Comparison of the effect of different vegetation indices on land surface temperature values

Abstract

Land surface temperature (LST), which is a very important parameter in many fields such as ecology, hydrology and climate studies, can be monitored on a regional and global scale with satellite images. This study aimed to investigate the effects of different Vegetation indices (VI) on LST. The study focused on to generate LST maps over North-Rhine-Westphalia and Rhineland-Palatinate states of German using 23 August 2016 dated Landsat 8 OLI&TIRS data. Additionally, Sentinel-FCOVER data was used to examine the effect of VI on LST. The LST values were retrieved with Mono-window method, and Land Surface Emissivity (LSE) maps were calculated with NDVITHM method. Normalized Difference Vegetation Index (NDVI), Renormalized Difference Vegetation Index (RDVI), Soil Adjusted Vegetation Index (SAVI), Enhanced Vegetation Index (EVI) were selected for the study. Additionally, soil emissivity values were calculated from LUCAS (Land Use and Coverage Area frame Survey) and ASTER Spectral Libraries datasets. Finally, the accuracy of FVC (Fractional vegetation cover) and LST maps were evaluated. Although the best FVC result was achieved with RDVI, LST maps showed similar results for all selected vegetation indices. Thus, transects were extracted from the LST maps for different land cover categories and the results were compared to determine the differences.