A modified land surface temperature split window retrieval algorithm and its applications over China

Due to the difficulties in correcting the influences of the atmosphere absorbability and the Earth surface emissivity diversification, the retrieval of LST (land surface temperature) from satellite data is a challenging task. In this paper, a modified Becker's split window LST inversion algorithm is developed for retrieving LST from the NOAA-16/17 AVHRR data. A new set of parameters for Becker's LST algorithm is proposed. The algorithm is developed from a surface brightness temperature dataset generated from the MODTRAN program, which uses a range of surface parameters and atmospheric quantities as inputs. The 10-day composites of the channels 4 and 5 brightness temperature data of NOAA-17 AVHRR (1-km resolution) are used to generate the clear-sky LST. As a validation of the algorithm, the retrieved LST is compared with MODIS LST of same period and area. The two LST products are found to be consistent, with the absolute difference being about 2.5 K for most areas. The NOAA retrieved LST is also compared with in-situ ground surface 0-cm measurements taken from 257 meteorological stations, which cover overall China area for the three periods of satellite observations. The comparison shows that the correlation between the retrieved LST and in-situ measurements is over 0.90 and the RMSE (root mean square error) is about 3.4 K.