An algorithm for retrieving land-surface temperature from modis data—A case study of Northern Hebei, China

The requirement for land surface temperature (LST) in environmental studies and resource management makes the remote sensing of LST an important topic. As a result, a number of studies about the development of methodologies for retrieving LST have been carried out. This paper attempts to use both the split-window algorithm and the Linear Spectral Mixed Model for computing vegetation fraction, in order to improve the algorithm to retrieve the land surface temperature from Moderate-resolution Imaging Spectroradiometer (MODIS) data. Compared results of surface temperature of the profile, the LST retrieved by our algorithm ranges from 298K to 308K, which have very consistent trend of temperature distribution with the MODIS LST product. They have very good correlation, and the correlation is significant at the 0.01 level.