Spectrum calibration of the first hyperspectral infrared measurements from a geostationary platform: Method and preliminary assessment

Abstract

AbstractThe level‐1 (L1) radiance spectra from the first Geostationary Interferometric InfraRed Sounder (GIIRS) have been publicly available since January 2019. On account of the inherent observation characteristics, operation modes, and capabilities, a complete spectrum calibration method of GIIRS/L1 products is originally proposed to form the latest version (V3) algorithm for implementation. Particularly, four targeted improvements in three aspects are independently established: subsample location alignment to yield integrated interferograms in both forward and backward directions with almost zero phase, rough spectral scale unification as well as accurate spectral scale correction to resolve spectral non‐uniformity due to a seriously asymmetric configuration of focal plane array, and an additional double‐reflected compensation to mitigate the influence of non‐ideal onboard blackbody reference upon radiometric accuracy. Preliminary assessments from both domestic and international sources indicate that the spectral and radiometric accuracies of the measured spectra from the latest GIIRS/L1 V3 algorithm show a well‐behaved performance in both longwave (LW) and midwave (MW) bands, that is, lower than 10 ppm of spectral scale errors, which is of sufficient accuracy for numerical weather prediction use, and around 1 K for most uncontaminated channels within the LW band. However, non‐linearity correction of interferograms and spectral quality improvements, especially for the MW band, should be developed further. In general, a feasible solution of spectrum calibration for a hyperspectral sounder on geostationary platform is provided in detail for reference, which is expected to benefit users of GIIRS data as well as designers responsible for L1 data processing of other similar sensors.