Bias Characteristics and Bias Correction of GIIRS Sounder onboard FY-4A Satellite for Data Assimilation

The geostationary interferometric infrared sounder (GIIRS) is the first hyperspectral infrared (IR) sounder onboard the geostationary meteorological satellite—FengYun-4A satellite. The GIIRS can provide atmospheric state information with a high temporal, spatial, and spectral resolution that could enhance the prediction of convective-scale regional models. This paper presents a panoramic analysis of the bias characteristics of the GIIRS with the high-resolution regional model WRF and its assimilation system WRFDA. Results reveal that the bias (observations minus simulations, or O−B) and standard deviation of long-wave infrared (LWIR) channels are generally smaller than those of middle-wave infrared (MWIR) channels, and both have some contaminated channels. The diurnal variation of biases is relatively unremarkable, and the O−B bias has a weak relationship with the satellite zenith angle. However, the biases in all selected channels are related to the value of brightness temperature observations and the scanning position of the satellite. In addition, the spatial distribution characteristics of the bias are relevant to the bias characteristics impacted by the scanning position. After re-calibration of the GIIRS in 2020, the quality of the GIIRS observation data is significantly improved compared with that in 2019. Based on this, further bias correction tests are carried out using correction factors relevant to the scanning position. Results reveal that both the systematic errors (bias) of O−B and O−A have been significantly reduced, implying the effectiveness of the bias correction scheme. This study provides implications for GIIRS radiance assimilation in global/regional models.