The impact of assimilating FY-3E MWHS-2 within 3DEnVar with surface humidity control variable on the forecast of Typhoon Doksuri

In this study, FY-3E MWHS-2 radiance data are assimilated using the Hybrid Three-Dimensional Ensemble Variational (3DEnVar) data assimilation (DA) system with surface humidity (SH) coupled as an extended control variable to investigate the impact on the forecast and analysis of typhoon Doksuri. The efficiency of adjusting low level humidity is confirmed by a single temperature observation test. Nine experiments were conducted to assess the impact of coupled versus uncoupled SH configurations on the data assimilation system performance, and the effect of different weights assigned to the flow-dependent background error covariance on forecast accuracy. The results show that the assimilation of FY-3E MWHS-2 data within the 3DEnVar system coupled SH significantly improves the accuracy of typhoon forecasts. This improvement is particularly evident in the forecasted track, central pressure, and maximum wind speed. In this study, the optimal weight of the flow-dependent background error covariance was determined to be 50 %, which resulted in the lowest average errors in track (36.73 km), central pressure (17.37 hPa), and maximum wind speed (−11.56 m/s). Moreover, the DA system coupled SH demonstrated better performance in reducing the root mean square error (RMSE) and BIAS across the zonal component of wind (U-wind), meridional component of wind (V-wind), temperature, and specific humidity. Additionally, the introduction of SH has positively influenced the simulation of precipitation distribution and intensity, it specifically improves the simulation accuracy for precipitation events with a 24-h accumulated amount exceeding 25 mm.