A SPLIT WINDOW ALGORITHM FOR ESTIMATING LAND-SURFACE TEMPERATURE FROM SATELLITES

The use of multichannel approaches has led to significant improvements in the accuracy of remotely sensed surface temperature determination. Nevertheless, even if with different targets from land versus ocean applications, some tasks have required further investigations. In this work a general multichannel algorithm, taking separately into account atmospheric water vapour, surface emittance and air-surface temperature difference, is described; thus, it results an extension of the split-window algorithm and is able to give a land surface temperature assessment by means of a formalism valid for a wide range of atmospheric water vapor and surface emittance. The surface-emittance effect has been analyzed in detail, making a comparison with other proposed algorithms; namely the knowledge of surface emissivity and its variation with wavelength has been considered. The algorithm has been set up by accurate simulations of NOAA satellite measurements. An analysis has been also carried out in order to study the feasibility of a remote sensing mission for detecting land surface temperatures for mid latitude regions.