Enhancing the Fast Radiative Transfer Model for FengYun‐4 GIIRS by Using Local Training Profiles. Issue 22 (19th November 2018)

• Record Type: Journal Article
• Authors: Di, Di; Li, Jun; Han, Wei; Bai, Wenguang; Wu, Chunqiang; Menzel, W. Paul
• Is Part Of: Journal of geophysical research. Volume 123:Issue 22(2018)
• Page Start: 12, 583
• Page End: 12, 596
• Publication Date: 2018-11-19
• Subjects: radiative transfer model -- RTTOV -- FengYun‐4; Atmospheric physics -- Periodicals; Geophysics -- Periodicals; 551.5
• DOI: 10.1029/2018JD029089
• Languages: English
• ISSNs: 2169-897X

Abstract

Abstract: With the successful launch of FengYun‐4A (FY‐4A), the first satellite in a new Chinese geostationary weather satellite series (FY‐4 series), which carries a high spectral resolution infrared (IR) sounder called GIIRS (Geosynchronous Interferometric Infrared Sounder), and vertical atmospheric profiles can be obtained frequently at the regional scale. A fast radiative transfer model is a key component for quantitative applications of GIIRS radiance measurements, including deriving soundings in near real time for situation awareness and radiance assimilation in numerical weather prediction models. The weighted least squares method on enhancing the accuracy of RTTOV (Radiative Transfer for TOVS) for GIIRS is developed. Besides, currently, fast radiative transfer models for IR sensors are based on global training profiles, since GIIRS is targeted for regional observations; it is beneficial for local weather related applications using local training profiles, which better represent the characteristics of that weather regime. A local training profile data set has been developed for GIIRS using the RTTOV approach, comparisons with line‐by‐line radiative transfer model indicate that weighted least squares method provides better accuracy (smaller root‐mean‐square error) in the brightness temperature simulation for the middlewave band of GIIRS than the ordinary least squares method, and the local training profiles have further remarkable improvements on brightness temperature simulation over the global training profiles, especially for GIIRS longwave band. The methods can be applied to RTTOV development for other IR sensors onboard the geostationary satellites. Key Points: The weighted least squares method improves the accuracy of RTTOV GIIRS simulation over the ordinary least squares method Local training profiles improves GIIRS simulation over the global training profiles, which is beneficial to local weather related applications The methods can be applied to develop the fast radiative transfer models for sensors onboard the geostationary weather satellites