The Application of the Principal-Component-Based Radiative Transfer Model for Satellite Intercalibration

Satellite based hyper-spectral instruments are currently used as reference sensors for the inter-calibration of GEO or LEO measurements in the Global Space based Inter-Calibration System (GSICS), and will be the critical part of future inter-calibration efforts towards achieving a rigorous SI traceability. Technical problems including spectral coverage difference, view geometry mismatch can greatly limit the inter-calibration accuracy and the application scope. The scene variability issue further imposes undesired uncertainties that complicates the correction for those errors. The Principal Component based Radiative Transfer Model (PCRTM) designed for a variety of hyper-spectral sensors can be used as a valuable tool to address those technical challenges. PCRTM can be used to simulate the errors in radiance caused by either the spectral sampling or the view geometry mismatch. PCRTM’s capability of doing cloud scattering simulation ensures the accurate quantification for scene dependent errors under all-sky conditions. We demonstrate that an accurate linear regression relationship between the errors and the corresponding spectral radiances can be established using the simulated results. PCRTM can be used for the simulation of hyper-spectral measurements in both the longwave and the shortwave region. Our study shows that the inter-calibration of broadband sensors such as MODIS using IR sounders including IASI, AIRS and CrIS can effectively reduce the spectral sampling errors to meet the requirement by the CLARREO mission. The error estimation and correction scheme has also been validated by the inter-calibration study using the real AIRS-MODIS data. The study on the CLARREO-VIIRS inter-calibration shows that the threshold of view geometry mismatch can be greatly raised after applying the correction for the off angle errors.