A spatial downscaling procedure of MODIS derived actual evapotranspiration using Landsat images at central Greece

Abstract

In this study, the Surface Energy Balance Algorithm for Land (SEBAL) was used to derive daily actual evapotranspiration (ET_a) distributions from Landsat and MODIS images separately. The study area is the Lake Karla basin in Thessaly, Central Greece. Meteorological data from the archive of Center for Research and Technology, Thessaly (CERETETH) have also been used. The methodology was developed using satellite and ground data for the period of summer 2007. Landsat and MODIS imagery were combined in order to have data with high temporal and spatial resolution (downscaling). The downscaling technique applied is the output downscaling with regression between images. This technique disaggregates imagery by applying linear regression between two MODIS products to the previous or subsequent Landsat product. After the calculation of a first order linear regression between two MODIS-derived ET_a maps the next step is the regression to the ET_a map derived from the prior Landsat image to predict the disaggregated subsequent Landsat ET_a map. The results are satisfactory, giving the general trend of ET_a derived from the original SEBAL procedure.