Development of a Hybrid Algorithm for Temporal Normalization of Polar-Orbiting Satellite-Derived Land Surface Temperature

Land surface temperature (LST) is crucial in many global and regional scientific studies and applications. The observation time differs along the scan line due to the intrinsic scanning characteristics of instruments, making satellite-derived LSTs incomparable. Although many algorithms have been developed to address this issue, they have many limitations and uncertainties in application. On the basis of the temporal evolution of clear-sky LST, this study proposed a hybrid and practical algorithm with good applicability for the temporal normalization of satellite-derived LSTs. The proposed algorithm was applied based on Aqua Moderate-Resolution Imaging Spectroradiometer (MODIS) data across the contiguous United States (CONUS) in 2020 and mainly validated by cross-comparisons with the Geostationary Operational Environmental Satellite R-Series 16 (GOES-R16) Advanced Baseline Imager (ABI) LST product over each season and various land cover (LC) types. The normalized MODIS LSTs had a superior agreement with the GOES-R16 LSTs. Especially for the temporal differences (original observation time minus the reference time) between 0.5 and 1.0 h, the root-mean-square error (RMSE) and bias of the normalized LSTs were improved by 0.32-1.03 K and 0.30-1.27 K, respectively. These results demonstrate that the proposed hybrid method has advanced potential for the temporal normalization of polar-orbiting satellite LST.