Three-dimensional variational assimilation of Lidar extinction profiles: Application to PM2.5 prediction in north China

Abstract

The capability of assimilating aerosol optical properties (AOP) using the three-dimensional variational (3DVAR) data assimilation (DA) approach is developed for the Weather Research and Forecasting (WRF) model's data assimilation (WRFDA) system by introducing WRF-Chem-based AOP observation operator and its tangent linear and adjoint. The new AOP DA capability of WRFDA- Chem is firstly applied to assimilate Lidar extinction coefficient profiles in Beijing City from 26 February to March 5, 2019. Cycling strategy of Lidar DA is important to allow a substantial positive impact on surface PM2.5 forecast over the whole evaluation domain (D02, 100 times area of Beijing city covered by the eight Lidar sites) for the entire 24-h forecast range. Compared to the control run without aerosol DA, the mean RMSE reduction and correlation coefficient increase of surface PM2.5 forecast can be up to 15% for the one-day partial cycling strategy. Lidar DA exhibits a clear benefit in improving the vertical distribution of aerosols over Beijing when comparing to surface PM2.5 DA. With the very limited coverage of Lidar data, however, adding Lidar data above surface PM2.5 data (covering the whole D02) does not show the further improvement of surface PM2.5 forecast over D02.