Land surface temperature estimates in the Tibetan Plateau from passive microwave observations

A new model for land surface temperature (LST) estimation in the Tibetan Plateau using passive microwave observations was presented. The new LST retrieval model was developed based on the strong linear relationship between the 36.5 GHz vertical polarized brightness temperature observations from Advanced Microwave Scanning Radiometer (AMSR-E) and the topsoil measurements from three networks named CAMP/Tibet, Maqu, and Naqu, established in the Tibetan Plateau. These networks were located in regions of different climatic conditions and vegetation cover. Then the new LST retrieval model was validated by ground measurements from the three networks and was also compared with the well-know land parameter retrieval model (LPRM). The results show that the new model can significantly improve the estimation accuracy of LST compared with LPRM in the Tibetan Plateau, and better results are achieved at the AMSR-E descending pass. The LPRM underestimates the LST at AMSR-E descending pass while it overestimates the LST at AMSR-E ascending pass in the Tibetan Plateau.