Impact of data assimilation and aerosol radiation interaction on Lagrangian particle dispersion modelling. (15th February 2021)

• Record Type: Journal Article
• Authors: Jia, Mengwei; Huang, Xin; Ding, Ke; Liu, Qiang; Zhou, Derong; Ding, Aijun
• Is Part Of: Atmospheric environment. Volume 247(2021)
• Publication Date: 2021-02-15
• Subjects: Lagrangian particle dispersion modelling -- WRF-FLEXPART -- Eastern China -- Data assimilation -- Aerosols-PBL feedback; Air -- Pollution -- Periodicals; Air -- Pollution -- Meteorological aspects -- Periodicals; 551.51
• DOI: 10.1016/j.atmosenv.2020.118179
• Languages: English
• ISSNs: 1352-2310

Abstract

Abstract: Lagrangian particle dispersion models (LPDMs) have been widely used in air pollution studies. However, substantial uncertainties still exist in LPDM simulations due to biased meteorological data, especially under stagnant and highly-polluted conditions. In this work, to better investigate the source contribution and formation of winter haze pollution in eastern China, we conduct a sensitivity study of WRF-FLEXPART by using different reanalysis data, applying observational meteorological nudging, and considering aerosols' radiative feedback on meteorology. We find that simulations driven by reanalysis datasets generally underestimate pollutant concentration, especially during periods with heavy haze pollution. The underestimation is directly caused by overestimated planetary boundary layer (PBL) height and lower PBL horizontal wind speeds. By assimilating meteorological data from surface and radiosonde observation, the WRF model can well represent the PBL dynamics and wind fields, especially those near the ground surface, which then substantially improves particle tracing in the LPDM. In addition, by including aerosols' radiative feedback in the WRF-Chem model, which significantly influences PBL evolution, the biases between LPDM modelling and observations are notably narrowed, particularly when the haze pollution is severe. Quantitatively, the accuracy increase of the simulations with aerosols' radiative effect accounted for 48% of the improvement produced by assimilating meteorological data. Overall, meteorological input is of great importance in LPDM modelling. In regions with intensive pollution like China and India, applying observational data assimilation or considering the feedbacks of aerosols to meteorology serve as an effective way to reduce the biases of LPDMs and better understand the source contributions as well as the formation and accumulation of pollution. Highlights: LPDMs driven by reanalysis datasets underestimate pollutant concentration. The underestimation can be minimized by assimilating meteorological data. Considering the feedbacks of aerosols to meteorology increases the accuracy of LPDMs.