Intercalibration of HY-2B SMR Using Double Difference Method Based on GPM GMI

• Published in: IEEE transactions on geoscience and remote sensing Vol. 62; pp. 1 - 18
• Main Authors: Wang, Shishuai, Yin, Xiaobin, Zhou, Wu, Bao, Qingliu, Li, Yan, Kong, Jian, He, Mingyao
• Format: Journal Article
• Language: English
• Published: New York IEEE 2024; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Brightness temperature; Calibration; Double difference (DD); Geographical distribution; Global Precipitation Mission (GPM) microwave imager (GMI); Haiyang-2B (HY-2B) scanning microwave radiometer (SMR); Incidence angle; Instruments; Intercalibration; Microwave imaging; Microwave measurement; Microwave radiometers; Microwave radiometry; Microwave theory and techniques; Oceans; Orbits; Periodicity; Radiative transfer; Radiometers; Satellite broadcasting; Surface radiation temperature; Uncertainty
• ISSN: 0196-2892; 1558-0644
• DOI: 10.1109/TGRS.2024.3499332

This article presents a study on the in-orbit calibration of the brightness temperature (TB) of the Haiyang-2B (HY-2B) scanning microwave radiometer (SMR), utilizing the Global Precipitation Mission (GPM) microwave imager (GMI) as a reference and employing the double difference (DD) method. The study specifically focuses on the calibration of seven SMR channels: 10.7 GHz-H, 10.7 GHz-V, 18.7 GHz-H, 18.7 GHz-V, 23.8 GHz-V, 37.0 GHz-H, and 37.0 GHz-V. To facilitate this, simulated TBs are calculated using either the National Centers for Environmental Prediction (NCEP) or WindSat data products as inputs for the radiative transfer model (RTM). The following conclusions are drawn from the study: 1) differences in Earth incidence angle (EIA) and center frequency preclude direct intercalibration between GMI and SMR; 2) when evaluating or calibrating the TBs of the SMR, it is essential to distinguish between ascending and descending orbits; 3) the different inputs of the RTM had minimal impact on the DDs and the NCEP can serve as the RTM input following the discontinuation of the WindSat data; 4) in fitting the DDs, first-order fitting proves to be more stable than second-order fitting; 5) the number and geographical distribution of matching datasets between GMI and SMR exhibit periodicity, with a recommendation to use at least three consecutive months of data for intercalibration; and 6) overall, a higher frequency correlates with greater uncertainty in intercalibration, with the 37.0-GHz-H channel demonstrating the largest uncertainty.