Preliminary Application Of Ensemble Optimal Interpolation Data Assimilation Method On O₃ Numerical Modeling In Xingtai

With the increase of pollution control,new air pollution situation has emerged in China,PM_2.5 pollution has decreased,and O₃ concentration has continued to increase.In order to improve the accuracy of model prediction,this paper proposes a pollutant prediction optimization scheme based on CAMx?Comprehensive Air Quality Model with extensions?.Combined with the observation data of the pollutant concentration of the observation station,the initial condition of the model O₃ prediction is subjected to the assimilation test using the EnOI method.details as follows:First,in order to provide an air quality model with an emission list that is closer to the real state,this study uses the 2017 as the base year to estimate the anthropogenic emissions inventory of Xingtai containing seven types of atmospheric pollutants by the emission factor method and the material balance algorithm.From the perspective of the contribution of various pollution sources to different pollutants,the focus of air pollution control is to reduce industrial emissions.Then,this study used the WRF,CAMx,and SMOKE models to build the Xingtai Air Pollutant Simulation System,and verified the simulation effect of the WRF model on meteorological elements and the CAMx model for the four typical months of Xingtai in January,April,July and October.The simulation effect of the temporal and spatial distribution of O₃ concentration changes.The CAMx mode has a low overall simulation with O₃,and the simulation effect on the daily maximum is not ideal,but it has a good simulation of the O₃ concentration change trend.Then we discussed the estimation method of background error covariance of EnOI method,and choose the overall prediction sample obtained during the long-term operation of the model to form the background error covariance field.From the distribution of the background error field high value area,the model simulates ozone.Some areas of Sichuan and Chongqing have the most uncertainties,followed by parts of Hubei,Hunan and Gansu.Sensitivity test analysis was carried out on the two parameters of the assimilation experiment,and the optimal localization scale and number of assimilation sites were 50 km and 390 stations.Assimilation test design In January,the analysis field after assimilation basically solved the situation that the O₃concentration value of the pattern background field in January was generally higher than the observed value and the simulated value when the O₃ concentration was greater than about 80?g/m³.The collective presents an underestimation.Finally,the sensitivity of the initial field in the CAMx mode O₃ concentration simulation is discussed.The assimilation experiment of the 8h forecast window is done.The optimization effect is discussed in Xingtai.In the initial condition assimilation test,the RMSE of the assimilation station and the inspection station decreased by 41.69%and 34.85%on average,and the RMSE decreased by 20.39%in the assimilation test of the 8h forecast window.The results show that the assimilation test can provide a more realistic situation for the air quality model.The initial condition and the improvement of the short-term forecast effect.