Improvement of COMS Land Surface Temperature Retrieval Algorithm

Land surface temperature (LST) is a key environmental variable in a wide range of applications, such as weather, climate, hydrology, and ecology. However, LST is one of the most difficult surface variables to observe regularly due to the strong spatio-temporal variations. So, we have developed the LST retrieval algorithm from COMS (Communication, Ocean and Meteorological Satellite) data through the radiative transfer simulations under various atmospheric profiles (TIGR data), satellite zenith angle (SZA), spectral emissivity, and surface lapse rate conditions using MODTRAN 4. However, the LST retrieval algorithm has a tendency to overestimate and underestimate the LST for surface inversion and superadiabatic conditions, respectively. To minimize the overestimation and underestimation of LST, we also developed day/night LST algorithms separately based on the surface lapse rate (local time) and recalculated the final LST by using the weighted sum of day/night LST. The analysis results showed that the quality of weighted LST of day/night algorithms is greatly improved compared to that of LST estimated by original algorithm regardless of the surface lapse rate, spectral emissivity difference ( De) SZA, and atmospheric conditions. In general, the improvements are greatest when the surface lapse rate and De are negatively large (strong inversion conditions and less vegetated surface).