Intercomparison of methods for estimating land surface temperature from a Landsat-5 TM image in an arid region with low water vapour in the atmosphere

This study compares the methods for retrieving the land surface temperature (LST) (T-s) from Landsat-5 TM (Thematic Mapper) data, including the radiative transfer equation (RTE) method, the mono-window algorithm (MWA) and the generalized single-channel (GSC) method in an arid region with low atmospheric water vapour content. In addition, T-s calculated without atmospheric correction of TM band 6 is also assessed. The intercomparison is divided into two parts. The first part is applying the methods at the Biandukou site (100 degrees 58' E, 38 degrees 16' N, elevation = 2690 m) and the second part is applying them at Binggou (100 degrees 13' E, 38 degrees 42' N, elevation = 3400 m) and Arou (100 degrees 27' E, 38 degrees 36' N, elevation = 2960 m) sites. Results demonstrate that these methods provide acceptable accuracies at the Biandukou site. At this site, GSC generates nearly the same accuracy as RTE; MWA estimations are slightly less accurate than RTE and GSC; estimations without atmospheric correction of TM band 6 exhibit the largest errors. On the other hand, MWA is a good choice for retrieving the LST at Binggou and Arou sites. In cases where the meteorological parameters are unavailable, it is an alternative option to calculate T-s directly from TM band 6 image without atmospheric correction at these two sites.