hour prediction experiment based on the initial field after assimilation shows that the improved effect of the three assimilation tests can last to more than 15hours.Among them,the improved effect of DA-BOTH test is the best.Within the first12 hours,the average RMSE of PM_2.5.5 was reduced by 16.7?g/m3,the average CORR was increased by 32.4%;the average RMSE of PM₁₀0 was reduced by 23.3?g/m3,the average CORR was increased by 28.9%.The improvement effect of the DA-AOD test and the DA-SUR test is basically equal after 6 hours.Within 6 to 12 hours,the average RMSE of the DA-AOD test was reduced by 4.8?g/m³,the average CORR of the DA-AOD test was increased by 22.1%;the average RMSE of the DA-SUR test was reduced by 5.7?g/m³,the average CORR of the DA-SUR test was increased by 25.8%.