A Novel Methodology for Downscaling Remotely Sensed LST Data

Remotely sensed high spatial and temporal resolution data are required for various fine-scale phenomena investigations such as Urban Heat Island (UHI) studies. However, the present availability of satellite-retrieved data is limited in both, temporal and spatial resolution. Where-as sensors such as Moderate Resolution Imaging Spectroradiometer (MODIS) or Advanced Very High Resolution Radiometer (AVHRR) provide an adequate temporal resolution of daily overpasses, the low spatial resolution of approximately 1km is less suitable for investigations at the urban scale. In contrast, sensors such as Landsat 8 provide an appropriate spatial resolution of 30m, however the temporal coverage ensues within a 16 days period. In order to bridge this gab, the present work pursuits the approach of developing a methodology for downscaling MODIS coarse resolution Land Surface Temperature (LST) to UHI reasonable finer spatial resolution using the relationship of various land-use / land-cover (LULC) types and LST.