Land surface temperature retrieval from MODIS and AMSR-E on the Tibet Plateau

Abstract

A simple linear regression algorithm to retrieve LST was presented in this paper. AvgSurfT from GLDAS was used to fill up the data gaps of MODIS LST due to cloudiness in the Tibet Plateau. The objective of our algorithm is to establish multi-liner regress equations between brightness temperature of AMSR-E and the combined temperature from MODIS and GLDAS under different land cover types, and then we retrieved LST using AMSR-E between brightness according to the coefficients obtained from regress equation. The retrieved LST and MODIS LST were examined to evaluate the algorithm. It was found that the retrieved results were good under the surface type categories of shrub, crop and grass due to the minimal variety of the land surface emissivity.