Exploring The Meteorological Impact Of Winter PM_2.5 Pollution In The Yangtze River Delta Region Based On Numerical Simulation And Data Assimilatio

In recent years,due to the implementation of air pollution control measures,the annual average PM_2.5concentration in the Yangtze River Delta（YRD）region has shown a downward trend.However,regional PM_2.5pollution events caused by unfavorable meteorological conditions in winter still occur occasionally.Meteorological conditions are the main cause of PM_2.5pollution when the emission source remains unchanged.Therefore,the accuracy of near-surface meteorological elements simulation has an important impact on PM_2.5simulation results.This work is based on the statistics of meteorological elements of polluted winter days in the YRD from 2017 to 2021.Then,the WRF-Chem model was used to simulate a typical PM_2.5pollution process in the YRD region from January 12 to 16,2019.Through sensitivity test,the influence of surface meteorological factors（temperature,wind speed,relative humidity）on the PM_2.5concentration in this area was quantitatively analyzed.Finally,the impact of the meteorological data assimilation and air quality simulation application in the YRD region were studied by using meteorological observation data and WRF-Chem with FDDA（four-dimensional data assimilation）to evaluate the improvement of meteorological field and PM_2.5pollution simulation.The main conclusions are as follows:（1）Air pollutant concentration in the YRD region showed a decreasing trend from2017 to 2021,and PM_2.5,PM₁₀and CO similar decreasing rates.The seasonal variation of particulate matter concentration was still in the order of winter>autumn>spring>summer.The average PM_2.5concentration in winter decreased by 21.89μg·m^-3.It shows that the air pollution in winter in YRD region has been improved obviously.The relative humidity of pollution days and clean days in the YRD region is 72%-82%and 68%-78%,respectively.The higher relative humidity on pollution days is favorable for increasing PM_2.5through hygroscopic growth.The average temperature ranges from 4℃to 8℃,and the temperature on the pollution day is about 1℃lower than that on the clean day.The lower temperature is favorable for maintaining more stable boundary layer and thus high concentration of air pollutants in the near surface layer.The wind speed was 1-3m·s^-1.On pollution days,north wind and northwest wind dominated the sea to the east,while north wind and northeast wind dominated the inland.The wind speed on the clean days was stronger and wind directions were from northeast for the whole region.In this case,clean air from the sea were blowed to inland.Conversely,lower wind speed occurred on pollution days and was conducive to the accumulation of pollutants.The boundary layer height on pollution days is about 300-600 meters,while that on clean days is about 300-800 meters.The low boundary layer height is conducive to the accumulation of PM_2.5in the lower layer and hinders the diffusion and dilution of pollutants to the upper atmosphere.（2）Six groups of sensitivity experiments were conducted for the pollution process occurred during January 12 to 16,2019 with WRF-Chem simulation to study the contribution of perturbed near-surface meteorological elements to the simulation of the PM_2.5concentrations in the three stages during the air pollution period.It was found that temperature,relative humidity and wind speed had the greatest influence on PM_2.5concentration,which were-1.31%,-2.82%,2.20%,2.97%,4.68%and 8.57%,respectively.Among the effects of relative humidity,temperature and wind speed on the main components of particles,NO₃^-concentration has the most obvious effect,and the spatial distribution characteristics of NH₄⁺and NO₃^-concentration are consistent and the spatial distribution characteristics of BC and OC were similar.（3）A data assimilation experiment,in which the weather observation of the surface automatic stations in the region were assimilated,was carried out for the January 12 to 16,2019 air pollution events of January 12 to 16,2019 with WRF-Chem coupled with FDDA.Data assimilation reduced relative humidity by 9.68 and it had a large impact in the high-humidity area in southeast Anhui province.Temperature assimilation showed similar results with RMSE decreased by 1.02℃.It effectively improved the simulation of the low temperature conditions in the northern part of the study area.The wind speed was also improved significantly after data assimilation and its RMSE was decreased by 0.35m·s^-1.After data assimilation,the accuracy of the simulated PM_2.5concentration was significantly improved in eastern Anhui and southern Jiangsu with its RMSE decreased by 9.17μg·m^-3.This study strongly confirms a need to assimilate weather observation for simulating,forecasting and controlling air pollution.