Land Surface Temperature Retrieval from the Medium Resolution Spectral Imager Thermal Data

A single channel land surface temperature (LST) retrieval algorithm named Single Channel Water Vapor Dependent (SCWVD) method was presented for Medium Resolution Spectral Imager (MERSI) thermal infrared band aboard FengYun-3A (FY-3A) satellite. Water Vapor Content (WVC) is the only input parameter in the algorithm assuming the surface emissivity is known. NCEP reanalysis monthly mean datasets were used to develop the SCWVD algorithm. Some tests, including global numerical simulations and validations with both in-situ measurements and MODIS LST product at Lake Tahoe, USA, were carried out to evaluate the algorithm performance. The simulation results showed that the LST could be estimated by the SCWVD algorithm with the root mean square error (RMSE) less than 0.8 K. In the inversion test, MERSI Level 2 water vapor product was employed, and the MERSI band emissivity was calculated using the MODIS band 31 and 32 emissivity by an empirical expression. The results show that the difference between the retrieved MERSI LST and the in-situ measurements is less than 1 K in most situations. The comparison with the MODIS LST products (V5) shows that the RMSE is about 2.3 K.