Improving Arctic Polar Low Forecasting through FY-3D MWHS-II Radiance Assimilation during Winter 2020–2021

During the winter of 2020–2021, extremely cold air activities broke out over the Arctic, leading to occurrences of multiple Polar Low (PL) events over the Norwegian Sea, which significantly affected the weather in East Asia and North America. In this study, the Polar Weather Research and Forecasting Model (WRF) and the associated three-dimensional variational (3D-Var) data assimilation system are employed to investigate the effects of assimilating clear-sky radiance data from the Microwave Humidity Sounder-II (MWHS-II) onboard the Fengyun-3C (FY-3C) and Fengyun-3D (FY-3D) satellites on two PL cases. Simulation experiments at 9-km resolution with one-way nesting for dynamical downscaling to 3-km resolution over the Norwegian Sea during the forecast period are conducted. The results reveal that assimilating the FY-3D MWHS-II data produces smaller observation minus background (OmB) results in one-month statistical averages, achieving accurate thermal and dynamic analysis variations with lower biases in variational bias correction (VarBC). Direct assimilation of the FY-3D MWHS-II radiance data enhances forecasts of ocean surface winds, pressure, PL tracks, and coastal precipitation along western Norway. Furthermore, it provides stronger 850-hPa relative vorticity, larger differences between the sea surface temperature and 500-hPa temperature (SST – T₅₀₀), and more accurate vertical potential vorticity and static instability, consistent with the observed development of the PL in Case 1. For Case 2, benefits are also observed from assimilating the FY-3D data. However, simultaneous assimilation of both FY-3C and FY-3D data does not demonstrate a superior performance. This investigation highlights the importance of satellite radiance data assimilation in analyzing and forecasting of PL events.