Rainfall Rate Measurement for Advanced Meteorological Imager of the GEO-KOMPSAT-2A Satellite

An operational rainfall rate (RR) algorithm for the advanced meteorological imager (AMI) on the GEO-KOMPSAT-2A (GK2A) satellite has been developed. This algorithm exploits a priori information, including rainfall data from the global precipitation measurement (GPM) dual-frequency precipitation radar (DPR) and infrared (IR) brightness temperature (TB) from GK2A. The performance of the RR algorithm is enhanced by incorporating a priori information that encompasses a wide range of precipitation systems. Additionally, retrieval accuracy can be improved by distinguishing between physically different precipitation systems during the retrieval process. To classify precipitating clouds, the RR algorithm uses brightness temperature differences between IR channels, accounting for the diverse radiative characteristics resulting from various hydrometeor and cloud thickness distributions. Consequently, the RR algorithm categorizes five types of precipitating clouds (one shallow and four nonshallow) and separates the databases into four latitudinal bands to capture regional variations. A Bayesian approach was applied to invert TB values from five IR channels to RR, based on a priori databases constructed using one year of collocated DPR and AMI data. The RR algorithm’s estimates were compared with those from DPR and GPM microwave imager over two months and twelve typhoon cases. The results indicate that the RR algorithm meets the operational accuracy requirement, with a bias of 9 mm/h at 10 mm/h. Additional validation with the ground radar observations over the Korean Peninsula confirmed that the retrieval biases were within the accuracy requirement.