Appraisal of radiative transfer model 6SV for atmospheric correction of multispectral satellite image towards land surface temperature retrieval

Land surface temperature (LST) is very important parameter for broad range of applications related to weather and climate, hydrology, etc. For LST, satellite could be a very promising solution, but need to be corrected especially for atmospheric noises present in bands such as in red and near-infrared (NIR) windows. The atmospheric noises occurred due to ozone, water vapor, and aerosols cause reduction in reflectance values of red and the NIR bands which makes the normalized difference vegetation index (NDVI) values smaller than their true value. NDVI considered as an important variable for emissivity estimation and any loss in reflectance causes error in LST retrieval. In this study, for atmospheric correction of satellite image, radiative transfer model 6SV (second simulation of satellite signal in the solar spectrum-vector) was used and then the corrected image was utilized for the LST retrieval using Landsat 8 satellite thermal and visible bands. Changes in the values are found in the atmospherically corrected images of NDVI, plant fraction, emissivity, and LST as compared to the atmospherically uncorrected images. Hence, the reported approach could be a better choice for retrieval of LST due to its robust nature.