Cloud-Cleared Radiances From Collocated Observations of Hyperspectral IR Sounder and Advanced Imager Onboard the Same Geostationary Platform

Abstract

The Geostationary Interferometric Infrared Sounder (GIIRS) onboard China’s Fengyun-4A (FY-4A) geostationary (GEO) meteorological satellite provides high-spectral-resolution infrared (IR) observations for targeted observing areas with high temporal resolution. Due to the high uncertainties in radiative transfer modeling of cloudy radiances, it is challenging to take full advantage of the thermodynamic information from GIIRS in all-sky conditions. The Advanced Geostationary Radiation Imager (AGRI) onboard the same platform provides a variety of cloud products with high spatial resolution. A bias-corrected optimal cloud-clearing (BCOCC) approach is introduced to generate GIIRS cloud-cleared radiances (CCRs) with the help of AGRI-collocated clear radiances (CLRs). The bias correction (BC) scheme is based on the inter-comparisons between GIIRS and AGRI for each field-of-view (FOV) and different scene temperatures. The BC method ensures the radiometric consistency between GIIRS and AGRI. Evaluations of GIIRS CCRs show that the mean biases are 0.09, −0.06, and 0.06 K when compared with the three AGRI IR bands, B12, B13, and B14. In addition, BCOCC significantly increases the data yields of successful CCRs by three times that of the optimal cloud-clearing (OCC) approach without BC. For 15 days from September 16–30, 2021, around 37% more GIIRS partially cloudy footprints than clear sky are cloud cleared successfully. The CCRs can be assimilated as CLRs in numerical weather prediction (NWP) models without worrying about the cloud impact. This study provides evidence of the importance of placing an advanced hyperspectral IR sounder and imager onboard the same GEO platform for better quantitative applications.