Publications

Li, N; Xu, JH; Li, X; Qin, BX; Wang, YP; Fu, DJ; Zhong, KW; Qin, ZH (2025). A Novel Land Surface Temperature Retrieval Algorithm for SDGSAT-1 Images. IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 63, 5000218.

Abstract
Land surface temperature (LST) is a crucial parameter influencing Earth-atmosphere interactions and energy balance processes. The Sustainable Development Goals Science Satellite 1 (SDGSAT-1) was recently launched to support the realization of the United Nations Sustainable Development Goals (SDGs), which provides worldwide three-spectrum wide-swath, high-resolution, and high-sensitivity thermal infrared (TIR) images. The objective of this study is to develop a modified three-channel split-window algorithm incorporating atmospheric water vapor content (W-TCSW) for LST retrieval from SDGSAT-1 images. This algorithm was developed from the existing split-window (SW) form. The parameters of the algorithm were determined based on the MODerate resolution atmospheric TRANsmission (MODTRAN) simulation results of 946 Thermodynamic Initial-Guess Retrieval (TIGR) atmospheric profiles. The W-TCSW algorithm was comprehensively compared with the SW and three-channel SW (TCSW) algorithms. The retrieval results of the three algorithms were validated with simulated datasets and in situ measurements from the Heihe Watershed Allied Telemetry Experimental Research (HiWATER) sites in China and the Surface Radiation Budget Network (SURFRAD) sites in USA. The SDGSAT-1 data retrieved by the W-TCSW algorithm was also intercompared with Landsat and ECOSTRESS LST products. The W-TCSW algorithm demonstrated the highest accuracy among the three retrieval algorithms (SW, TCSW, and W-TCSW). The influences of atmospheric water vapor content (AWVC) and land surface emissivity (LSE) as well as land use and land cover (LULC) on retrieval algorithms were discussed in a long-term time series. This study introduces a novel LST retrieval algorithm considering AWVC for SDGSAT-1 images and elucidates comprehensive validation and comparative assessment, expanding the application of high-spatial resolution TIR remote sensing data.

DOI:
10.1109/TGRS.2024.3514359

ISSN:
1558-0644