The Effect Of Modified Meteorological Fields Simulation On The PM Simulation By WRF-Chem Model Over Valley Terrain

Air quality model is used to accurately simulate the concentration of particulate matter to quantify the contribution rate of various emission sources to air quality is the basis of in-depth control of air pollution and precise policies.Under the condition of constant emission source,accuracy of meteorological field simulation has an important impact on the simulation effect of air quality,thererfore,this study starts from improving the accuracy of meteorological field simulation to improve the simulation effect of ambient air quality.Firstly,WRF-Chem model was used to simulate the particle concentration in the main urban area of Tianshui City in December2019,using MYJ boundary layer parameterization scheme,USGS land use type data and NCEP reanalysis data,and the simulation ability of WRF-Chem model to simulate the meteorological field and particle concentration in valley cities was tested.Secondly,the effects of different meteorological fields on air quality simulation results are studied by setting up comparison schemes from three aspects:boundary layer parameterization scheme,the characteristics of the underlying surface and model initial field.Finally,a combination of the above results with better simulation results is selected to quantitatively evaluate the degree of improvement of the air quality simulation effect by the modified meteorological field simulation.The conclusions are as follows:(1)The near-surface meteorological field and particulate matter concentration simulated by WRF-Chem model were in good agreement with the observed values,and the emission source inventory combined with the local high-resolution inventory of Tianshui and MEIC inventory was adopted.The model simulation results are reliable.(2)Among the four boundary layer parameterization schemes,the MYJ scheme and QNSE scheme are better than YSU scheme and MYNN2.5 scheme in simulating near-surface temperature in the main urban area of Tianshui City.The MYNN2.5 scheme has the best simulation effect of the near-surface wind speed in the main urban area,followed by YSU scheme and MYJ scheme,and QNSE scheme has the worst simulation effect.The MYNN2.5 scheme has the best simulation effect for the near-surface particle concentration in the main urban area of Tianshui City,followed by MYJ scheme and YSU scheme,and QNSE scheme is the worst.(3)Among the two land use types,the simulation results of using USGS and Tianshui City land use type data respectively,and considering the urban canopy scheme(UCM)show that,using higher resolution land use type data of Tianshui City,the simulation effect of near-surface temperature in the main urban area of Tianshui City is modified,but the simulated near-surface temperature is still underestimated.There is no significant improvement in the simulation effect of near-surface wind speed and particle concentration.(4)The results of numerical simulation using NCEP-FNL and ERA5 reanalysis data respectively show that the simulated near-surface temperature,wind speed and particle concentration using ERA5 reanalysis data are more consistent with the observed values,and significantly improve the phenomenon that NCEP-FNL simulates low near-surface temperature.(5)Modeling results of different boundary layer parameterization scheme using ERA5reanalysis data show that:The QSNE-ERA5 scheme has the best simulation effect on near-surface temperature,and the MYNN2.5-ERA5 scheme has the best simulation effect on near-surface wind speed when the ERA5 reanalysis data is used to provide the initial meteorological field and boundary field.The simulation cross-check results of the QSNE-ERA5 scheme and the MYNN2.5-ERA5 scheme show that the QSNE-ERA5 scheme has the best simulation effect on the particle concentration,Compared with the baseline scenario,the correlation coefficient between the simulated PM₁₀concentration and the observed value in QSNE-ERA5 scheme increased from0.402 to 0.421,and the correlation coefficient between the simulated PM_2.5concentration and the observed value increased from 0.328 to 0.389.