Cross-Track Infrared Sounder Spectral Gap Filling Toward Improving Intercalibration Uncertainties.
The cross-track infrared sounder (CrIS) is a hyper-spectral infrared sounder with 2211 (full-spectral resolution) channels over three wavelength ranges. Due to its excellent calibration and high spectral resolution, CrlS radiances are also used as an infrared reference to check calibration accuracy of other narrow- or broadband instruments. However, there are spectral gaps between the CrlS bands, which impact the accuracy of intercomparison between CrlS and other instruments. To fill up the spectral gaps, this paper develops a new method to predict the CrlS gap channels. A training data set was first built based on the spectra selected from different seasons to represent different atmospheric and surface conditions. The principal component regression method is then developed to derive the prediction coefficients between the CrlS measured and the gap channel spectra. The prediction accuracies are around 0.005 K ± 0.18 mK, 0.06 K ± 0.45 mK, and 0.05 K ± 0.71 mK for the long-, middle-, and short-wave gap channels, respectively. It should note that the predicted channels are intended to be only used in the intercomparison purposes instead of providing exact information of these gap channels. When comparing the CrlS gap radiances with atmospheric infrared sounder and infrared atmospheric sounding interferometer, consistent results are observed between CrIS and other sounders. The further intercomparison results with visible infrared imaging radiometer suite and advanced baseline imager suggest that the predicted gap radiances are good enough for the intercalibration purposes and this method can be confidently applied in global space-based intercalibration system community.