Publications

Liu, WH; Cheng, J; Wang, Q (2023). Estimating Hourly All-Weather Land Surface Temperature From FY-4A/AGRI Imagery Using the Surface Energy Balance Theory. IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 61, 5001518.

Abstract
Thermal infrared (TIR) observations from geostationary satellites enable the retrieval of the diurnal variations in land surface temperatures (LSTs) linked to the land-atmosphere energy exchange and the water cycle. However, cloud cover obstructs TIR signals and leads to missing gaps that generally account for more than 50% of TIR LST maps. This study proposed an effective method to estimate hourly all-weather LST from the Advanced Geosynchronous Radiation Imager (AGRI) data under the framework of surface energy balance (SEB) theory. An improved temperature and emissivity separation algorithm was first used to obtain the high-quality clear-sky LST, which plays a decisive role in ensuring the accuracy of recovered cloudy-sky LST. Then, we proposed a unique way to solve the temperature difference (Delta LST) between the cloudy-sky LST and hypothetical clear-sky LST caused by cloud radiative effects (CREs). The bias and root mean squared error (RMSE) of the estimated AGRI hourly cloudy-sky LST are 0.10 K and 3.71 K during the daytime, and -0.20 K and 2.73 K during the nighttime, respectively. The overall bias (RMSE) of the estimated AGRI all-weather LST is 0.02 K (2.84 K). The estimated hourly all-weather LST not only captures the rapid variation in diurnal LST but is also promising for temporal upscaling. The temporally upscaled daily mean LST shows a bias (RMSE) of 0.03 K (1.35 K). This study provides a promising solution to generate diurnal hourly all-weather LST for AGRI and other geostationary satellites.

DOI:
10.1109/TGRS.2023.3254211

ISSN:
1558-0644