LST Estimation Over Rugged and Heterogeneous Areas Using TIR Data

Resumen

Land Surface Temperature (LST) is an important indicator of earth-surface balances and it is a key parameter in modern climate studies. The use of satellite remote sensing represents a significant progress in order to improve our knowledge of global environment. However, it introduces uncertainty in satellite-derived data generated from spatial heterogeneities and complex topography. These circumstances reduce its applicability to flat and homogeneous areas, without taking into account a wide range of surface types. Multi-resolution at coincident height, time and nadir observations have been used with the purpose of studying the uncertainties in satellite-derived data over rugged and heterogeneous surfaces from Tenerife Island. So, we have used information from ASTER and MODIS, onboard TERRA platform. Information from surface topography has been obtained from a digital elevation model (dem) with an accuracy of 3", dividing each scene into 8 layers according to its elevation. LST has been retrieved using two different separation methods, TES and TISIE, for ASTER and MODIS data, respectively. The comparison between LST derived with and without dem data shows significant mean differences. For ASTER/TES data we have obtained differences ranging from 0.14 o C to 5.54 o C. For MODIS/TISIE data this range raise up to values from 1.36 o C to 8.03 o C. A further study of these differences can lead us to an improvement in LST estimation from satellite data and a wider application of remote sensing information over complex surfaces.