Spatial Sharpening of Land Surface Temperature for Daily Energy Balance Applications

Abstract

Daily high spatial resolution assessment of actual evapotranspiration is essential for water management and crop water requirement estimation under stress conditions. The application of energy balance models usually requires satellite observations of radiometric surface temperature with high geometrical and temporal resolutions. By now, however, high spatial resolution (~ 100 m) is available with low time frequency (approximately every two weeks); at the opposite daily acquisition are characterised by poor spatial resolution. The analysis of vegetation index (VI) and land surface temperature (LST) spatial relationship, shows in substance a scale invariant behaviour; this consideration allows the application of spatial sharpening algorithms of thermal data, by means of a combination of high spatial resolution data in VIS/NIR range with high temporal acquisition on TIR. In this paper, a sharpening algorithm was applied using the thermal bands of MODIS (MOderate resolution Imaging Spectroradiometer) and vegetation indices derived by ASTER (Advanced Spaceborne Thermal Emission and Reflection Radiometer) sensor; the choice of this sensors is justified by the simultaneous acquisition time. The results of this sharpening process was firstly compared against LST estimation (at the same spatial resolution) by means of the ASTER simultaneous data; then the derived high spatial resolution LST distribution was used in order to investigate the effect of the disaggregation on the outputs of surface energy balance models. The above described application was performed on a Sicilian study area.