Influence of land surface parameters and atmosphere on METEOSAT brightness temperatures and generation of land surface temperature maps by temporally and spatially interpolating atmospheric correction

Tendencies toward desertification or changes of the land surface can be detected by monitoring land surface temperature (LST), but accurate retrievals require good knowledge of the atmosphere and land surface parameters. Here, the effect of land surface emissivity, LST, and ground height on the error of LST associated with atmospheric correction is modeled for the thermal infrared (TIR) channel of METEOSAT using MODTRAN. The atmospheric conditions of midlatitude summer (MLS) and midlatitude winter (MLW) are considered. The results confirm that for accurate atmospheric corrections temperature and height variations have to be extended by an emissivity variation. LST maps for larger areas are generated using atmospheric corrections derived from EMCWF reanalysis and Digital Elevation Model (DEM) data. The corrections are spatially and temporally interpolated using Shepards' method and a model of the diurnal LST wave, respectively. The interpolations allow the generation of spatially smooth LST maps for any time of the day. Modeling the diurnal wave can partially compensate for the adverse effect of cloudiness and radiometric noise. (C) Elsevier Science Inc., 2001.