Effect of WRF-3DVAR Meteorological Data Assimilation on WRF-Chem Forecasts over East Asia

The online coupled chemistry-meteorology modeling (CCMM) used for air quality forecasting improves the forecasting performance by reflecting the interaction between meteorological and air quality variables. However, uncertainties in meteorological and chemical initial conditions cause uncertainties in the model forecasts. Meteorological data assimilation (metDA) can reduce the uncertainties of meteorological initial conditions in the CCMM, and further increase the air quality forecast accuracy.

In this study, Weather Research and Forecasting model coupled with Chemistry (WRF-Chem) and WRF three-dimensional variational data assimilation system (WRF-3DVAR) were used. To investigate the effect of metDA on the air quality forecasts, the experiment with metDA (DA) and that without data assimilation (noDA) were performed and compared. Additionally, the experiments with three different assimilation domains were conducted to examine the effect of assimilation domains on the air quality forecasts.

The root mean square error (RMSE) and bias of particulate matter (PM) concentration forecasts with respect to observations were smaller for the DA compared to the noDA. The forecast error growth of PM concentration was relatively small for the DA. The root mean difference total energy (RMDTE) of the upper meteorological variable, the RMSE of the near surface meteorological variable, and their growth were smaller for DA compared to noDA. The bias and RMSE of PM concentration forecasts and RMSEs of upper and near surface meteorological variables were small when the meteorological observations were assimilated in the outermost domain of the nesting domains. When metDA was applied over the outermost domain, the model synoptic weather was generally improved. This reduced the uncertainties of the initial and lateral boundary conditions of the inner domain that was the forecast verification domain.

Acknowledgements

This work was supported by a grant from the National Institute of Environment Research (NIER) funded by the Ministry of Environment (MOE) of the Republic of Korea (NIER-2022-01-02-076), a National Research Foundation of Korea (NRF) grant funded by the South Korean government (Ministry of Science and ICT) (Grant 2021R1A2C1012572), and Yonsei Signature Research Cluster Program of 2022 (2022-22-0003).